Research on Reading Development of Adult English Learners: An Annotated Bibliography -- Alphabetical Listing

The nature of schema and its application to the reading process. The researchers synthesized several models of schema theory to give a unified overview of the importance of schema in both l1 and L2 reading.

Generally, schemata are knowledge structures. The researchers here propose that they are organized hierarchically, such that larger, more general categories subsume smaller, more specific categories. For example, under the category Òa restaurantÓ one would find the smaller categories of Òfast food,Ó ÒItalian,Ó Òdiner,Ó etc. Under these categories would be other, more specific categories as well. When information from texts is processed, all information must be taken into account. For correct schema activation to occur, the reader must use the incoming data to locate possible specifics in the schema. This activates the top down processing, which searches for the appropriate schema to account for all the details in the input.

Schema processing can be blocked in two ways and often are in L2 reading.Ê First, it is data-limited and depends on the readerÕs ability to receive data. therefore, factors such as a lack of vocabulary knowledge can derail it. Secondly, the processing system is resource-limited. It cannot use more resources than those available in the working memory. If both the vocabulary and the structure of the text, for example, are unfamiliar, the reader will not be able to allocate sufficient working memory resources to top-down and bottom-up processing. The reading will therefore short-circuit because the process demands of the reading exceed the resources of the learner.

The researchers used a short text, an AesopÕs fable entitled ÒStone Soup,Ó to illustrate how schema is invoked at different levels of comprehension by a reader. They argue that schema plays a role even in orthographic processing, where it is invoked to more quickly process words. Skilled readers do not need to look at each letter to process a word. Rather, they apply their schema of English orthography. A similar phenomenon occurs at the level of syntactic processing, in that the learner invokes background knowledge of English sentence structure to more quickly process the syntactic relationships within a sentence. At the semantic level, the reader fills in the details that are not in the text. For example, the fable opens with the statement that a poor man went to a large house. The reader must invoke her/his knowledge of Òpoor menÓ (that they donÕt have wealth) and Òlarge houseÓ (that they cost a lot of money) to understand that the house does not belong to the poor man. Semantic processing of this sort must happen throughout the text in order for comprehension to occur. Finally, schema of the nature of fables (that they normally have a moral) is applied at the interpretive level for the reader to understand the point of the story.

This is an excellent synthesis of the various schema theories that were proposed in the psychological and educational literature in the seventies and eighties. It clearly explains the notion of schemata and how they apply to the reading process at all levels. Unlike many subsequent top-down processing articles, this does not neglect the importance of bottom-up processing strategies, but rather indicates that they are essential for the activation of schema.

Schema theory not only applies to meaning-based concepts, but to all knowledge structures. Application of schema in reading requires background knowledge of concepts as well as background knowledge of the language and orthographic systems.

Researchers could investigate techniques for helping L2 adult learners to apply schema at in at levels of processingÑphonological, semantic, syntactic, interpretational, etc. The impact and role of culture in applying schema could also be researched.

Alderson, J.C. (1984). Reading in a foreign language: a reading problem or a language problem? In J.C. Alderson & A.H. Urquhart (Eds.), Reading in a foreign language (pp. 1-27). New York: Longman.

The previous research that attempted to characterize the nature of reading difficulties and abilities in second languages. While the fact that reading in a second language is a difficult task for most second language learners had been well established in the literature, it was unclear whether this difficulty stemmed from problems in learning the language or problems in reading. Alderson surveyed the studies to determine what factors contribute to reading difficulties.

Alderson surveyed empirical studies of second language reading from the seventies and eighties to determine whether difficulties in reading in a second language were prompted by imperfect understanding of the second language or by poor reading strategy use. The studies he included examined the second language reading of a variety of learner populations, ranging from adolescents to adults. Some studies were conducted in a second language context and others in a foreign language context. Alderson used the findings of these studies to form two major hypotheses and two modified hypotheses, then searched for evidence from previous studies to confirm or reject the hypotheses.

The first major hypothesis is that poor reading in the second language is caused by poor reading in the first language. This hypothesis is also examined in a modified form, that poor reading in a second language is due to the transfer of strategies from the first language that are not effective in the second language.

The second major hypothesis is that poor reading in the second language is caused by inadequate knowledge of the second language, both oral and syntactic. This hypothesis is also examined in a modified form, that poor reading in the second language occurs when readers have not crossed a threshold of second language ability, beyond which reading strategies from the first language can be applied to second language reading.

While some studies have shown evidence of the transfer of reading ability from one language to another, only very moderate or low correlations were found between first and second language reading when the same subjects are tested in both languages. This indicates that reading ability in the first language may not be the best indicator of reading ability in the second language.

For the modified hypothesis about the transfer of ineffective strategies from the first language to the second, the parallel processing hypothesis is examined. Proponents of this hypothesis believe that bilinguals have separate syntactic processors for dealing with each language. At early levels of language learning, learners might transfer the language-specific processing strategies of their first language to reading in their second language. In areas where the languages differ syntactically, they will have poor reading comprehension. After a certain level of language learning is attained, the readers gain syntactic processing strategies specific to the second language. While some studies seem to support this hypothesis, Alderson presents several studies that show that readers at even low levels of language proficiency are able to apply second language syntactic constraints to the interpretation of texts, refuting the parallel processing hypothesis.

The second hypothesis, that poor reading in a second language is due to inadequate knowledge of the second language, seems to be supported by most of the studies Alderson surveyed. However, the strongest support for this hypothesis comes from studies involving learners at lower levels of language proficiency. When readers with high levels of language proficiency are included, reading differences between native and non-native speakers are no longer easily attributable to language proficiency issues. This provides tentative evidence for a threshold hypothesis, but also indicates that language proficiency may only be the cause of poor second language reading among low proficiency readers. At higher levels of proficiency, it is possible that reading processes are more influential. However, as Alderson points out, we know virtually nothing about reading in a second language at higher levels of proficiency. More studies are needed to determine how the reading process changes as learners increase their language proficiency in the second language.

As Alderson points out, the study is limited because few of the studies surveyed are designed to answer the exact questions he has asked. While his interpretations of results of other studies appear to be reasonable, studies specifically designed to answer this question are needed before these questions can be adequately addressed. However, AldersonÕs survey is one of the first attempts to address the question of language learning and reading ability, which should be at the core of second language reading research.

If indeed second language reading is constrained more strongly by second language proficiency than by first language reading ability, literacy teachers should strive to increase their studentsÕ language proficiency, especially through the acquisition of vocabulary. This is most important at lower levels of reading ability. It is possible that, until students reach a threshold of language proficiency, reading strategy instruction will not be helpful. Linguistically modified texts that use restricted vocabulary, however, could be employed to help beginning learners read despite their inadequate language proficiency. These studies also indicate that the reading needs, and by extension instructional needs, of beginning and advanced students may be very different in terms of the balance of language instruction and strategy instruction. Teachers should be aware of the different needs of their students.

Reading skills transfer
Reading strategies
Second language proficiency
Vocabulary

Further research could address the needs of advanced level readers. Also, longitudinal research could be used to determine how the reading behaviors of learners change over time and how their instructional needs evolve.

Brown, C. (1993). Factors affecting the acquisition of vocabulary: Frequency and saliency of words. In T. Huckin, M. Haynes, & J. Coady (Eds.), Second language reading and vocabulary learning, (pp. 263-86). Norwood, NJ: Ablex Publishing Co.

The likelihood of different words being acquired by students from dual oral/visual input. The researcher wanted to know if the frequency of word use (either in general usage or in the specific input) affected studentsÕ acquisition of a word, and if the saliency of the word (either because it was central to comprehension or because the learners recognized that they did not previously know the word) would affect the likelihood of the word being acquired.

The learners were all students at an intensive English institute in the United States. Students from each of the centerÕs five levels participated in the research. Eighty-five students completed all the pre- and post-tests for the study. The learners included in the study came from mixed ethnolinguistic backgrounds. Approximately one-third of the participants were male; the other two-thirds were female.

An interactive videodisc program based on the movie Raiders of the Lost Ark and designed for ESL learners was used in the study. The videodisc allowed the students to watch scenes while reading the transcript. They could also check glosses of key words used in the scene, and could choose to do exercises that asked questions about the scene. All students had access to the videodisc in the computer lab.

To find out which words were most frequent in the specific context, a frequency test was run on the transcript for the video. The words from the video were divided into fifteen levels according to frequency. The most frequently used words fell into level one, then the second most in level two, and so on. A random sample of fifteen words was selected from each level for possible inclusion in the study.

To find out which words were most frequent in general, the Brown corpusÑa commonly used corpus of naturally occurring languageÑwas used. A frequency test was also run on the Brown corpus, which was also then divided into fifteen levels as in the case of the video script. Words were randomly selected from each level of the Brown corpus and matched with words from the corresponding level of frequency for words in the video. Random selection continued until there were ten words for each level. For each group of ten words, two occurred in Raiders but not in the Brown corpus, two occurred more frequently in Raiders than in Brown, two occurred equally in Raiders and Brown, two occurred more frequently in Brown, and two occurred in Brown but not in Raiders.

The words were divided in terms of saliency according to two variables: focus and gap. Focus was measured by whether or not the word was used in exercises or in glosses; gaps were considered to be created when the word was used in the exercises without appearing in the scriptÑwhen this happened, it was assumed that the student would recognize the presence of a gap because of seeing the concept without hearing the word.

Once the words were selected, a multiple-choice vocabulary test was created to test them. All the students were given the vocabulary pretest before they began working on the videodisc. Students were taught how to use the videodisc program and could choose to use itÑrecords were kept of who used the program and how much during the term. At the end of the term, the students completed the same vocabulary test as a post-test.

In general, frequency appeared to significantly affect the likelihood that a word would be acquiredÑgenerally frequent words were more likely to be acquired than words that were only frequent in the context of the program. However, an analysis of covariance indicated that the relationship frequency in general and acquisition was at the lower end of significance (F(1,1)=4.27, p=.040).

Saliency seemed to have a stronger impact on the acquisition of vocabulary. Almost all of the words that were most likely to be gained were ones that were central to the plot of the program. Many of these were low frequency words both in general and in the context, indicating that saliency in context may strongly influence the acquisition of words. Closer analysis of the data indicated that saliency caused by focus in the exercises or glosses did not affect acquisition (F(2,1)=.93, p=.395), but that saliency caused by creating gaps did (F(2,1)=5.25, p=.006). This indicates that saliency is more likely to contribute to the acquisition of vocabulary when learners recognize that they need a word that has not yet been acquired.

The report is very technical. The statistics are carefully done and well explained. While it is possible to say with some certainty that the high saliency words from the program itself were most likely acquired from exposure to the oral/written data, it is harder to be sure that the generally frequent words were acquired because students had varying degrees of exposure to this input. In the ESL context, the students most likely had many opportunities to learn generally frequent words (in everyday speaking situations and in the reading for their ESL courses). It is impossible to say whether the acquisition of these words came from exposure to the computer program or from some other source. Ultimately, the findings on frequency indicate that people learn the words they hear and read the most, which is not a novel finding.

Learners are most likely to learn words that are necessary for comprehension, or when they have a concept prior to learning the word. They are also more likely to learn words when they recognize that there is a gap in their knowledge. It is possible that teachers could increase vocabulary learning by creating a need for specific words before teaching them, perhaps through pre-reading exercises.

Researchers could compare the acquisition of words made salient through explicit instruction and words made salient through gapping in reading exercises to determine which condition is more conducive to language acquisition.

Carrell, P.L. (1989). Metacognitive awareness and second language reading. Modern Language Journal, 73 (2), 121-134.

The metacognitive awareness of second language readers and their reading comprehension in their first and second languages. Research on first language reading has indicated that metacognitive strategies, or the way in which learners organize their learning, can play an important role in reading comprehension. This study sought to determine what role metacognition played in second language reading.

There were two groups of students in the study. One group was comprised of 45 native speakers of Spanish studying ESL either at a pre-academic IEP (n=28) or in university composition classes for non-native speakers (n=17). The second group was composed of 75 native speakers of English studying Spanish at an American university. This group was composed of first year students (n=39), second year students (n=23), and third year students (n=13).

The students completed a metacognitive strategies questionnaire designed for this study. The questionnaire consisted of thirty-six statements about silent reading strategies. Using a 1-5 scale (1 = strongly agree, 5 = strongly disagree), students were required to indicate how closely each statement reflected their own reading behaviors. The questionnaire was designed to investigate studentsÕ confidence in reading in the language (confidence), their strategies when they do not comprehend what they are reading (repair), the reading strategies they considered effective (effective), and aspects of reading they found difficult (difficulty). Each student completed the metacognitive strategies questionnaire twice: once about their native language reading and once about their second language reading. All questionnaires were administered in the studentsÕ native language. An example of a statement is: ÒWhen reading silently in English, if I donÕt understand something, I keep on reading and hope for clarification later on.Ó

The students were also tested on reading comprehension in their first and second languages. The test for each language consisted of two reading passages (controlled for difficulty) followed by ten comprehension questions. All the passages were on the general topic of language, to control for schematic knowledge. The texts were also controlled for textual organization, so the tests were parallel.

For first language reading, there were no significant correlations between reading comprehension and confidence or repair. For the Spanish L1 participants, there was a negative correlation between some of the effective items and reading ability. Therefore, better L1 readers tended to claim not to use certain strategies. These tended to be local, bottom-up strategies. Students who claimed to not use bottom-up strategies and who also claimed that bottom-up processing caused them no problems tended to be the best L1 readers.

For second language reading, confidence and repair strategies seem to significantly influence reading comprehension. Specifically, students who felt able to distinguish main and supporting points and to critically question the author tended to be better readers. Students who reported giving up and stopping reading when unable to comprehend tended to have lower comprehension. Also, for L2 readers, focusing on sentence-level syntax seemed to be related to higher reading comprehension.

The students were further grouped as having ÒlocalÓ (using bottom-up, decoding) or ÒglobalÓ (using top-down, background knowledge of content) preferences based on their answers to both the difficulty and effective items on the questionnaires. For the English L1 group, it seemed that local processing was related to higher comprehension in Spanish reading; for the Spanish L1 group, it seemed that global processing was related to higher comprehension in ESL reading. The authors attribute this difference to the relatively higher proficiency of the ESL group compared to the Spanish as a second language (SL) group, and to the differences between ESL reading and foreign language (FL) reading. They indicate that the lower proficiency readers who also had little outside exposure to Spanish literacy might have to rely more on bottom-up, textual processing while the higher proficiency readers with extensive English literacy exposure might be better able to integrate their world knowledge and the texts.

The study is well designed and carried out:There were an adequate number of participants tested for both groups and their targeted readings were comparable (all about language). From the study, it seems that effective second language metacognitive processing strategies may be dependant both on student proficiency and on the context in which the language is learned. In other words, for metacognitive strategy instruction, it is important for teachers to understand that there may not be one set of universally effective strategies, but rather that strategy efficacy can be dependent on other factors.

The findings on the differences between SL and FL strategy use are only suggestive and should be interpreted with caution, because the SL and FL groups (as the authors point out) are not strictly comparable in level.

Teachers should be aware that metacognitive processing can influence reading comprehension. Students should be encouraged to monitor their metacognition in reading and focus on effective processing strategies. Teachers and students should be aware that effective strategies may vary from one context to another, and from one proficiency level to another.

To directly compare effective strategies in ESL and EFL context, this study could be replicated using an ESL group from the same L1 background (as this study did) as well as an EFL group from that same L1 background with similar educational backgrounds and English proficiency.

Carrell, P.L. (1991). Second language reading: Reading ability or language proficiency. Applied Linguistics, 12, 159-179.

The effects of both first language reading ability and second language proficiency on learnerÕs second language reading ability. The study was conducted in response to AldersonÕs (1984) article calling for research that tested reading ability in the first and second languages of the same participants to separate out the effects of first language reading proficiency from those of second language proficiency in second language reading.

There were two groups of participants in the study. The first group was composed of ESL students from an intensive English center in the United States and of students enrolled in composition courses for non-native speakers at a university in the United States. All of the forty-five ESL participants spoke Spanish as their native language. Based on the English Proficiency Chart of the Consortium of Intensive English Programs (CIEP), they fell into three proficiency groups: eight at level three, twenty at level four, and seventeen at level six (these seventeen were the matriculated university students).

The second group of participants in the study was comprised of native speakers of English studying Spanish at a university in the United States. These seventy-five students were rated by their teachers according to the same rubric as the ESL students. Based on these results, these students were also divided into three groups: thirty-nine at level two, twenty-three at level three, and thirteen at level four. Thus, the overall L2 proficiency level of the English L1 students was generally lower than that of the ESL students, and direct comparisons were only possible at levels three and four.

Two Spanish passages and two English passages (each of approximately 315-350 words) were selected for the study. The passages also had similar levels of reading difficulty: one for each language was rated at tenth grade, while the second English passage was rated at fifteenth grade, and the second Spanish passage was rated at twelfth grade. One passage from each language was written in a comparison/contrast style, the other for each language was written in a problem/solution format. In all the passages, the authorsÕ intent and personal opinions were clear. To limit the effects of topic, all the passages were on the topic of language, since this was assumed to be an interesting topic for second language learners. Ten multiple-choice questions that required the students to draw inferences from the readings were prepared for each passage.

All the students completed the instruments first in their second languages and then in their native languages. Their scores on the multiple-choice questions were used as their second language comprehension score.

Predictably, Spanish speakers out-performed English speakers on the Spanish test; English speakers out-performed Spanish speakers on the English test. The overall mean for the English test was higher than that of the Spanish test, most likely because of the overall lower proficiency level of the L1 English participants. Both first language reading ability and second language proficiency level were significant predictors of second language reading ability. In terms of the relative importance of each of these factors, first language reading ability was a more important predictor of second language reading ability for the ESL learners; however, for the L1 English participants, second language proficiency level was a more important predictor of second language reading ability. Carrell gives two possible explanations for this discrepancy. First of all, part of the difference could be attributable to the nature of second language learning (in the case of the ESL participants) and foreign language learning (in the case of the students of Spanish). Also, the L1 English learners of Spanish had overall lower second language proficiency than the ESL learners. It is possible that second language proficiency is a more important predictor of second language reading comprehension until a certain level of proficiency is attained, after which first language literacy becomes a more important predictor of comprehension.

The study filled an important gap in the literature by directly comparing first and second language reading for the same group of students. The comparisons drawn are very carefully explained.

It is unclear why the more advanced texts at the same level of reading difficulty were not chosen for the study. It may be that the difference between reading a twelfth grade text and a fifteenth grade text influenced studentsÕ reading comprehension scores.

Also, it is possible that the order in which the students did the tasks influenced the scores. Rather than having all students perform the tasks in the same order, the design should have been counter-balanced to control for design-order effects.

The most troubling limitation of the study was the mismatch between the groups. While it is possible to draw inferences about the second language reading of each individual group, it is impossible to compare the ESL and L1 English learners, since the proficiency levels of the groups were so disparate.

One of the most interesting conclusions of this study was that, at all the levels included in the study and for all subjects, both second language proficiency and first language reading ability were predicting factors of second language reading comprehension. This implies that, at all levels, reading teachers need to address both language learning and reading skills and they need to be aware ofÊ their studentsÕ L1 reading ability.

The research could be replicated with a counter-balanced design and with subjects from a broader range of proficiency levels. Further research could also include sensitive second language proficiency measures in order to more clearly relate language proficiency, first language reading, and second language reading comprehension.

Carrell, P.L., Pharis, B.G., & Liberto, J.C. (1989). Metacognitive strategy training for ESL reading. TESOL Quarterly, 23 (4), 647-678.

The effect of strategy training on second language reading. The study addressed the questions of whether strategy training enhanced L2 reading, whether a particular type of strategy training was more facilitative of L2 reading, and whether the effectiveness of strategy training was related to student learning style. The styles of strategy training examined were making pre- and post-reading semantic maps to help students focus on the details and relationships among them and the Experience/Text/Relationship (ETR) approach. In the experience step of the ETR approach, the teacher leads the students in a discussion of how their own knowledge or experience relates to the passage to be read; in the text step, the students read the text in chunks, with the teacher asking content questions about each section; in the relationship step, the teacher directs the students to make connections between what they already know about the subject (what they discussed in the experience step) with what they have learned about the topic from reading the text and answering questions about its content (text step). The students are directed to draw relationships between the text and their expectations.

Twenty-six ESL learners of mixed ethnolinguistic backgrounds (Greek, Arabic, Japanese, Malaysian, various African groups, Chinese, and Spanish) in an intensive English program at an American university participated in this study. Their ages ranged from 19-43. Nineteen were male, seven were female. Seventeen students were preparing to begin undergraduate studies; the others were preparing to begin graduate studies. Their Test of English as a Foreign Language (TOEFL) scores ranged from 470-524. (Most universities require a score of at least 600 for admission to credit bearing classes.)Ê

The students came from four reading classes from the same level. At the beginning of the study, all the students completed a reading comprehension test based on three TOEFL reading passages. For each passage, the students answered three multiple-choice questions and two short-answer questions. They were also required to complete a partially constructed semantic map (a diagram of vocabulary related to topic to be read) after each of the first two passages. After the third passage, they were asked to construct a semantic map without any scaffolding.

The students also discussed learning styles in their classes, and after the discussion completed a learning styles inventory, the Inventory of Learning Processes (ILP). The inventory consists of 62 true-false questions, divided into four subsections. The subsections are designed to measure studentsÕ tendencies to critically evaluate information, personalize information, memorize information, and engage in a drill-type study behavior. it is thought that people who exhibit higher tendencies to abstract and evaluate information on the inventory should be better able to comprehend texts well.

During the treatment phase of the study, one class (nine students) participated in the semantic mapping activities over four days of course, using three different texts. The second class (nine students) participated in the ETR activity using the same texts. The last two classes (three and five students) held regular classes based on the same texts (no discussion is included of their activities these days).

Following the instruction, the students retook the same pre-reading test. On the pre- and post-tests, the multiple choice questions were scored as correct or not. Semantic maps were scored based on the percentage correct of possible responses. The short answer questions were graded on a scale of 1-3 based on how well the studentÕs answers demonstrated understanding.

The control group did not significantly increase their scores on any part of the comprehension tests. The group that participated in semantic mapping improved their comprehension as measured by open-ended questions and by creating a semantic map without scaffolding. The group that participated in ETR significantly improved their comprehension as measured by open-ended questions and on the partially completed semantic maps.

A significant interaction was found between improvement on the post-tests and learning style type. The relevant learning styles dealt with the critical evaluation of information and the tendency to personalize information. In the control group, only students with tendencies to critically evaluation information scored better on the semantic mapping questions.

The results should be interpreted with caution for several reasons. There were very small numbers in each of the groups, so these results should be generalized with caution. The number of students in each group is technically too low for the assumptions of the statistical measures used, undermining the validity of the findings.

Secondly, one of the strongest findings, the increased performance of the semantic mapping group on making semantic maps, is most likely due to the practice effect rather than strategy training.

Lastly, the control groupÕs activities were not explained. It is by no means sure that their instruction during the treatment days contained similar amounts of reading input and practice to that experienced by the treatment groups. Their improved comprehension scores could therefore be the result of increased input, rather than strategy training.

The study does seem to indicate that strategy training can lead to improved second language reading comprehension. Teachers and curriculum designers need to consider the role of reading strategies and strategy instruction in the development of ESL reading comprehension. They also need to look at providing a variety of strategies to meet the learning styles of the students.

Other types of strategy instruction should also be systematically evaluated in classroom settings to determine what types of instruction might help students better improve their ability to read and comprehend English texts. Researchers could also look at the impact of further developing strategies that students exhibit a preference for, and developing strategies that students do not indicate a preference for.

Carroll, S.D. (1999). Storytelling for literacy. Grand Rapids, MI: Annual Meeting of the Michigan Reading Association. (ERIC Document Reproduction Service No. ED 430234)

The effect of oral story telling on ESL adult literacy and language development. Some researchers believe that a lack of exposure to oral genres like storytelling can contribute to reading difficulties for both first and second language learners. The author sought to establish a link between learning or practicing an oral genre like storytelling in the second language and the development of literacy skills in that language.

The learners were Chinese adults enrolled in a three-month community-based ESL course. The class focused on developing oral and written storytelling skills.

The researcher was the teacher of the course, and was committed to the concept of developing understanding of the storytelling genre.

The course lasted three months. During this time, students participated in storytelling several times. In each case, the stories came from their own experiences in China. During the study, the researcher recorded qualitative impressions of the studentsÕ progress in processing oral and written input.

The teacher reported that the students increased their confidence in speaking in public and developed more native-like oral language. They also increased their reading comprehension when reading narratives.

It is difficult to support the claims that the studentsÕ language or reading skills were more target-like by the end of the study because only impressionistic data was collected and reported. It is not clear how the teacher was measuring the increase in learner confidence.Ê Moreover, the lack of control group or any comparison data prevents us from drawing conclusions about the effectiveness of this technique as compared with others.

This study was inconclusive. However, increasing studentsÕ exposure to oral genres may help them better comprehend related written genres. This may be especially worth considering in the case of pre-literate learners, as oral practice is more familiar and accessible to them ad may serve to support their reading.

The thesis of this study could be investigated with more structure and design validity, including pre- and post-program measures of language use and reading comprehension.

Carson, J.E., Carrell, P.L., Silberstein, S., & Kuehn, P. (1990). Reading-writing relationships in first and second language. TESOL Quarterly, 24, 245-266.

(a) The relationship between reading in the L1 and in the L2, (b) the relationship between writing in the L1 and in the L2, (c) the relationship between reading and writing in the L1, and (d) the relationship between reading and writing in the L2.

48 native speakers of Chinese and 57 native speakers of Japanese enrolled in pre-academic intensive English courses or in university basic writing courses in American universities. Students were at the low intermediate to advanced proficiency level (TOEFL scores ranged from 400-525) and were literate in their first languages.

The students each completed four tasks in a two-week period: writing a short essay in the L1, writing a short essay in the L2, completing a cloze task in the L1, and completing a cloze task in the L2. Essays were scored based on the 6-point rubric of the Test of Written English (part of the TOEFL exam) øwhere 6 points Òclearly demonstrated competence in writing and 1 point was ÒincompetentÓ in writing; cloze tests were scored based on exact-word scoring.

There were significant correlations between L1 reading and L1 writing and between L2 reading and L2 writing for both language groups at all proficiency levels. There was also a significant correlation between L1 reading and L2 reading across language groups and proficiency levels. However, there were only weak or non-significant correlations between L1 writing and L2 writing. This could be attributed to insufficient variance.This indicates that while first language reading ability may affect second language reading proficiency, first language writing ability may not affect second language writing ability. This suggests that students may be able to transfer their reading skills from their native language to English, but may be less able to transfer writing skills. It also suggests that studentsÕ reading ability in both their first language and in English can be an important influence on their writing ability.

Several variables, including educational background in the first language and the length of time students had studied in the United States, may have also affected studentsÕ first and second language literacy skills.

Students from both L1 groups scored lower on average on the L1 cloze than they did on the English cloze. This may indicate that the L1 cloze tests were not reliable for measuring L1 reading ability. This is a possibility because of the nature of Japanese and Chinese writingÑthere are no clear defining orthographic characteristics of words in these writing systems (such as the spaces preceding and following words in English writing). As such, constructing Chinese or Japanese cloze exercises is awkward. Also, the participants were likely unfamiliar with cloze exercises in their L1.

There were not enough participants at different proficiency levels to directly compare their performances. This study does not show when the transfer of reading skills from first to second language is likely to take place.

The study indicates that students can transfer literacy skills from their first to their second languages, and possibly should be encouraged to do so.

The authors assume that proficiency levels are an indication of the transfer of reading skills. However, L2 reading proficiency might be built on a separate set of abilities. A more direct measure of reading comprehension, such as recall, should be used.

Chervenick, E.C. (1992). Schema theory: Teaching U.S. history to beginning amnesty students. Unpublished masterÕs thesis, Biola University, La Miranda, CA. ERIC Document Reproduction Service No. ED 354 787.

Effects of teaching to build new knowledge schemata. Schema theory has conclusively shown that students demonstrate higher levels of reading comprehension when reading on a topic for which they have prior or cultural knowledge (Jagar-Adams & Collins). This study investigated the effect on reading comprehension of teaching students about a new topic to help them build schemata.

The learners were thirty-eight beginning level amnesty students. The amnesty program was designed to give illegal aliens in the United States an opportunity to establish residency in the United States during a transitional period in immigration legislation. The adult students were required to attend 40 hours of a 60-hour ESL course or take a test on United States history similar to the United States citizenship tests. The goal of the amnesty classes was to provide adult ESL students with the coursework necessary to receive a green card, and to help prepare students to eventually go on to complete the citizenship process if they chose to do so.

The teachers in this study were the researcher and one other amnesty teacher. Both were masterÕs students in teaching English to speakers of other languages (TESOL).

All the students were given one of two reading passages (one complex and one simple) about Abraham Lincoln and asked to complete an objective test, a written free recall, and an oral guided recall based on the reading. Five days later, the students divided themselves into two groups to receive special history classes. The experimental group spent two class periods discussing Abraham Lincoln, while the control group spent the time learning about Martin Luther King, Jr. using similar materials. All the students then again read one of the two passages about Abraham Lincoln and repeated the pre-test measures.

The written recalls were scored by counting the words in each studentÕs recall that were reproduced from the original text. Points were deducted for incorrectly spelled words. The experimental groupÕs improvement on the written recall between the pre- and post-tests was significantly better than that of the control group, indicating that they were able to more exactly recall the text after having learned about Abraham Lincoln. On the objective test and the oral recall, which was scored based on the ideas replicated from the reading, there were no significant differences between the control and experimental groups. The author interprets the performance of the experimental group on the written recall as evidence for the effectiveness of helping students to build schematic knowledge for improving reading comprehension.

The researcher allowed the students to self-select group membership in the experimental or control groups, rather than assigning half of the students who had used the simple text to each group and half the students who had used the complex text to each group. It was therefore impossible to separate out the effect of the text complexity from the effect of receiving instruction. It cannot be determined from the analysis whether the difference in the scores is caused by the instruction or by having read a particular text.

The written recalls were scored according to the words reproduced from the text, as opposed to according to the number of ideas remembered from the text. Using this sort of scoring, a student who didnÕt comprehend the text well but memorized some words would score higher than a student who understood the text well enough to explain it in his or her own words. It is not clear that this is an appropriate measure of comprehension. This is further complicated by the deduction of points for poor spelling. While spelling ability may be a reflection of reading comprehension, it may also be related to the studentsÕ writing ability.

The control groupsÕ topic for the days of instruction, Martin Luther King, Jr. and the civil rights movement, may not have been sufficiently different from the experimental groupsÕ topic, Abraham Lincoln and the Civil War. For the students to understand the civil rights movement, they would have to receive some information on slavery and the Civil War, as well as subsequent political movements involving race. The schema-based instruction for the control group was therefore not entirely controlled.

This study encourages teachers to teach students the background knowledge associated with the texts they are reading. It suggests that this may help students increase their reading comprehension.

This study should be replicated using a controlled design to separate the effect of text complexity from the effect of instruction. A replication should also include more adequate measures of comprehension and should involve a control group that works with a completely unrelated topic.

Longitudinal research could investigate the long-term effects of instruction to help students build schema. Delayed post-tests, for example, could be used to determine how long the effect of the instruction lasts.

Cho, K.S. & Krashen, S.D. (1994). Acquisition of vocabulary from the Sweet Valley Kids series: Adult ESL acquisition. Journal of Reading, 37, 662-667.

The effects of pleasure reading on the reading habits and vocabulary of adult ESL learners.

Four adult female learners participated in this study: three Koreans and one Spanish speaker. All had studied English for several years; all but one had lived in the United States for several years. All were literate in their native languages and had varying levels of English literacy. None reported engaging in pleasure reading in English before this study.

The researchers hypothesized that if they gave adult ESL students comprehensible and interesting reading material in English the students would read more and increase their language proficiency. They chose books from the Sweet Valley series. This series of popular childrenÕsÕ books consists of three separate seriesÑSweet Valley High (sixth-grade reading level), Sweet Valley Twins (fourth-grade reading level) and Sweet Valley Kids (second-grade reading level). All participants began by reading books from the Sweet Valley Kids series. They were told simply to read as much as they liked. They were not asked to study new vocabulary, and they were not told that they would be tested for comprehension.

To aid their comprehension, the researchers gave the participants background knowledge about the series and occasionally discussed the plots of the stories and relevant aspects of American culture. The women also were asked to underline words they did not know the first time they encountered them in the readings. They were not given any instructions about the use of dictionariesÑsome chose to look up words; others did not.

Additionally, one of the researchers discussed the plots and characters of the books in Korean with the Korean-speaking participants. During the conversations, the plot and characters of the books were discussed. These conversations most likely increased the Korean readersÕ comprehension of the texts. This likely affected the findings

All of the participants were very enthusiastic about the readings. They read between two and six books a week in their free time and reported that they were eager to read more. They also reported that they had noticed improvements in their oral communication while participating in this research. The participants were tested on the words that they had underlined as new words in the books. Their scores indicated that they had learned between 60 and 80% of the new words they had found.

The study did not employ rigorous pre- and post-testing procedures, without which conclusions about vocabulary acquisition are questionable. Also, the researcherÕs hypothesis was that learners would gain vocabulary knowledge through comprehensible input, not through a focus on vocabulary learning. Having the participants underline new words while they read changed the nature of reading. We cannot say whether the growth in participant vocabulary is the result of the reading or of focusing participantsÕ attention on the vocabulary as they read. Furthermore, the reading done was discussed with a teacher, so it was a modified version of individual extensive reading.

The study suggests, however, that pleasure reading may give students more access to comprehensible language. It should be remembered, however, that some pleasure reading depends on cultural and genre knowledge that English language learners may not have yet acquired.

Students are more likely to engage in reading if the reading materials are interesting to them. Narrow reading (reading works from the same genre or by the same author) may be more likely to facilitate pleasure reading, as it continues student engagement in the topic. The study also suggests that engaging students in pleasure reading can facilitate incidental vocabulary learning, especially if the words are repeated often.

This research could be replicated among other populations, using other texts relevant to them. Also, while the books may be a good resource for conversational language -- dialogue can be hard to read in L2 if you arenÕt familiar with the spoken language it tries to simulate --Ê they most likely do not facilitate the learning of academic or professional language. Narrow reading should be examined using books that are more directed to professional learning.

Coady, J., Mgoto, J., Hubbard, P., Graney, J. & Mokhtari, K. (1993). High frequency vocabulary and reading proficiency in ESL readers. In T. Huckin, M. Haynes, & J. Coady (Eds.) Second language reading and vocabulary learning (pp. 217-228). Norwood, NJ: Ablex Publishing Corporation.

Explicit instruction in high-frequency vocabulary and its effect on ESL reading comprehension. The researchers hypothesized that explicitly focusing student attention on words from the 2,000 most common words in English may help them to automatize lexical recognition of those words while reading in English. Automaticity in processing certain text elements increases the cognitive resources available for processing text elements that have not become automatic yet. The hypothesized effect of this was increased reading comprehension.

All the learners involved in this study were students enrolled in an intensive English program. No information was given on their L1 educational backgrounds. For phase one of the study, there were twenty-two subjects in the experimental group and twenty in the control group. No information was given on subjectsÕ English proficiency level, although descriptive statistics from the study indicate that there was a range of proficiency levels. Also, the group selection mechanisms are not explained, although is seems most likely that existing classes were used for the study. In phase two of the experiment, forty-two students were involved in the study. For this phase of the study, there was no control group.

In phase one of the experiment, all the students completed two pre-tests: a thirty-six item vocabulary multiple choice synonym test derived from a basic vocabulary list and the Degrees of Reading test from the College Entrance Examination Board (CEEB) (this multiple choice reading comprehension test was created for native speakers). The control group received no specialized reading or vocabulary training other than the regular English instruction they received along with the experimental group.

The experimental group spent one hour per week for eight weeks working with a computerized vocabulary-learning program. The program presented them with words from the 600-2,000 frequency range of the most frequent words in English. Each time a student began a lesson, he would see a list of twenty words from this lesson. The student could select any unknown words to see definitions and sample sentences, and to add his own mnemonic cue to remember the word. The computer would generate multiple choice practice tests and unit tests. Selected words and words missed on tests would be added to a personalized review file, which could be used to create individualized units.

At the end of the eight-week period, all students completed the same vocabulary and reading measures that were used for the pre-test.

In phase two of the experiment, all students had access to the vocabulary-learning program. The vocabulary and reading comprehension measures were altered for this phase: the vocabulary was selected randomly from the computer program, and a different version of the reading test was used, one that included readings targeted at lower levels of reading comprehension. The phase two students also completed a short, closed-ended questionnaire asking them to evaluate the computer program.

The results from both phases of the experiment indicate that studying high-frequency vocabulary does increase reading comprehension. In phase one of the study, students from the experimental group experienced significantly greater gains in both reading comprehension and vocabulary knowledge than the control group (F(41,1)=9.55, p<.005). In the second phase of the study, the students significantly increased their reading comprehension scores and their vocabulary scores, according to matched t-tests (p<.05 for each reading comprehension and vocabulary). On the questionnaire, the majority indicated that they enjoyed using the program and felt that it had helped them to learn vocabulary and improve their reading comprehension.

The study seems to indicate a link between studying high frequency vocabulary and reading comprehension. It would have been strengthened by separately considering the performance of students at different proficiency levels rather than grouping them into unified control and experimental groups. It is possible that learning high frequency vocabulary is more effective for learners at certain levels of reading proficiency than for others.

Also, closer examination of the statistics reveals that the control and experimental groups were not equivalent at the beginning of the study. The experimental group started the study with a much lower mean score on the reading comprehension test than did the control group. Both groups improved their reading comprehension during the study, but the control groupÕs gains were more pronounced than the experimental groups. On most language proficiency measures, more dramatic gains are made by lower-proficiency learners; thus, it is not possible to determine whether the experimental group had more dramatic reading proficiency gains because of exposure to the vocabulary-learning computer program or because of the disparity of learnersÕ proficiency levels before the treatment.

Also, it is important to remember that this study does not investigate the effectiveness of vocabulary learning on reading comprehension in general, but rather the effect of one specific, individualized vocabulary-learning system. These findings may not be generalizable to other methods of vocabulary teaching.

The findings of the study give preliminary indications that reading comprehension can be facilitated when students are encouraged to increase their familiarity with high frequency vocabulary. This offers an example of how a bottom-up text processing practice such as vocabulary building can support reading comprehension.

This study could be replicated controlling for student proficiency levels. It also could be replicated with a larger number of students at different levels of proficiency to determine what effects, if any, proficiency level has on the impact of high-frequency vocabulary learning on reading comprehension. Different methods of instruction of high-frequency vocabulary learning could be compared to establish which methods do positively influence reading comprehension.

Coady, J. (1997). L2 vocabulary acquisition through extensive reading. In J. Coady & T. Huckin (Eds.) Second language vocabulary acquisition (pp. 225-237). Cambridge: Cambridge University Press.

The possibility of learning vocabulary through reading in a second language. While proponents of the input hypothesis claim that the best way to learn second language vocabulary is through extensive reading, some research has indicated that this may not be true. This researcher sought to explain the research on extensive reading and how extensive reading might be best applied.

Proponents of incidental vocabulary learning claim that vocabulary is learned gradually with repeated exposure. They estimate that second language learners need about ten to twelve exposures to a word in a comprehensible context over time in order to learn it well. While some L2 vocabulary acquisition research indicates that this can happen through reading, most of these studies did not provide adequate controls and were conducted with intermediate to advanced level learners. Other studies have failed to show any benefits from extensive reading to vocabulary acquisition. These explain the foundations of the beginnerÕs paradox, which is CoadyÕs focus here.

The beginnerÕs paradox comes from two competing ideas in the L2 literacy acquisition literature. The input hypothesis states that a second language learner can learn vocabulary through reading. However, bottom-up processing research has indicated that there is a vocabulary threshold for reading. Learners cannot read in a second language at a level of comprehension sufficient to learn new words from context until they have gained a certain threshold of vocabulary in the second language. This threshold has been estimated to be in the range of 3,000 to 5,000 word families, or 5,000 to 8,000 lexical items. Before attaining this level, it will be impossible for readers to comprehend with accuracy. This is the paradox: beginning language learners need to read to gain vocabulary, but they need to gain vocabulary to read.

Coady suggests that special attention be given to vocabulary learning until learners have reached the vocabulary threshold. He also suggests that top-down processing will be more effectiveÑand better able to overcome the effects of limited vocabularyÑif the student reads texts that are personally interesting and familiar. He suggests that the bottom-up processing instruction in vocabulary and structures be complemented with an emphasis on reader interest and a match between the text and the readerÕs background knowledge.

The article clearly explains the nature of the beginnerÕs paradox and the need to support beginning students through both vocabulary and reading instruction. It articulates rationale for attending to both bottom-up and top-down processing in literacy instruction. It also emphasizes the fact that different learner background knowledge may preclude the possibility of effective homogenized instruction at the beginning level. Beginning level readers especially need to read texts that match their background knowledge, as they will already have the structure needed to understand the reading, thus freeing resources for lower-level vocabulary processing. When learners have mixed backgrounds and interests, instruction will have to be individualized to allow for comprehension and vocabulary building.

The study suggests that teachers should incorporate vocabulary instruction in high-frequency English vocabulary for their learners. Also, teachers should match texts to student experiences, to facilitate the more exact transfer of schema when students may not have sufficient bottom-up processing capacity to activate appropriate schema on their own.

Different methods of teaching vocabulary, with emphasis on the more common words of English and on textual support of vocabulary instruction should be evaluated.

Constantino, R. (1995). Reading in a second language doesnÕt have to hurt: The effect of pleasure reading. Journal of Adolescent and Adult Reading, 39, 68-69.

The effect on reading comprehension and writing fluency of introducing adult ESL learners to pleasure reading in an intensive language course.

The learners were adult low intermediate speakers from Asian, Latin American, and European countries. They had various educational backgrounds, ranging from sixth grade to high school education. All could read in their native languages; however, all felt apprehensive about reading in English. At the beginning of the class, the students indicated that they considered reading in English to be the end result of English study, not a means of learning English. They expected a reading class to include book reports and tests for vocabulary and spelling.

The practitioner researcher was the teacher of the course. She designed the course to have the students engage in pleasure reading. There were no tests or quizzes and no assigned writing.

The first twenty minutes of each class were devoted to pleasure reading. They were not allowed to use their dictionaries. During this time, the students could read anything they brought with them or any of the books, magazines, and articles the teacher supplied. In addition, the students were assigned reading outside of class on topics and books they had previously selected as a class. The students could elect to keep reading journals in which they summarized and reacted to the readings. They were required to write questions and answers about the books they were reading. The teacher used these as a means of ascertaining their comprehension of the books. Class time was spent in discussing the assigned readings or in reading short articles the teacher had brought into class to supplement the books they were reading at home.

The teacher noticed changes in the studentsÕ reading behavior during the course. When reading in class, they began to rely less and less on dictionaries, and made fewer queries about word meanings. They began to read in English for meaning, similar to what they did in their native languages. As the course progressed, more students elected to write reading journals about the issues the books brought up that related to their own lives, and they expressed more confidence in their writing abilities. Students became more confident in their speaking abilities and were more willing to participate in class. They also reported that they felt more comfortable speaking with friends and neighbors outside of class. They also indicated that they felt more comfortable reading in English and that they were reading books not required by the class.

The study suggests that behavioral and attitudinal changes may be associated with pleasure reading. Because the language development evidence is anecdotal, it is unclear whether there are language development benefits from reading extensively in English. Given that fact that the students also discussed the books and wrote about them, this may have also contributed to the the behavior and attitudinal changesÊ observed by the teacher.

Students may experience and recognize reading proficiency development by engaging in pleasure reading. The benefits of pleasure reading may also extend to language development in oral communication as well.

Pleasure reading
Extensive reading
Written-oral/reading relationship
Vocabulary
Reading skills transfer

As with much practitioner research, this research could be replicated among a larger group of participants incorporating pre- and post-testing and control groups.

Cotterall, S. (1991). Meiko: Case study of a second language reader. Reading in a Foreign Language, 8(1), 611- 627.

The strategies used by a second language reader in a program designed to promote reading strategy development.

The learner in this case study was Meiko, a Japanese learner of English at an intensive English program in New Zealand. She was a 21-year-old female who had no previous experience of study in an English-speaking country. While she had a good knowledge of English grammar, her oral English was not fluent. The videotapes of the class sessions showed she was very interested in the class discussions, although she contributed little verbally to these discussions.

No information is given on the teacher. It seems likely, but is not stated, that the researcher was also the teacher of this course.

The case study consists primarily of information gathered from MeikoÕs contributions to the reading strategies class. Additionally, she was interviewed in English and was asked to complete a reading strategy inventory.

The reading strategies course employed the procedure of reciprocal teaching. This technique focuses readerÕs attention on strategies that can be used in order to monitor and promote comprehension. When using this technique, the class examines a text together. For each paragraph of the text, the students and teacher together clarify areas of miscomprehension, identify the main idea, summarize the paragraph, and predict what will happen next. The technique is built on the features of metacognitive awareness (helping learners to understand their own strategy use), modeling of strategy use (helping students to understand how to actively apply strategies), scaffolded instruction (helping learners to collectively deal with texts that they would not be able to process alone), and overt monitoring of comprehension (helping learners to recognize communicative breakdown and identify the source of miscommunication).

The researcher found that Meiko tended to request clarification for discrete words using a formulaic request, and that she sought to find a single word synonym or Japanese gloss, paying little attention to context. When she attempted to clarify for other students, she also tended to supply single words. When seeking the main idea of the text, she tended to focus on macro discourse markers like ÒAn important point . . .Ó or Òthe problem lies in the fact . . .Ó Her main idea statements tended to be closely modeled on the wording of the text. This text reliance was also demonstrated in her summarizing behavior. Here, she generally tended to restate the main idea previously identified and add a few details. When predicting, she normally indicated that the following paragraph would give more details about one of the ideas from the summary. In all of her responses, she tended to rely heavily on the text rather than background knowledge or experience.

Her answers to the reading strategy inventory (created by the researcher) indicated that she did not tend to integrate reading with her background knowledge, and that she focused mostly on word level phenomena in reading. In interviews with the researcher, Meiko indicated that she used summarizing as a way of dealing with miscomprehension, and that she felt that the instruction had helped her to find main ideas more easily. The researcher concluded from this that Meiko did not use the clarifying stage well, since that is the area intended for clearing up miscomprehension (rather than summarizing) and that Meiko failed to Òappreciate the potential for applying the strategies in which she had received training to other contexts, or of the purpose of reading being to make sense of the ideas represented in textÓ (p. 625).

The study fails to separate the learnerÕs strategy profile from the effects of instruction. It is impossible to tell whether MeikoÕs behaviors and perceptions are the result of her initial strategy use or of the strategy instruction. Some pre-instruction characterization would have been appropriate. There is also no evidence presented of the effectiveness of the instructional technique used, orÊ that the strategies encouraged by the technique are actually effective strategies for L2 readers. Some (i.e., summarizing) seem to have little connection to reading comprehension, and the analysis of MeikoÕs strategies in this section seem to relate more to her language proficiency and production than to her reading comprehension or use of reading strategies. Finally, the analysis of MeikoÕs interview data is disappointingly brief and lacks detail. This is odd, considering that the participant was interviewed on four occasions. The conclusions (i.e., that the learner did not appreciate the potential for applying strategies) drawn from the interview seem to be based more on what the learner did not say than what she actually said. The overarching conclusion that Meiko, as a relatively low-proficiency reader, tended to use more text-based processing than top-down processing, is neither new nor particularly interesting. The researcher concludes that strategy development for Meiko would necessitate a change in attitude; however, it appearsÊ that no attempt was made to determine her current attitude towards reading strategy use.

Lower proficiency learners may tend to rely more on bottom-up processing than on top-down knowledge application. They should be encouraged to use top-down processing in their reading strategies.

An in-depth, richly reported case study of a learnerÕs use of reading strategies, attitudes toward reading strategies, or of a learnerÕs response to strategy training could be highly useful. Such a study should employ careful analysis of the qualitative data and draw only merited conclusions. The participants also should be assessed and interviewed both pre- and post-intervention to see what effect the intervention has had.

Devine, J. (1988). A case study of two learners: Models of reading and reading performance. In P.L. Carrell, J. Devine, & D.E. Eskey (Eds.) Interactive approaches to second language reading (pp. 127- 151). Cambridge: Cambridge University Press.

The reading behavior of two adult ESL learners who have different internalized models of the reading process. Previous work by this author has indicated that almost all learners have an internal model of what it means to read and that these models influence the way people read and the reading difficulties they are likely to experience. The focus of this study was examining this process in depth with two learners who expressed different orientations towards reading.

The learners were recent immigrants to the United States who were studying at a community-based ESL program in the Midwest. These learners attended classes for three hours daily, five days a week, and were at the high beginning level of English proficiency when the study began. One participant, Stanislav, was a Polish male who held advanced degrees in chemistry in his native language. The other participant, Isabella, was a Spanish female who had no education beyond high school.

In the previous study, twenty learners had been asked to describe the reading process. The learnersÕ responses fell into three different categories: sound-centered models, word-centered models, and meaning-centered models. The author argues that readers who have internal sound-centered models of reading focus on correctly matching sound-symbol correspondences. Learners who envision reading as a word-centered process focus on recognizing and understanding individual words. Having a meaning-centered modal of reading indicates that the learner focuses on understanding the meaning of the text as a whole. Miscue analysis of learnersÕ oral reading errors indicated that their internal models of reading significantly affected their oral reading proficiency. Sound-centered readers, for example, closely followed the print on the page, but did not recognize when they had pronounced non-sense words. Internalized models of reading also affected recall of the text. Meaning-centered learners demonstrated better recall of the texts than did sound- and word-centered learners.

For this study, two subjects from the previous study were chosen. Stanislav has a sound-centered model of reading, while Isabella has a meaning-centered model of reading. For the study, each read a passage out loud and then orally recalled what they had read. The researcher analyzed the number and types of miscues that each used and rated the recall for each learner.

The author hypothesized that learnerÕs internal models of reading might influence the reading process by determining which L1 reading skills can be transferred to the L2 or by allowing or restricting schema activation. Learners with a meaning-centered internal model of the reading process might be more able to transfer top-down skills from their L1 reading, making limited language proficiency less of a barrier to reading comprehension. They also might be more likely to activate appropriate schema, as they are more interested in constructing meaning and are more willing to move beyond the printed text. In all, this indicates that learners with meaning-centered models of reading might be advantaged in reading comprehension, explaining why learners with similar language proficiency can have very different experiences with reading.

Stanislav and IsabellaÕs divergent internal models of the reading process seem to correspond to elements of their reading performance. StanislavÕs oral reading contained very few miscues (his miscue per word rate was just over 3%), most of which were caused by spelling irregularities in English. However, a high percentage of his miscues were semantically impossible in the given context, and almost one fourth of the miscues were non-words, indicating that he was not monitoring for meaning while he read. Predictably, his recall was very poor, as he was not able to summarize the basic plot of the story.

On the other hand, IsabellaÕs oral reading contained almost four times more miscues than StanislavÕs (her miscue per word rate was around 13%). However, her miscues were more likely to be semantically possible in the given context, and she used almost no non-words, indicating that she did monitor for meaning more than form while reading. Her retelling of the story was very complete, including the plot line, the characters, and most of the major details, as well as inferences she drew from the story itself.

The study suggests that language proficiency is not the only factor in determining reading comprehension. However, without knowing the levels of component reading skills of each reader in English, we cannot be sure what the differences in their miscue patterns were due to.Ê

Raising learner awareness of the different aspects of reading and of their own internal models of reading and how they may impact their reading ability could help sound-centered readers to focus more on meaning and comprehension.

The effects of different instructional methods on learners with different preferences of reading models could be investigated.

Eskey, D.E. (1997). Models of reading and the ESOL student: Implications and limitations. Focus on Basics, 1(B), 9-11.

The possibility of applying psycholinguistic and processing models of reading (which have traditionally been based on native language reading) to second language reading.

Current models of the reading process focus on the interaction of top-down language and background knowledge with bottom-up text processing. These models indicate readers need to decode and interpret simultaneously. For ESL students, this indicates that if they have deficiencies in either component -- decoding skills or with language or background knowledge -- they will have difficulties reading in English.

Eskey points out that second language learners most often have deficiencies in both of these areasÑthat they are likely to lack both the schema and the appropriate decoding skills. He suggests that ESL teachers need to consider both of these issues in reading instruction. One of his suggestions was that teachers needed to take these issues into consideration in the selection of reading materials, judging both the language and the content of the reading.

Eskey also reminds teachers that reading is not only a psycholinguistic, but also a social process, as readers are inducted into the community of readers in a language. The uses for literature can vary from culture to culture, and studentsÕ expectations for the use of literacy can differ from the reality of literacy use in English.

Eskey explains simply the applications for reading models in the ESL classroom. He also points out that psycholinguistic processing models will never cover the totality of literacy. While his observations are interesting, data on the success of the various techniques he advocates would be helpful.

Eskey advocates that teachers use varied pedagogical techniques to help students increase their use of both top-down and bottom-up processing strategies. Top-down strategies can include pre-reading activities to build knowledge schemas; bottom-up strategies include helping students to chunk text to process it more rapidly.

The various strategies that Eskey recommends could be investigated with ESL learners to determine their effectiveness. Also, texts for ESL learners could be evaluated and rated for difficulty level with respect to both studentsÕ decoding skills and their language and background knowledge, to help teachers select appropriate reading material.

Fitzgerald, N.B. & Young, M.B. (1997). The influence of persistence on literacy learning in adult education. Adult Education Quarterly, 47(2), 78-91.

The effect of persistence, or the number of hours of instruction on literacy outcomes in three different adult education settings: ESL, ABE, and adult secondary education. As each type of setting was studied separately, only the design issues and findings relevant to ESL will be reported here.

The learners in this study were participants in 44 adult education centers in the United States. As the study made use of existing test scores for participants in adult education literacy classes, subjects were only included in the study if the centers could provide valid, matched pre-test/post-test scores for that individual. In all, data were collected for 614 students. Of these, 349 were ESL learners. Approximately half of the ESL participants were male and half female. Seventy percent were native speakers of Spanish; the remaining thirty percent represent a wide variety of ethnolinguistic backgrounds.

The authors collected pre- and post-test scores on the Comprehensive Adult Student Assessment System (CASAS) literacy test for the subjects. They alsoÊ collected information on the background of the students and the structure of the adult education program from the from the adult education centers included in the study. Multiple regression analysis was performed on the data. The study looked at possible effects based on the studentsÕ background (ethnicity, years of previous education, and initial English literacy level), on the program structure (whether the program was open or highly structured, the cost of the program per student), and on student persistence (hours per week of instruction, hours of lab instruction, and total hours of instruction).

For English language learners, persistence had a positive effect on literacy acquisition. On average, the ESL students experienced significant gains in reading comprehension by the end of the instruction. While their average pre-test scores indicated that most had only the literacy skills suitable for entry-level jobs, their average post-test scores indicated that most had attained the literacy level necessary for participation in job training programs that involve comprehension of basic English texts.

Several of the variables studied affected the studentsÕ gains in reading comprehension. The most important variable was the studentsÕ initial level of English literacy. Students with higher initial literacy were much more likely to significantly increase their literacy level during the program (pre-test beta=.66). Students with more years of previous education in their first language benefited more from the instruction than students with less educational background (total beta=.14)). Also instructional intensity (measured in hours per week of instruction) did not contribute directly to improved reading comprehension. Actually, more intense reading instruction had a negative effect on reading comprehension: English reading achievement tended to decline after about nine hours of instruction per week.The authors point out, however, that more intensive instruction was also associated with greater use of the lab and with larger classes. Smaller classes more focused on classroom instruction seem to be more effective than larger classes with a split class/lab approach. Increasing the total hours of instruction, or overall persistence, did positively impact reading comprehension scores (total beta=.20). Students who participated in more total hours of instruction ø spread out over a longer period of time -- were more likely to increase their reading comprehension than those who did not.

Several of the other variables addressed also seemed in impact reading comprehension gains.

The study makes good use of large amounts of quantitative data to understand the effects of different student background and program variables. The study integrated the investigation of student variables and program variables well. However, large quantitative research design prevented researchers from closely examining different aspects of the actual instruction in these programs. While this indicates that increasing instructional hours in literacy can help ESL learners to improve their reading comprehension, it does not indicate how the instructional time could best be used.Ê Furthermore, a problem exists with CASASÕs generalizability: as many people have pointed out, pre-post gains seem heavily related to whether the CASAS curriculum was used and not be related to gains on other literacy tests, even similar functional tests like the Test of Awareness of Language Segments (TALS).

Literacy instruction does affect adult ESL reading comprehension. Increasing the hours of instruction that a student experiences can be related to reading comprehension. The impact of student persistence underscores the negative effect of absenteeismÑby being absent, students decrease the hours of instruction they received and decrease their opportunities for improving their reading comprehension. Although it may seem obvious, these data seem to indicate that teachers should make students aware of the negative consequences of absenteeism. The data from this study suggest that smaller classes, with more highly trained teachers, structured curricula, and support services are more successful in improve the literacy level of the English language learners.

A further study could more closely examine the effects of persistence in different types of classrooms, in order to better understand the interaction of persistence and pedagogy.

Garreton, R. & Terdy, D. (1991). Correlation study of adult English as a Second Language (ESL) and Adult Basic Education (ABE) reading tests. Boston: Adult Learning Resource Center. (ERIC Document Reproduction Service No. ED 352 849)

A comparison between ESL adult studentsÕ scores on an ESL testÊ and their scores on two common frequently used adult basic education reading tests in order to standardize the reporting of adult education outcomes for minority language students.

Over 1500 adult ESL students enrolled in high-beginning through high-intermediate classes in community education programs in a major metropolitan area were tested. Subjects were included in the study only if they had taken all three of the tests. In the end, 300 sets of matched test scores were included in the analysis.

Reading comprehension sections from three major tests commonly used in community adult literacy programs were administered to the participants. The students took an ESL specific test (the CELSAÑComprehensive English Language Skills Assessment) and two adult basic education tests (the ABLEÑAdult Basic Learning ExaminationÑand the TABEÑTest of Adult Basic Education). Basic demographic information was collected from the students as well.

There was a correlation between the scores on the CELSA and the TABE. There was a weaker correlation between the scores on the CELS A and on the ABLE. However, the correlation between the CELSA and the ABE tests was not strong enough to predict scores on one test based on the scores on the other. Higher levels of language proficiency (based on class placement) were associated with stronger correlations between the scores on the two tests.

The study includes a careful explanation of the research method and the statistics. Test statistics are responsible and clearly explained.

While there is some correlation between the scores of this adult education test and the ESL test, especially for higher level students, this correlation is not strong enough to justify substituting the ABE test for ESL learners. ESL tests should be used rather than adult basic education tests for monitoring student progress in literacy learning. Given the higher correlation between the scores for the higher level learning, it is possible that learners could benefit from earlier mainstreaming into adult basic skills classes.Ê

Researchers could experiment with the effects of earlier mainstreaming of adult ESL students into other adult education training courses.

Goldberg, R. (1997). Deconstructing the great wall of print. Connections: A Journal of Adult Literacy,1, 8-13.

The effectiveness of pre-reading schema-building techniques on student comprehension of English texts.

The learners were adult ESL students enrolled in adult basic education classes. All the participants were Asian immigrants. The students sought to improve their English proficiency in order to continue their education at the community college or skills training program level in order to increase their job opportunities in the United States.

The researcher was the teacher of the class, and had been teaching at the center for several years. He compared his past experiences with student difficulties with reading to the experiences of this class.

The researcher implemented new reading techniques in the same curriculum he had taught in past years and reported on the success of the techniques based on comparisons of student reading behaviors from year to year. The techniques he employed were 1) previewing for schema building and 2) teaching students to read for information. Previewing was accomplished by introducing students to the themes of a text and asking them to free write on some of the issues in that text, as well as helping students make predictions about the text. To teach students to focus on the informational content, the teacher helped students to exploit headings, tables, and charts in order to understand the main points of the reading and to find information more efficiently.Ê

The researcher found that the students were less likely to be frustrated by the reading and less dependent on bilingual dictionaries if the texts were appropriately previewed before the reading began. He also found that reading for information content helped the students to process larger amounts of text more rapidly and with greater comprehension.

The study contains good descriptive detail of the desired reading behaviors and the ways in which the relatively simple innovations encouraged better reading strategies. The claims would be stronger if comprehension measures had been included ø and if the researcher had measured the studentsÕ actual rate of reading.

Relatively simple reading instructional techniques like previewing texts are often not included in curricula because of a lack of instructional time. This report suggests that it is more effective to use the time in previewing to help students comprehend the main points of the text than to spend time later helping them deal with the text word by word.

This work could be replicated with comparison groups and comprehension measures.

Goodman, K. (1988). The reading process. In P.L. Carrell, J. Devine, & D.E. Eskey (Eds.) Interactive approaches to second language reading, (pp. 11-21). Cambridge: Cambridge University Press.

The process of reading. Goodman sought to establish a psycholinguistic model of the actual process of connecting meaning to printed language.

Goodman begins with the premise that reading is more than simply decoding and that better understanding of the actual reading process will lead to better literacy pedagogy. Goodman defines reading as an active process in which the reader uses linguistic cues embedded in the text to reconstruct a message that the writer has encoded in written language.

GoodmanÕs reading model is based on oral miscue analysis. In oral miscue analysis, the reader is asked to read a text aloud. Differences between the written text and the oral reading are considered miscues. It is assumed that miscues are not random mistakes, but rather the result of the reading process unsuccessfully used. Thus, miscues can give researchers insights into the reading process itself.

GoodmanÕs model describes the reading process as a cyclical model, in which the mind processes the optical structure (the written symbols), the syntactic structure (relationships among words), and the meaning structure (the message the reader reconstructs). Processing the different structures to arrive at meaning is the goal of each reading cycle. Proficient readers move seamlessly from one cycle to another as they progress through a text. As readers move through cycles of reading, they employ certain processes. These are: recognition-initiation, which occurs when the brain recognizes written symbols as text and begins the reading process; prediction, which is the process of actively predicting what will follow in the text; confirmation, in which the brain seeks to verify previous predictions based on input; correction, which occurs when predictions are not verified and must be modified; and termination, which occurs when the reading is finished or stopped. While initiation and termination normally only occur once in a reading situation, prediction, confirmation, and correction occur cyclically throughout the reading. The success of the reading process is based on confirming or correcting predictions; in other words, successful reading ends in the reconstruction of meaning.

When meaning is not derived from a text, the reading has been unsuccessful. Goodman characterized unsuccessful reading as Òshort circuitsÓ. When reading is too difficult, for example, readers may short circuit meaning, seeking to understand only the syntactic structures. This is seen in the ESL classroom when readers are able to find the subject and object of a verb, but do not seem to process the meaning of the text.

GoodmanÕs original model of reading was an early psycholinguistic model of the reading process. Goodman asserts that previous models of reading had considered only linguistic cues, leaving out the cognitive processes involved in reading. This revised model makes use of the top-down predicting processes on which the first model was based, but also concedes a place for the bottom-up processing of written symbols. However, the model is still very focused on meaning-based, rather than print-based, aspects of reading. For example, of the five processes Goodman describes as occurring during reading, only oneÑinitiationÑis based on the recognition of linguistic cues. The others are more dependent on the meaning that the reader brings to the reading experience.

Any model of reading based on miscue analysis is more likely to explain unsuccessful reading than the processes of successful reading. Furthermore, GoodmanÕs assertion that he can correctly identify the reason for any given miscue and fit it into his categories for miscue analysis is dubious, at best.

GoodmanÕs model suggests that successful reading depends on good predicting skills. Readers who predict well will need fewer cues to comprehend text than poor predictors. Good predicting skills make the reading process easier and faster. By building good predicting skills, ESL learners may be able to compensate for lower language proficiency. ESL literacy teachers should help their students make predicting meaning part of the reading process. However, since these prediction skills are operative only if the reader can both decode and know the meanings of 90-95% of the words, prediction activities would have the most value if used where the readers know both the topic and the language of the readings.

Additional research could focus on the effectiveness of teaching predicting and confirming skills in improving studentsÕ reading comprehension and reading efficiency.

Gough, P.B. (1985). One second of reading. In H. Singer & R.B. Ruddell (Eds.) Theoretical models and processes of reading (pp. 661-86). Newark, DE: International Reading Association.

The nature of reading from a bottom-up, information-processing standpoint. Gough attempts to describe what happens in approximately 700 milliseconds, from the time that the eye fixates on the page until a sentence is processed.

In GoughÕs explanation of reading, it begins with eye fixation on text. At this point the visual information is formed into an icon, or a section of the printed page that is processed. This space is assumed to encompass twenty letters in the space of a two-inch by one-inch oval on the page. Once the icon is formed, the process of letter identification begins. Letter identification is extremely rapidÑGough estimates from letter acknowledgement studies that letters are identified at a rate of ten to twenty milliseconds per letter. Combined with eye fixation times, this yields an average reading time of three hundred words per minute. These speeds cause Gough to conclude that we do read letters in sequential order.

Next, the letters are mapped to the phonology system, either as full words orÑas Gough suggestsÑas systematic phonemes (morphophonemic units). The phonemic mapping allows the reader to match the letters to a word from the spoken lexicon. Invoking the phonology system allows for the unification of the lexicon; it assumes that readers have one lexicon for both spoken and written language, and that this lexicon is accessed through the phonology system for both. Once the words have been retrieved, they are transferred to primary, or working memory, where they are analyzed as a sentence or clause. After sentences are understood, they are transferred from working memory to free space for new incoming words. Gough theorizes that the sentences might be held in secondary (or short-term) memory, but in light of the paucity of evidence on this point, labels the transfer-point for processed sentences PWSGWTAU (the Place Where Sentences Go When They Are Understood). These sentences are still available to memory, so that they can be combined with the rest of the text to understand the text as a whole.

The model gives insights into the necessary equipment for reading. The beginning reader must have a visual (or in the case of blind readers, tactile) system capable of fixation on the textual input. Also, the model assumes that the reader must come to the task of reading with a lexicon, a comprehension device, and a phonological system. While these may not be fully developed, they must exist in at least rudimentary forms for reading to occur.

With these resources in place, the beginning reader must master both letter recognition and decoding (or the conversion of recognized letters into systematic phonemes). Gough suggests that this process is mastered through cryptanalysis, in which the beginning reader is presented with equivalent data in spoken and written form in a sufficient amount over time until connections are made. For children reading in their first language, this may be accomplished when they follow the text of a book while adults read out loud to them. At this point, the child can presumably make form-meaning connections, using the phonological system to call words from the lexicon and processes them as sentences.

GoughÕs work is one of the most complete and best substantiated of the various bottom-up models of the reading process. Readers must accomplish the tasks he suggests in order to process text. The problem with the model comes from its complete reliance on sequential processing. In light of studies of letter and word string processing, it is clear that information from texts is not uniformly processed in a sequential manner. As Rumelhart points out, sequential systems like this, while they elucidate the necessary component processes of reading, do not allow for non-sequential influence of components on each other. They thereby fail to capture the rich interactions between knowledge and skill systems that characterize proficient reading.

This model is useful for teachers, in that it explains the components of reading that can form the basis for strategies readers must develop to be successful. Whether this processing occurs in the sequential, non-interactive order that Gough explicates or not, readers must develop all of these skills in order to read successfully. Second language teachers of non-literate students need to help their students learn the basic processes of letter recognition and decoding; even literate students need practice mapping form to meaning in a second language. While bottom-up processing not constitute the whole of second language reading, it is part of the processes and should not be neglected in instruction.

The process and speed of letter recognition, the process of phonemic mapping, and the process of lexical retrieval in ESL reading could be studied. It would be interesting to contrast these processing strategies for learners with different levels of first language literacy.

Grabe, W., & Stoller, F. L. (2002). Comparing L1 and L2 reading. In Teaching and researching reading (pp. 40ø63). Harlow, England: Longman.

Differences that exist between reading in the first language (L1) and second language (L2). The chapter also looks at how these differences might influence classroom instruction. The information presented does not generally specify whether it pertains to K-12, adult, or university-level L2 readers. It is presumed that it is applicable to all L2 readers.

The research on reading comprehension in the L1 is extensive and complex for individuals from the age of three through university level. Yet the range of emphases in the research and the numerous learner variables make it difficult to make applications from the literature to any specifsic classroom context. The issues are even more complex in reading research in the L2.

Grabe and Stoller discuss 14 ways, in 3 basic areas -- linguistic and processing, individual and experiential, and socio-cultural and institutional -- in which L1 and L2 reading contexts may differ.

Linguistic and processing differences refer to vocabulary, grammar, discourse, orthography, metalinguistic and metacognitive issues, and amount of exposure to the language

L1 readers generally start to read with a base knowledge of 6,000 words. When they begin to read, they have been learning their language for 4-5 years. L2 learners, whether children or adults, typically learn to read simple sentences in the target language at the same time they are learning the language orally. They lack a lexical, grammatical, and discourse base in the language.

L2 readers may benefit from the explicit instruction they have received in the target language (as opposed to L1 readers for whom knowledge is tacit). L2 learners can use the vocabulary, grammar, and discourse knowledge they have learned in instruction to help them in reading comprehension. Additionally, some L2 learners may be able to draw upon what they have learned about reading in the L1 (e.g., what strategies work for them), to help them in reading in the L2.

Most L1 readers build fluency in reading in L1 through years of exposure to print. Most L2 readers do not have that degree of exposure to print in the L2.

The differences between L1 and L2 will influence reading comprehension in the L2. Differences in orthography from L1 to L2 (for example, an Arabic speaker learning to read in English) may cause difficulties in word recognition, reading rate, and fluency.

The language threshold hypothesis states that students must have a sufficient level of language proficiency in the L2 (vocabulary, grammar, and discourse) to be able to access the skills and strategies they use in L1 reading comprehension. The language threshold hypothesis says that L2 linguistic knowledge is more important than L1 reading ability until learners reach a level of fluency in L2 reading (Alderson, 1984). Research on L2 reading supports this (see, for example, Carrell, 1991). The difficulty with this hypothesis is that there is not a fixed set of language knowledge that must be reached. Rather the amount of linguistic knowledge and fluency needed varies with the task, topic, and text to be read. A student may pass over the threshold for one task, but may be below it for another.

The transfer of L1 reading knowledge may support or interfere with L2 reading. As for the transfer of skills and strategies learned in L1, research seems to show that this transfer is not automatic (Carrell, 1991; Koda, 1999; Strucker, in press). In other words, word-recognition skills, vocabulary-learning strategies, cognate use, and comprehension strategies may not occur without direct instruction.

The fact that reading in L2 involves the interplay of two languages is seldom discussed and has little research. It is a factor that affects such reading components as word recognition, reading rate, and strategies for comprehension. and others.

Individual and experiential differences refer to resources and experiences that influence L2 reading comprehension including learnersÕ proficiency level in L2 literacy, their personal experiences with and motivation for L1 and L2 reading, their attitude towards authentic texts, and their training in the use of various supporting resources.

Learners in L2 reading classes have differing levels of L1 reading ability. This should be examined in order to determine which skills and strategies to promote in direct instruction.

Differences in motivation are based on varying academic goals, prior educational background, and prior cultural frameworks and concepts of literacy.

L2 readers are not likely to be exposed to the full range of texts to which they are exposed in L1 in terms of length and complexity of language used in the texts.

ÊL2 readers often use bilingual and learner dictionaries while reading. They also may write out translations in the L1. Thirdly, they may access cultural knowledge (e.g, proverbs, special and sacred texts) from L1 to help them. These types of resources are not typically used in L1 reading.

Socio-cultural and institutional differences refer to cultural and social issues that operate outside the classroom context such as parental and community attitudes towards reading and the uses of literacy.

What literacy means, who is literate, and how literacy is used varies among cultures. For example, in some cultures, literacy has traditionally been the domain of a few letter writers and scribes. In others, everyone is expected to be literate. These attitudes toward what literacy is and who accesses it can influence learning to read in the L2.

L2 texts may not be organized in ways that match learnersÕ L1 reading experiences. Rhetorical organization may differ, as may use of nominalizaton and pronouns.

Expectations L2 readers bring with them from L1 may not match the realities of L2 instruction in terms of classroom management, teacher expertise, teacher-student relationships, and classroom size. These differences may affect their ability to continue to develop reading skills in the L2.

Instructors need to become informed about differences between L1 and L2 students. This can help instructors interpret reading research correctly in relation to the particular situations of their L2 students. Instructors need to know the level of literacy their learners have in the L1, their L1 language backgrounds, and their educational backgrounds. When learners are able to employ strategies in L1 reading, instructors need to identify and build on these strategies and skills (e. g., the use of word-recognition skills, vocabulary-learning strategies, and comprehension strategies) in the L2, providing direct instruction in their use. Instructors need to know the extent of learnersÕ exposure to print in the L2. If studentsÕ L1 languages have many cognates with English (such as French and Spanish), they should be instructed on ways to use the cognates in reading English. In order to take advantage of the skills and strategies of the L1, students should be exposed to sufficient reading in L2. Some of the reading should be at a fairly easy level and should be enjoyable to read in order to help improve reading fluency in the L2. Instructors need to assist learners in finding and using resources such as bilingual and learner dictionaries, so that learners will know the limitations of their use as well as the values. Instructors need to raise learnersÕ awareness of the rhetorical organization of text in the L2. Direct instruction should be given in language components such as nominalization and the use of pronouns in L2 reading.

Although Grabe and Stoller say there is no Òstraightforward blueprint for how teachers should adapt instruction for all L2 contexts, Ò (p. 62) they have discussed the many implications for practice described above. These implications seem to point to the fact that, given the variety of learners, learning contexts, and tasks, there is a need to explain to students why they are learning what they are learning and why they are learning it the way they are learning it. This is especially true for adult learners who are by nature motivated, bring prior experience to the situation, and are self-directed.

This article was clearly written and explained terms used in reading research well. It seemed, however, to be most useful in providing suggestions for instruction of learners at more advanced levels of L2 ability in that it discussed strategies and skills extensively at that level.

Grabe and Stoller discuss the need for classroom-based research. Especially for adult learners, large-scale research had not been conducted on reading in the L2, and instructors should not wait for this type of research but rather conduct research in their own classrooms, with their own learners, to see what works for them.

Areas for future research could include the usefulness of vocabulary study, the use of dictionaries to improve vocabulary development and reading comprehension, the usefulness of direct instruction in the development of reading strategies, and the usefulness of extensive reading. It seems, however, that research in all areas of L2 reading comprehension is warranted.

Griffin, S. M S. (1990). The effect of media choice on English literacy acquisition for preliterate Hmong adult learners. Unpublished doctoral dissertation, University of Washington, Seattle, WA.

The effectiveness of media used in ESL literacy acquisition with literacy-level learners. A comparison was made of learner-selected versus teacher-assigned media, and media that appeal to learnerÕs preferred perceptual modes compared with media that do not match preferred modes.

Fifty-four Hmong adult English language learners between the ages of 24 and 65 from Spokane, WA, and Fresno, CA, began the study. Of these, 37 ø 9 males and 28 females øcompleted it. The learners had no schooling beyond ESL instruction. Some had up to five years of sporadic instruction in English, but all had fewer than 200 hours of ESL instruction. Participants also had no literacy in their native language, and minimal literacy in English. All scored below 25 (out of possible 42) on the literacy section of the Washington State Adult Refugee Project ESL Oral Placement TEST (WSARPESLOTP). This means they were unable to identify all five of the single-digit numbers and five double-digit numbers on the test, as well as all five of the uppercase letters and five lower case letters.Ê This test was chosen because it distinguishes several levels of proficiency among learners who posses basic literacy skills in English.

Participants had been in United States from 6 months to 5 years, and were socially and linguistically isolated from the English-speaking community. They had not received literacy instruction or ESL instruction in the refugee camps.Ê Here in the United States, they had not responded to traditional methods of ESL literacy instruction previously provided. None of the 37 had obvious disabilities that may have affected their ability to acquire literacy. All were enrolled in ESL classes at the basic literacy level. They had been exposed to the Roman alphabet in ESL classes and recognized some environmental print (e.g., signs and labels).

Potential Hmong subjects took the WSARPESLOPT. Those scoring below 25 on the literacy second of the test were deemed eligible for the study and randomly assigned to one fo three groups: Groups 1 and 2 were given the Kerby-Learning Modality Test (chosen because it does not require literacy) to determine their preferred learning modalities. Group 3, the control group, was not given the test.

Media available for instruction was shown to Groups 1 and 2 by the teacher-researcher. Translators explained to participants that they would be asked to choose one medium for three instructional sessions.Ê Then pre-instruction interviews were conducted in Hmong with participants from Groups 1 and 2 to find out their experiences with language learning, their experience in using audio and video tape players, their opportunity to acquaint themselves with the media used in the study, and their preferences for one of the media ø visual, auditory, or kinesthetic ø offered for use in three ESL literacy instructional sessions. Subjects in Group 3 were shown flash cards with English words on them and returned to their ESL classes.

The so-called media available to the participants were 1) auditory- mode: a video presentation of the vocabulary items presented on flashcards for 15 seconds. The word were pronounced in English and then after a short silence translated into Hmong. The students repeated the vocabulary item. The tape was shown three times.Ê 2) visual-mode: Color photographs were presented simultaneously with flash cards displaying the name of the objects in the photographs.Ê As these items were displayed, the teacher pronounced the vocabulary item in English and the students repeated it. After this was done a second time, the students were to match the photographs to the vocabulary items.Ê 3) kinesthetic mode: the teacher displayed the flashcard , pronounced the vocabulary items printed on them. Wile the students repeated the teacher demonstrated the activity described on the flash card or held up the item named. This was repeated. The third time theÊ students were given flashcard and encouraged to act out the activity or pick up the object names.

For three one-half hour segments on consecutive days, learning sessions were conducted with group 1 using media they had chosen, and with group 2 using the media assigned by teacher-researcher. A total of 16 vocabulary items and 12 numerals were presented in these sessions. After each session the groups returned to their regular ESL classes. The teacher-researcher made observation notes, took photographs, and audio-taped the participantsÕ reactions to the three learning sessions. The participants were interviewed, again in their native Hmong, and asked if they felt the media instruction had helped them learn the English words and phrases on the flash cards. Finally all three groups were given a Òliteracy achievement testÓ where four 9Ó X 12Ó manila flash cards were show to the participants. On each card, a simple task was described in the vocabulary that had been taught (e.g. ÒPick up the penÓ). Participants were told in Hmong to read the flash card aloud in English, then explain in Hmong the meaning of the words or phrases on the card, and perform the task indicated.ÊÊÊ

ParticipantsÕ pre-instruction in English literacy significantly affected their further acquisition of English literacy: ParticipantsÕ pre-instruction literacy scores on the WSARPESLOPT were significant predictors of their post-instruction achievement scores. For the entire population, there was a strong, positive correlation between pre-instruction literacy scores and achievement scores. This was the most powerful indicator of further acquisition of English literacy

Participants whose classroom instruction in ESL was not supplemented by media-based instruction did not acquire as much English literacy during the course of the study as participants with comparable pre-instruction proficiency in English literacy who had participated in these supplemental sessions.

Participants who were instructed in media they chose did not perform significantly differently from those instructed in media chosen for them by the teacher. Similarly, participants who expressed belief in pre-interview that the media they used would help them acquire more English literacy did not score significantly different from those who did not have this belief.Ê

Participants who were instructed with the photographs and those instructed by bilingual video scored, on the whole, 25 more points then those instructed by live teacher. This was in spite of the fact that the combination of oral presentation and demonstration is a traditional method of instruction in Hmong culture.ÊÊ

Given both learner and program factors (e.g., lack of familiarity with western research methods, desire to be included in all activities, tolerance of noise and closeness, relative lack of privacy among classes due to crowded rooms and barriers that did not stop noise, etc.) a truly clinical study was not possible. The teacher-researcher was well aware of this. She mentions that many of the learners who did not receive the additional sessions, would ø during their breaks ø crowd around those in the instructional sessions and watch and practice the video activities with them. In addition, the n is very small for the students who did not receive instruction returning for the post-test.Ê At one site, only 5 of the 15 returned for literacy achievement test. The teacher-researcher discusses all of these issues at length.

What is not discussed, and is perplexing is why the students were not pre-tested with the achievement test with which they were post-tested. Whereas it is true that the participants all had similar literacy scores as determined by the WSARPESLOPT, the post-test given was not the same as the pre-test.Ê

Given that previous literacy instruction seemed to be the best predictor of success for these learners, literacy level learners should receive instruction separately from those with previous literacy acquisition. Classes should be small, as those in the instructional groups 1 and 2 both received instruction in groups of 2-3, whereas the ESL classes had 25-30 persons. If it is not possible to offer separate small classes, there should be tutorials available for literacy level learners. Whereas live teacher demonstrations appear to be effective for facilitating the acquisition of oral/aural skills, they may actually inhibit the acquisition of English literacy skills, by overloading learners with verbal and nonverbal information. Literacy instruction should focus on literacy, not on speaking. The most effective media for teaching English literacy skills to Hmong adults appear to be those that focus their attention on literacy while linking speech and writing. For example, in this study, a bilingual video-tape showed flashcard images or flash cards showed photographic images. Furthermore, translation of English words into the learnerÕs native language appears to be an effective method of instruction in the earliest stages of literacy learning.

This study could be replicated with a larger group of learners. Also, it could be done with another group, perhaps a mixed language/culture group, to see if similar results are yielded on teacher demonstration versus lack of distraction of media, or if this is unique to Hmong learners. If the study were done again, the same pre- and post-test should be used.

Hardman, J.C. (1999). A community of learners: Cambodians in an adult ESL classroom. Language Teaching Research, 3, 145-66.

The English literacy development of Cambodian immigrants to the United States. The researcher discusses the involvement of the participants, their families, and their teachers in the development of literacy by drawing on multiple ways of sharing knowledge and creating learning environments.

The research took place at a community association building, located in the heart of the Cambodian community. The learners were mostly Cambodian females in their twenties and thirties. All lived within a few blocks from the school and many were mothers of children who attended school nearby. Most of the students came from farming families in rural areas of western Cambodia. Few had more than three years of schooling in Cambodia. While most were able to read definitions from a Khmer-English dictionary, there was wide variation in their literacy beyond that. It is unclear how many students were included in the study because the author does not give this information. Rather, he says that 10 ø15 students were in the Khmer teachers room daily, while 5 were in the authorÕs.Ê It is unclear whether they were always the same 10 ø 15 and the same 5 students in class, so the total number of students in this study cannot really be determined.ÊÊ

For the most part, the study draws on observations of two teachers in the center. One is the researcher as well, who did participant observations of his classes. He is an American with very basic knowledge of Khmer. In his intermediate level class, he saw his role as that of a helper and a resource. Similarly, he saw the role of older Cambodian children who accompanied their parents to class as a resource to help their parents learn.

The other teacher was a Cambodian who had learned English over several years in the United States as a Buddhist monk. He closely followed a typical style of Cambodian teaching in which the teachers are very strict. His students did not turn to their children for help with English. The teacher used Khmer extensively to support the English language materials used in class.

The data for this study were collected as part of a five-year ethnographic study of a Cambodian community in the United States. The researcher observed and interviewed the participants and their families at home, and, as a teacher in a community center ESL program, was a participant observer of the literacy learning process in two different classes over the course of a year. He also interviewed other teachers and staff from the school, and collected Òartifacts of literacyÓ including student writings.

Based on the data drawn from field notes and interviews, the researcher created a typology of interactive structures among the Cambodian students. The structures deal with the behavior of the adult students when working with each other, with the children, and with their teachers.

The adults interacted with each other as Òknowledge holders.Ó In this role, they assisted one another with answers in class, on the homework, and on tests. They also interacted as Òfacilitators,Ó translating information relating to classroom management such as instructions. To do so, they drew on their knowledge of both oral and written English, as well as oral and written Khmer.

Children frequently accompanied their parents to class. The children often had higher English literacy and better English oral proficiency than their parents, but were not literate in Khmer. They also interacted with the adults as knowledge holders and facilitators, but also functioned as Òevaluators,Ó correcting their parentsÕ mistakes in class both in written and oral English. While adults often used Khmer writing to exchange knowledge, children only used oral Khmer to give information to their parents.

In addition to fulfilling similar roles as the Cambodian children, the native Cambodian teacher also acted as an authority figure. Having extensive previous education in Khmer and Vietnamese, he was seen as both a Cambodian and English authority. His teaching style included a great deal of copying and oral reading on the part of the learners, which he prompted, corrected, and praised. Like the American teacher, he presented and explained the language, but he did so making use of both Cambodian and English. Because of his ability to fulfill both teaching and translating roles, the interactions in his class were mostly between himself and the students, not between students or between students and their children. Because of this, his classroom more closely reflected the cultural norm of Cambodian classroom, whereas the monolingual teacherÕs classroom more closely resembled the studentsÕ home environment.

This study reports on extensive data collection with rich, detailed reporting, painting a very clear picture of the classes and the interactions that take place in them. Supplementing the observations with interviews with the students, teachers, and family members clarifies the nature of the interactions that take place. The additional observations at home add more complexity and completeness to the descriptions.Ê For teachers not familiar with the culture, the information on school attitudes of Cambodians could be useful.

Although this is an excellent descriptive study, it does not document the effectiveness of the different sorts of classroom interactions mentioned in terms of language learning. There is no assessment data of any kind for this.ÊÊ

The classroom is one of the interactional settings for members of a minority community. Teachers should strive to understand the interactions of their students in this and in other settings to understand how the classroom builds on and contributes to the literacy experiences of the students.

The effect of these different interactions and of the different types of classrooms could be studied to determine whether one or the other is more beneficial for literacy acquisition.Ê

Haynes, M. (1993). Patterns and perils of guessing second language reading. In T. Huckin, M. Haynes, & J. Coady (Eds.) Second language reading and vocabulary learning (pp. 46-64).

The effectiveness of guessing the meanings of words in second language reading. Many ESL reading textbooks advocate the abandonment of bilingual dictionaries, asserting instead that students should be encouraged to guess meaning from context. The author wanted to evaluate the effect of this strategy on reading comprehension.

Sixty-three learners from an intensive English center in the United States volunteered to participate in the research. Learners spoke Spanish, Japanese, or Arabic as their native languages (Tunisian Arabic speakers were considered separately from other Arabic speakers because of their extensive educational background in French). The learners had a variety of English proficiency levels, according to the centerÕs proficiency exam. Most of the Spanish speakers were clustered in the lower levels of proficiency, while most of the Japanese and Arabic speakers were clustered into the higher levels of proficiency.

The students read two stories about animals. The stories had parallel structures and syntax. Each story had two nonsense words embedded in it. One of the words was only used once and could be guessed using the immediate context, through a local context clue (Laufer, 1997). A local clue is a context clue in the same sentences as the unknown word. For example, if a student didn't know the word "partridge", and read the following sentence" The partridge spread her wings, ruffled her feathers, and chirped loudly." the words "wings" "feathers" and "chirped" would all be local context cues. The reader might well be able to infer that the animal being described is a bird. However, in most texts not designed specifically for ESL speakers, sentences do not contain much redundant information in the same sentence. In these cases, there tend to be global cues, which are context cues derived from understanding the meaning of the passage as a whole. An example of this is the passage "It was early morning when the partridge lifted her head. The first rays of the sun fell through the branches of the tree and lit up the leaves surrounding the nest. Through the morning mist, she could see the others taking flight." The reader cannot tell from the text what kind of an animal a "partridge" is and will have to infer the meaning of the word not from the immediate context of the sentence but from the meaning of the passage as whole. In this study, many of the clues were global: All the word was used repeatedly (in each story, it was the type of animal); its meaning had to be guessed using the global context of the story. The students were instructed to read the passages and then retell the stories to the researcher. After reading and retelling both stories, the students were shown each passage again and asked to point out words that made the story difficult to understand. As the students pointed out words, they were asked to guess their meaning and explain the reason for their guesses.

Across all groups, students were significantly more likely to guess the meaning of the nonsense word correctly when the contextual clues were local rather than global. Their ability to guess the meaning of words in local contexts, however, was undermined in the first passage because many of the students did not know the meaning of the word in the sentence that was used as a contextual clue. This is one of the biggest limitations of the use of contextual cluesÑseveral times the researcher found that students were not able to make use of clues because they did not understand the clues themselves. Considering the text as a whole, and not just the nonsense words, students were unlikely to guess word meanings in global contexts; they were better able to guess word meanings when the unknown word appeared along with comprehensible contextual clues.

When contextual clues were unavailable or incomprehensible, students commonly resorted to word analysis to guess meanings. While this was successful with some words, it was problematic for many others. For example, Spanish and Tunisian learners often analyzed the word campfire according to Spanish and French cognates which have military connotations. They then interpreted a passage about a scavenging animal as a story about a soldier running away from battle. This illustrates the negative impact guessing word meanings can have on reading comprehension. Japanese and other Arabic students also made unsuccessful guesses when they sought to analyze words. For example, several Japanese students tried to related habitat to habit, guessing its meaning to relate to habitual actions. Similarly, students from all language backgrounds tried to break the word offspring into its component parts, arriving at guesses like Òthe end of spring.Ó

A third problem with guessing came from mismatching background knowledge to the written stimulus; the word they processed from memory was spelled/pronounced differently than the word in the text. For example, several students unsuccessfully guessed the meaning of tapped because they related it to top. Strikingly, many mismatches were syntactically impossible in the context, which could indicate that instructing students to guess at word meaning diverts attention from syntactic processing.

Lastly, processing time was an issue. Students needed time to decide if a word was new or unfamiliar, indicating that vocabulary recognition was not automatic for them in English at this point. Once they determined that they needed to guess word meaning, they needed more time to evaluate and guess. This could indicate that guessing words in context is not a time-saving strategy that allows students to process more input faster, as other reading theorists have asserted.

The researcher carefully explains the problems that exist when readers rely on guessing word meanings as the major strategy for reading in a second language. The results are carefully explained and well exemplified.

Because student proficiency levels and language background were not separated as variables in the sampling, it is not possible to draw conclusions on the effect of language proficiency or language background on successful guessing of word meanings.

Teachers should be aware that guessing from context may not always be productive and reading comprehension may be impaired students are urged to rely too heavily on guessing meaning from context. Vocabulary development should not be neglected by reading teachers. Use of monolingual and bilingual dictionaries should not be ruled out.

Researchers could investigate the longitudinal impact of guessing for meaning in context by tracing the development of learners who do and do not use dictionaries in their reading.

Hilferty, A.G. (1996). Coding decoding: Predicting the reading comprehension of Latino adults learning English. Unpublished doctoral dissertation, Harvard University, Cambridge, MA.

The relationship of L2 decoding skills to other reading and language sub-skills in the reading performance of 42 adult Latino learners of elementary English as a second language reading. The study was performed in two phases. Phase I was a correlational study where data from the study was analyzed quantitatively to identify patterns of relationships among selected reading sub-skills and personal variables. Phase II was an observational study of the ESL instruction of adult Latino ESL readers, and explored the utility of the correlational data in the classroom context.Ê

Forty-two Spanish-speaking male and female adults attending or recently completing ESL classes for adults in the Cambridge/Somerville area were selected for the study. Participants ranged in age from 16 ø 53, with a mean level of 35.7. Level of education ranged from four to 20 years with a mean level of 13.7 years. Length of stay in the United States ranged from three to 216 months with a mean length of stay of 73.31 (about 6 years). Length of attendance in ESL classes ranged from a mean length of one to 60 months. At least 24 of the subjects were employed in full or part-time jobs. Thirty-eight of the learners completed the study.

Instruments were chosen to measure language and literacy skills in Spanish and English, including tests of reading comprehension, listening as a measure of oral proficiency, phonemic awareness, decoding through letter/word recognition skills and pseudo-word recognition skills, and a letter-word recognition time measure. Reading comprehension in English was the dependent variable. Decoding skills in English, as measured by letter/word and pseudo-word recognition were the question variables.

1. Reading: Bateria Woodcock Psico-Educatovio en Espa.Hol (BWPEE) and the Woodcock-Johnson Psycho- Educational Battery (WJPEB). The BWPEE had been adapted from the WJPEB.

2. Oral proficiency: Because measure of oral proficiency were not yet available for use with Woodcock øJohnson batteries, oral proficiency was measure using Spanish and English listening subtests from the Language Assessment Battery (LAB) of the New York City Public Schools. In the LAB, the subjects select the correct response to an oral stimulus by reading the possible answers silently and marking them in a test booklet.

3. English phonemic awareness: Rosner Test of Auditory Awareness Skills (TAAS). Spanish phonemic awareness: an adaptation of the TAAS developed by staff of the Cambridge Community Language Center. In the TAAS, the subject responds orally and the researcher records the answer.

All tests had corresponding grade levels attached to them. Participants were also interviewed in Spanish to find out information about educational levels in Spanish, time in the United States, and whether or not they were working.

The data were analyzed in two ways: 1. SubjectsÕ patterns of performance were viewed in a correlation matrix displaying relationships among all the variables in the study. 2. Relationships between the measures and the outcome variable were examined in multiple regression analysis.

SubjectÕs mean grade equivalences in English listening were substantially higher and closer to the counterpart Spanish grade equivalents than were their mean grade equivalences in English word attack, reading comprehension, and phonemic awareness.

The four measures that correlated most highly with English reading comprehension were English letter/word identification, English listening, English pseudo-word identification, and English phonemic awareness. In an analysis of the Spanish and English language and reading sub-skills, English decoding accounted for 15% of the subjectsÕ reading comprehension.

In short, the major reading variables are related to each other across languages: Spanish and English reading comprehension were correlated, and Spanish and English letter/word identification were related to each other.

However the variables related to oral and phonetic aspects of language were not related to each other: Spanish and English listening, Spanish and English phonemic awareness and Spanish and English word attack are not related. Some of the variables that were unrelated to each other across the languages were related to other sub-skills across the language. For example, although Spanish phonemic awareness was not related to phonemic awareness in English, it was related to English reading comprehension, suggesting a Spanish to English effect. Furthermore, oral proficiency and literacy skills in both languages interact with reading sub-skills.

Four predictors that had significant correlation with English reading comprehension were motherÕs education, Spanish listening, Spanish reading, and Spanish phonemic awareness.

English reading comprehension had a weak relationship with subjectÕs education and no relationship with the following: fatherÕs education, age of arrival, the Spanish word-naming variables, Spanish letter/word identification, and Spanish pseudo-word identification.

The implications of the results of the correlational study crossed two areas of application: instruction and research. The two questions to be answered by an observational study were:

1. If ESL decoding is a significant predictor of reading comprehension for L1-litearcy Latino adults reading ESL at approximately a grade 3 level, what are the implications for class groupings and for instruction?

Five participants, all women in their 40s, were selected for the case study research for the observational study. The participants came from three Latin American countries. Two had been in the U.S. for ten years, and had delayed beginning their ESL studies for over eight years after arrival. The other three had lived in the U.S. for less than two years, but had begun studying English within a year of arrival. All had least high school education in their countries some had vocational and one had a college degree. They all had scored fairly high in Spanish oral proficiency and fairly low in English oral proficiency. All had scored low on English reading comprehension.

The observational case studies took place over a period of four months from December 1995 ø March 1996. The purpose was to visit the classes of these five learners to see whether creating a double context of statistical and observational data would lead to new information about adult acquisition of second language reading and new ideas for further studies.

Descriptive information was gathered from observations of classroom activities with literacy components by a team of two observers: one English dominant and one Spanish dominant.

2. Are their levels of L2 decoding skill apparent in their classroom performance?

5.Is it feasible to accommodate decoding in ESL instruction in the adult ESL class?

Only one of the classes (which included two of the subjects) included a high number of activities relevant to decoding. Two of the classes had a few activities linked to decoding and one had almost none. Although all classes observed contained some elements of instruction related to decoding, decoding activities were not usually delivered through direct instruction, and were seldom part of teachersÕ long-term plans for instruction, as determined through the class observations and the teacher interviews. Furthermore, none of the learners interviewed specified decoding as a goal. Theys were much more focused on listening and speaking and were happy with the instruction they were receiving.

All of the women observed had scored low in English reading comprehension, however those who had delayed their study after arrival in the United States for several years had especially low measures of English letter-word identification and phonemic awareness.Ê

1. Would some instruction contain elements of the instruction considered optimal for normal child L1 readers?

Some, yes. But in most of the classes, these elements of instruction were incidental to other goals of the teachers and were not part of the longer-term, conscious goals on the part of the teacher.

Ê2. Would the information from this correlational study be usable in the planning and conducting of the instruction observed? In others words, would it be feasible to teach English decoding in the ESL classroom?

Yes. There were many opportunities to use decoding in the ESL classrooms. There was a lot of environmental print as well as instructional print. Class activities such as one using the song ÒI just called to say I love youÓ (Stevie Wonder) provided lots of opportunities for decoding and word recognition that were not exploited.

1. Adult Latino learners at the beginning level of reading in English would benefit from instruction and practice in English decoding as well as in English listening. This is consistent with the theory ø prevalent from research with L1 reading in children ø that suggest that mastery of lower-level reading skills is necessary for the development of higher-level skills in reading.

2. The transition from decoding Spanish to decoding English is likely to be more difficult than the reverse. The evidence of a possible transfer from L1 to L2 decoding suggests that for such students, it may make more sense to master the more transparent Spanish decoding system before tackling the English system. However, there appears to be no full, automatic transfer of Spanish decoding to English decoding skills. It appears that students independently acquire second language word recognitions skills.

3. It is quite possible that the relationship between reading comprehension and decoding is reciprocal. Therefore, classroom activities should include instruction in both skills.

4. Even for subjects who have fairly high levels of education in their first language, working with printed language in English is hard. Oral skills do not automatically transfer to print.Ê Deliberate instruction is needed in reading.

5. Teacher education should include instruction on the importance of decoding to beginning-level readers of English and in ways to foster this.

Ê6. Delaying the study of a new language for several years may inhibit the development of word recognition skills.

This study seems to be well planned and well executed. The combination of the correlation study and the observations is worthwhile. Of particular interest was the importance decoding appeared to play in reading comprehension and the fact that it did not appear to automatically transfer from Spanish to English. Also of interest was the fact that decoding was not being directly taught in the classrooms.

However, the small sample size (38) and the very small number of case studiesÊ caution against too much generalization from the findings. Also problematic was the use of some of the tests: the English listening test had been normed on children only. Furthermore, it was not a true listening test in that it was a multiple-choice pencil and paper test that required reading ability. Since there were no beginning level literacy learners this was probably not an issue for this study, but would be if students with fewer L1 and L2 literacy skills had been tested.

Decoding
L1 literacy
Reading skills transfer
Reading tests
Second language development
Second language proficiency
Teacher Education
Written-oral/reading relationships

More research should be undertaken on the role of decoding in adult ESL acquisition ø when and how is it most useful. This study could be replicated with a larger group of learners. It could also be done with learners whose L1 does not use the Roman alphabet. Further research could focus on comparing adult patterns of learning and the patterns of the child L1 learner. Research should be done on whether there are some adults for whom initial instruction in L1 decoding should precede L2. Research is needed on the effects of different kinds of class composition on language learning ø whether placement by oral proficiency only meets the needs of learners and whether compartmentalizing English is a good idea, or if integration of skills is not preferable. Research is needed on the effects of delaying instruction for years after arrival in the United States: A study could be done testing the reading skills of adult learners at similar levels of total English proficiency who vary by length of time spent in the United States before beginning instruction.

Hsueh-chao, M.H. & Nation, P. (2000). Unknown vocabulary density and reading comprehension. Reading in a Foreign Language, 13(1), 403-430.

The relationship between vocabulary knowledge and text comprehension for ESL readers. The researchers attempted to determine what percentage of words in the text learners needed to understand in order to engage in reading with comprehension without looking up unknown words.

The learners in the study were adult ESL learners in a pre-university English course in an English-speaking country. Most were preparing to study at an English speaking university. They were chosen based on their scores on the Vocabulary Levels Test (Nation,1990). They needed to score more than 14 of 18 at the 2000 word level as well as more than 10 out of 18 at the 3000 level.

Four versions of a text were prepared for the study. The text was 673 words long and was written as a first person narrative. In three versions of the text, low frequency words in the text were replaced by nonsense words (nonsense words were used in order to guarantee that no learners knew the words). One version (95% coverage) had 5% of the words replaced by nonsense words. The 90% coverage version had 10% of the words replaced and the 80% version had 20% of the words replaced. Words were chosen based on Francis and KuceraÕs (1982) frequency list. They were replaced in reverse order of their frequency, with the least frequent words replaced first. Words appearing multiple times were always replaced by the same nonsense words. After the replacements were made, the texts were simplified to guarantee that only the nonsense words would be unknown. All words that had not been replaced by nonsense words had to be within the most frequent 2000 words of English (well within the vocabulary level of the subjects).

The students were randomly assigned to read one of the four passages. Following their reading, they each completed three tasks. The first task was a multiple choice test containing fourteen items dealing with explicit and implicit information from the text. The questions dealt with ideas determined by a native speaker to be in the top 50% of the idea of the story based on importance of the idea to the text. One point was given for each correct answer.

The second task was a cued-written recall. Students were given questions as prompts. Their answers would supply information that served to reconstruct the story. The aim of the written recall was to get learners to produce as much as they could about the story. In all, 27 cues (all in the form of questions) were given. The recall was scored by weighting each idea according to its importance in the story. If a student recalled one of the top 25% most important ideas, four points were awarded. For recall of the next 25% most important ideas, three points were awarded, etc.

Finally, learners were asked to predict what their scores on the tests would be. The purpose of this was to determine how the different amounts of unknown vocabulary of the text affected student perceptions of comprehension.

On the multiple-choice measure, the mean score for each group clearly decreases as the percentage of nonsense words increases. No student who read the 80% coverage text, for example, scored above 10/14, while this is the lowest score for students who read the 100% coverage text. The results for the recall were very similarÑas the percentage of nonsense words increased, student ability to recall decreased. Statistical measures indicated that 48% of the variation in student scores could be explained by the difference in the amount of nonsense words, indicating that the percentage of unknown words in a text is an important factor in comprehension. The authors looked for evidence of a vocabulary threshold. A sudden jump in comprehension at a particular level of vocabulary understand would indicate that a threshold had been crossed. The study does not find evidence for a threshold, in that there is no sudden leap in comprehension at any one point, but rather a steady relationship between the percentage of words known and comprehension. Across groups, student predictions of their scores were fairly accurate, indicating that students are aware of miscomprehension caused by unfamiliar vocabulary.

The study is very well controlled, and the measures are well designed. The recall measure seems to be somewhat redundant to the multiple choice measure, since the students were asked so many questions that they were more likely to simply answer the questions than attempt any independent recall. The recall could be better considered as a short-answer test than a true recall. This similarity likely accounts for the high correlation of multiple-choice scores and recall scores. The scoring of recalls is somewhat counter-intuitive. The top 25% most important ideas, which were most likely also the most salient ideas, were awarded the most points. It would seem more logical to award more points for comprehending the less salient ideas. No rationale for this decision is given. In addition, the results, while suggestive, are not necessarily the same that would be found if a students encountered similar percentages of unknown actual words, where they could have applied intraword analysis or other word-level strategies to guess meanings.

Vocabulary knowledge, while not the only component of reading, seems to be one of the most important knowledge areas for second language readers. Based on the regression analysis performed by the researchers, students reading narrative texts need to know at least 98% of the vocabulary they encounter in the texts in order to read with comprehension without needing to use a dictionary. Other reading skills like guessing meanings in context will not be able to compensate for vocabulary knowledge lower than this level. Given this high level of comprehension needed before reading in context strategies are effective, pre-teaching of unknown vocabulary seems to be indicated for most learners with most texts.

This is a very well designed research study. It could be replicated among a different population, or it could be replicated using texts on different topics with different groups to determine how vocabulary knowledge and schematic knowledge interface in reading comprehension.

Hudson, T. (1982). The effects of induced schemata on the Òshort circuitÓ in L2 reading: Non-decoding factors in L2 reading performance. Language Learning, 32, 3-31.

The effects of inducing schema activation on studentsÕ reading comprehension. Much research has shown that second language readers, in applying background knowledge to a text, will ignore or alter contextual clues to bend the text to the schema they wish to apply. The author wanted to determine whether inducing the correct schema either before or during the reading process could increase learnersÕ comprehension of a text.

The participants were ninety-three students enrolled in an intensive English language institute. The students came from various linguistic backgrounds. They were divided into beginning, intermediate, and advanced levels based on the reading class in which they were currently enrolled.

Three reading passages dealing with scientific topics were prepared for each proficiency level. For each reading passage, three different types of schema-inducing activities were designed. The first, pre-reading, involved discussion of pictures related to the topic of the reading. The second, vocabulary, involved students receiving and discussing a list of relevant vocabulary before reading. For each of these treatments, the schema-inducing activity preceded reading the article, which was followed by a multiple choice comprehension test. In the third treatment, the students read the article and answered the multiple-choice questions. They were then given the article again and retook the original test. The purpose of this treatment was to see if post-reading activities (answering comprehension questions after reading the text) could induce schema.

Each proficiency group was given all three types of reading treatments for the articles selected for their proficiency level. Within each proficiency level, the students were divided into three groups, so each group performed the activities in a different order and with different articles. The treatment took place during normally scheduled reading labs. Looks like a good design.

Judging by the scores on the post-test when compared to those of the pre-test, for the beginning students, the pre-reading discussion treatment was significantly more effective than the pre-reading vocabulary treatment or the post-reading treatment. For the intermediate students, both the pre-reading treatment and the vocabulary treatment were more effective than the post-reading treatment. For the advanced learners, however, the post-reading treatment was the most effective, although there were no significant differences among scores received by those the treatments. The researcher interprets these findings as evidence that schema-induction can have an effect on comprehension at different proficiency levels. He also concludes that the reading strategies students use differ by level. The advanced students, for example, seem to apply schema more flexibly, as they are able to modify their schema after reading, while beginning and intermediate students are not. Thus, schema-induction should be attuned to the proficiency level and strategy use of the students.

The study is very well designed and the findings are carefully interpreted. The researcher also wisely cautions readers to consider the possible effects of conducting this research in a laboratory, rather than a classroom setting and of using multiple-choice tests as a comprehension measure. A student can possibly have high scores on a multiple-choice test without having successfully reconciled the schema to the reading. Multiple-choice tests reveal information about general comprehension and about test-taking skills, and may not be fine-tuned enough to detect the effects of poorly reconciled schemata. It is likely that the effects for advanced learners were small for all treatments in part because they may have already gotten 90% of the texts, so there was not much room for improvement.

Teachers can help students build appropriate schema to comprehend new texts. This could increase student confidence and comfort in reading at any level, and increase the amount of comprehensible input through reading the students receive. However, the level of the student should be taken into consideration when choosing schema-building activities. Beginning and intermediate students may benefit more from activities that encourage them to build schema before reading. Advanced level students appear to be able to benefit from schema-building activities whenever they are presented, whether before or after the reading.

If students do change their reading strategies as they become more proficient, longitudinal studies could possibly give us information about when and how strategies are changed.

Huntley, H.S. (1992). The new illiteracy: A study of the pedagogic principles of teaching English as a second language to non-literate adults. Unpublished manuscript. (ERIC Document Reproduction Service No. ED 356685)

Different learner types defined in terms of native language literacy and the pedagogical implications of working with each type of learner. The discussion is based on both previous research and the authorÕs experience.

Much research has indicated that non-literate adult ESL learners have special literacy needs that are often not addressed in traditional ESL classrooms. Huntley begins the study with definitions of literacy in terms of functional ability and with a consideration of the dual burden that non-literate ESL learners face of learning both oral and written language at the same time. After considering proposals that written language should not be taught until after learners have achieved a highly functional command of oral language, she concludes that the most effective manner of instruction for non-literate ESL adults is one that integrates oral and written language learning. She then discusses different types of ESL non-literate learners and addresses pedagogical approaches to meet their needs.

There are four types of ESL learners who are in need of ESL literacy instruction: pre-literate, non-literate, semi-literate, and non-roman alphabetic literate. Pre-literate ESL learners come from cultures in which literacy is rare or non-existent. These learners have had no or extremely limited exposure to written language. Pre-literate learners generally have a slow-paced learning rate, and require repetition and re-teaching. They need to begin literacy instruction with practice in discriminating different letters and with comparing sight words before they will be ready for literacy instruction. The teacher needs to make it clear to this group of learners that the written symbols are a method of communicating, and that reading will eventually be meaningful.

Non-literate ESL adults come from cultures where literacy is fairly common, but have never learned to read. While they do not read, they have had the opportunity to notice printed language, and are more aware of the role of reading in society. These learners likely feel stigmatized because they cannot read, and many will be unwilling to disclose their illiteracy. Teachers should be very sensitive to these learnersÕ affective needs.

Semi-literate ESL adults have not progressed beyond elementary level literacy in their native languages. They are likely as well to have acquired basic English literacy skills, but not to have acquired functional literacy to meet their personal or employment needs. Teachers can help these learners build on their literacy skills through extensive reading at an appropriate difficulty.

Non-roman alphabetic ESL adult learners can be highly literate in their first language, but are unused to the Roman alphabet. Their first language literacy is still an important advantage; these learners can still transfer cognitive skills of reading, but need practice with forming letters and numbers and with sound-symbol correspondences.

The study gives a good overview of some of the important issues teachers face in adult ESL literacy situations. Reviewing student needs in terms of different types of literacy exposure and literacy needs is a helpful exercise for determining what and how to teach reading to different students. The categories Huntley has suggested should be useful for teachers.

Teachers should be aware of the different types of non-literate students in their classes and needs of those students. Teachers and tutors can customize the literacy instruction for students with different literacy needs.

Researchers could investigate different pre-literacy techniques to determine the effectiveness for pre-literate students. Very little research has been done on how to help students who read in non-Roman alphabets or those who read ideograms learn to read with the Roman alphabet.

Joe, A. (1998). What effects do text-based tasks promoting generation have on incidental vocabulary acquisition? Applied Linguistics, 19 (3), 357-377.

The relationship between reading and recall tasks and incidental vocabulary acquisition.The researcher wanted to see if reader engagement in reading and recall would help the learners acquire vocabulary without specifically focusing on vocabulary learning. The researcher also wanted to determine whether instructing learners on reading strategies to help them integrate new and known information would increase the effectiveness of reading tasks for incidental vocabulary learning.

There were a total of 48 adult English learners who were studying in two different non-credit courses in Wellington, New Zealand. Forty-one were studying in a 3-month intensive English proficiency course; the remaining 7 were participating in an ESL program in a technical institute. Fifteen were males, 36 were females and their ages ranged from 19-46 years. Approximately three-quarters of the subjects were Asian; the remaining fourth were mostly Samoan. They were in low intermediate to advanced-level English classes. At the beginning of the study, their scores on a generalized test of the first three thousand most common words in English, their scores ranged from 23 to 95 out of a possible 102. Only four students scored over the minimum level (86) that indicated mastery.

The students all took a pre-test at the beginning of the study. The pre-test consisted of an interview in which the students reported their knowledge of words associated with the topic of Òpain.Ó In the interview, the studentsÕ knowledge of several words from the first and second thousand general word lists was estimated. The students also completed a reading and recall task, and the researcher investigated the subjectsÕ language proficiency during the interview.

After the pre-tests, the learners in the study were randomly assigned to one of three treatment groups (There were 16 students per group. The groups had an even distribution of students who had scored in the high and low vocabulary range on the pre-test). One group received explicit instruction on integrating new and known information, and then read a text on pain control and retold it without referring to the original text.

The second group did not receive strategy instruction. They read the same text as the first group and retold it. They were allowed to refer to the text while retelling the story.

The final group did not receive reading instruction and did not do the reading and retelling activity.

At the end of the study, the participants completed post-tests consisting of individual interviews and two multiple choice tests that targeted ÒpainÓ related words from the reading.

Both the experimental and comparison groups performed significantly better than the control group and did not differ significantly from each other. This indicates that participation in read and retell tasks stimulated learners to process written information in a way that facilitates vocabulary acquisition.

Students whose pre-test measures indicated that they had high background knowledge on the topic gained substantially more vocabulary than those who had lower levels of background knowledge.

A possible reason why there are no significant effects for the explicit strategy instruction is that the design of the study is confounded. The treatment for the two groups should have been the same, except for the variable of strategy instruction. By adding the extra variable of allowing the second group access to the story during the retelling and not allowing that to the first group, the researcher obscured the issue of the effect of strategy instruction. However, the study does show that engaging learners in reading and retelling tasks does help them increase their vocabulary. Also, as the Carrell,1991 study in this review suggested, learnersÕ abilities to profit from certain kinds of instruction may be dependent on their proficiency levels. By mixing high and low English proficient students, these possible effects were lost.

Literacy tasks that promote learner engagement, such as reading and retelling tasks, can promote incidental vocabulary learning. Additionally, the learnerÕs prior knowledge about a topic is a predicting factor in learner vocabulary acquisition. If incidental vocabulary acquisition is the goal, pedagogic tasks should be structured around readings for which the learners have a high degree of background knowledge.

The study could be replicated separating the effects of explicit language instruction and access to the text during the recall on incidental vocabulary acquisition.

Johnson, K.E. (1992). The relationship between teachersÕ beliefs and practices during literacy instruction for non-native speakers of English. Journal of Reading Behavior, 14, 83-108.

ESL literacy teachersÕ beliefs about teaching literacy to adult ESL learners and how these beliefs shaped their instructional styles. The researcher argues that teacher beliefs may strongly influence their decision-making processes. This article seeks to clarify the relationship between literacy beliefs and literacy teaching and thereby emphasize the importance of teacher development.

The participants in this study were 30 ESL teachers from urban and suburban areas of the northeastern United States. Some were associated with K-12 education in the school districts; 8 taught in continuing education programs for adults or in programs for refugees.

The researcher first defined three separate theories of language learning common in the second language teaching literature. These were the empiricist explanation (language learning through habit formation, in this study skills-based), the rationalist explanation (language learning through rules), and the communicative explanation (language learning through meaningful communication, in this study function-based).

In order to understand the teachersÕ beliefs about ESL teaching, the teachers first recorded an explanation of what they considered the ideal ESL instructional setting. Their characterizations of an ideal teaching situation were coded as skill-based (dialogs and drills), rule-based (explicit language rules) or function-based (realistic materials and interaction). They were then given three separate lesson plansÑone skills-based, one rule-based, and one function-basedÑdesigned to teach the same grammatical principle. They were asked to select the lesson plan that best reflected their own instructional beliefs or style. Finally, they were given a beliefs inventory, a set of fifteen statements about language teaching. The inventory contained five statements reflecting each of the three theories of language learning and teaching. The teachers were asked to select the five statements that best reflected their own beliefs.

Three teachers (one from each of the theoretical orientations toward teaching) were selected to be observed teaching in a literacy setting. The transcripts of these observations were divided into instructional episodes that were coded as skills-based, rule-based, or function-based.

Teachers fell into one of three categories in terms of their beliefs: those with a dominant theoretical view, those with a dual theoretical view, and those with mixed theoretical views. Almost sixty percent of the participants had a dominant theoretical view; in all, almost fifty percent had a dominant belief in function-based instruction. Twenty percent of the participants had a dual theoretical view that incorporated a function-based belief system with another of the methodologies.

The teachers that were observed overwhelmingly followed their belief systems in their instructional style. Almost seventy percent of the instructional episodes in the skills-based teachersÕ class were skills based. For the teacher that professed a belief in rule-based instruction, over fifty percent of the instructional episodes in his class were rule-based. The teacher with a function-based theoretical orientation used function-based instruction in over ninety percent of the instructional episodes identified.

The study impressively establishes a link between teacher beliefs and teacher practices in the ESL literacy classroom. As the study focused on instruction in ESL literacy classrooms, the relationship between belief and practice might have been more clearly defined had the belief measures also focused on ESL literacy instruction beliefs, rather than on general ESL instruction beliefs.

TeachersÕ beliefs seem to strongly impact their practices. Teachers should examine their own underlying beliefs in order to better understand their own literacy teaching practices.

A more in-depth qualitative analysis of both teacher beliefs about literacy and instructional practices in literacy classrooms could be done. This might help illuminate the ways in which beliefs influence decision-making. Teachers could also be asked to reflect on the sources of their teaching beliefs, so that the process of forming beliefs could be better understood.

Jones, M.L. (1996). Phonics in ESL literacy instruction: Functional or not? Philadelphia, PA: World Conference on Literacy. (ERIC Document Reproduction Service No. ED 436 104)

The effectiveness of integrating phonics instruction into adult ESL literacy instruction. While phonics has been extensively debated in first language reading research, it has not been often discussed in the second language reading literature. However, many adult ESL learners have limited literacy in their first languages and/or have literacy in non-Roman alphabets or in non-alphabetic languages. These learners need to acquire decoding skills as efficiently as possible.

The learners in the study were adult ESL students who participated in an open-entry community education class that met twice weekly over 18 months. Because of the nature of the class, many learners started and stopped attending the class during this time. Qualitative data was collected as the study progressed from the students participating at any given time. Most of the students were Spanish speakers from Mexico. Many had limited educational backgrounds and were semi-literate in Spanish.

The teacher of the course was a trained ESL teacher with a background in phonics instruction. The teacher believed that introducing adult ESL students to phonics would more effectively help them to increase their English reading and writing skills. This teacher also created the materials used to present phonics to the students.

The researcher began the study by presenting the students with some basic phonics rules, similar to those presented to young children learning to read English. The teacher referred to these rules throughout the course, drawing student attention to the spelling of the words used in content activities. The class also participated in activities designed to draw learner attention to sound-symbol correspondences, such as learning various forms of verbs or reading and writing short paragraphs. Samples of student essays were collected, along with the teachersÕ impressions of the class.

Based on an examination of the essays, the researcher concluded that the learners gained a better command of English spelling and were more willing to write in English when instructed in phonics. She observed language development in their writing as well.

Many researchers stress the importance of interactive models of reading, in which top-down predicting skills are integrated with bottom-up decoding skills. However, there is little discussion of how this is to be achieved, especially with non-literate or semi-literate students who cannot transfer their L1 decoding skills to English reading. Research into decoding instruction is valuable for the insights it gives teachers about pedagogical techniques for helping students develop literacy.

While the researcher suggests that phonics instruction improves reading ability, no direct evidence was collected on studentsÕ reading proficiency. The effects of phonics instruction on reading have to be inferred from studentsÕ writing ability. Also, since there was no comparison control group, it is not possible to say that the improvements in student spelling were prompted by the phonics instruction, and not by some other instructional technique.

Students with limited literacy exposure in their native language or in the Roman alphabet could most likely benefit from some phonics exposure. It is important to remember that, while bottom-up decoding skills are not the entirety of reading, they are still essential and should not be neglected in instruction.

Researchers could investigate the effectiveness of different techniques for teaching students decoding to determine whether the phonics approach or another is most useful for adult ESL literacy instruction. Also, pre- and post-testing for reading speed and comprehension could be used to quantify the effects of phonics instruction.

Kessler, C., Cohen, B., & Walsh, R. (1996). Classroom interaction for adult literacy. Netanya, Israel: The Second International Conference on Teacher Education. (ERIC Document Reproduction Service No. ED 405743)

The effectiveness of a program designed to train teachers to use more interaction in ESL literacy classrooms in terms of changing teachersÕ attitudes toward teaching reading.

One-thousand four-hundred three ESL reading teachers from Texas community-education centers. The teachers were involved in the training as part of a statewide initiative to increase literacy and literacy levels. The program was implemented after nationwide studies indicated that nearly half of the people of the United States adult population read at the lowest two levels of literacy as designated by national standards. The teachers in this study were learning new techniques and theories about ESL teaching in order to improve their own practice. The participants ranged from professionally trained ESL teachers to teacher aides and community volunteers. How many?

Three teacher trainers from a large Texas university with expertise in TESOL and communicative pedagogy. Ideologically, the teachers aligned themselves with literacy views that emphasize the need to move beyond functional literacy, or the ability to function in society, to critical literacy, or literacy for personal and cultural empowerment. They sought in their teacher education sessions to help teachers re-evaluate the nature of the student/teacher relationship and to make the literacy curriculum relevant to the social issues that studentsÕ face.

A needs assessment was sent to each educational center, and the teacher education workshops were modified to fit the needs as expressed by teachers atÊ each center. Then the teacher educators held a professional development workshop. At the beginning of each workshop, participants were asked to rate their needs for help or new ideas on ten aspects of the critical literacy model using a Likert scale. The workshops were designed to be participatory and to involve group investigations and decision-making. To evaluate the overall effectiveness of the workshops, the researchers then asked the participants to rate the effectiveness of different aspects of the program on a Likert scale and to give comment on the most and least helpful aspects of the program. To gauge attitude shifts during the conference, the participants rated the same ten items related to critical literacy that they had examined at the beginning of the conference. This time, they were asked to indicate how useful they felt these concepts were to their own teaching.

In terms of the overall effectiveness of the teacher development workshops, the researchers found that the participants overall were very positive about the interactive nature of the workshop and were eager to apply this pedagogical model to their own teaching. Their comments on the workshops indicated increased understanding of the concept of critical literacy and the importance of a participatory approach to teaching ESL literacy to adults. In terms of the attitudinal shift, the researchers found that the participants overall rated the aspects of critical literacy much higher after participating in the workshops, which they interpret as evidence of the effectiveness of participatory workshops towards introducing teachers to literacy pedagogy.

All of the assessments in the study were completed immediately following participation in the workshops, so there is no evidence of long-term attitudinal shifts. Since all the data collected was self-report data on attitudes, there is no evidence that the workshops had any effect on the participantsÕ practice. Also, the use of statistics with their instruments is troubling. Comparing scores on one scale based on teachersÕ estimates of their need to learn about a particular concept with their estimates of the usefulness of the conference is not a fair comparison to uncover attitudinal change. A teacher could find a concept very useful in practice, for example, but feel that he already sufficiently understands the concept. The differences between the Likert scale ratings of critical literacy concepts is as likely dues to reasons like this as to any attitudinal shift.

Finally, no evidence is presented of the effectiveness of the critical literacy model as a model for ESL literacy pedagogy. While the model is theoretically and socially appealing, there is no evidence that it is effective or that it is more effective for helping ESL speakers to read than any other pedagogical model. Without such evidence, it is hard to see why changing teacher attitudes toward literacy is a necessary or even positive step.

Teacher development practices can be more effective when they model the pedagogy that is advocated. There is a need for more understanding of literacy pedagogy, and many practitioners are eager for more information about teaching literacy.

Research should be conducted on the impact of different literacy pedagogy models on learner attainment in literacy. Also, long-term investigations of the relationship between teacher understanding of pedagogy and teacher practices should be undertaken to evaluate the impact of teacher education.

Klassen, C. & Burnaby, B. (1993). ÒThose who know:Ó Views on literacy among adult immigrants in Canada. TESOL Quarterly, 27, Ê377-397.

The beliefs about L1 and L2 literacy of non-literate immigrants in an English speaking community and of their strategies to compensate for their lack of literacy in both their first language and in English.

Nine monolingual Spanish L1 adult immigrants in Toronto, Canada. Four of the nine participants had little or no literacy skills in Spanish; none were functionally literate in English.

The researchers first summarized national-level census statistics about immigrant populations in Canada, their education levels, and implications about their literacy. They concluded that national-level statistics could only supply a surface understanding of immigrant populations and their literacy needs. To supplement the census findings, they interviewed nine Spanish monolinguals who were non-literate in English about their literacy needs and beliefs.

Speakers without English literacy skills used several strategies to function in society. These strategies included establishing relationships with neighbors, ESL classmates, or officials who could help them read or write letters or fill in forms or complete commercial transactions. Those without Spanish literacy skills used these same strategies for written communication in Spanish as well. They also relied more heavily on memory, memorizing subway or bus routes because they could not decode signs, or by memorizing product logos in order recognize bills or choose food at the stores.

All of the participants felt confident in their abilities to perform their jobs without being literate, usually because they could use visual cues to do their work. However, they felt that they were ineligible for promotions at work because of their limited ability.

The participants who were also illiterate in Spanish expressed dissatisfaction with their experiences in adult ESL courses. Many were ashamed of being illiterate. They could not participate in activities that involved reading and writing in English, and felt lost in grammar lessons. Because they could not write down new words or phrases, they could not retain as much or review at home, or efficiently use Spanish-English dictionaries.

The participants universally agreed that it was more important for them to become literate in Spanish than in English, and doubted that they could improve their English proficiency without first becoming literate in Spanish. They also felt that they could not be accepted by either the Spanish-speaking immigrant community or by their own children without gaining literacy.

The focus on functional literacy (literacy skills needed on the job or in official transactions) perhaps is aimed at the areas of literacy for which learners can already compensate. Teaching personal literacy, such as letter or personal history writing, may be more relevant for adult ESL learners, especially those with low native language literacy.

The effectiveness of conventional ESL methods is constrained by learnersÕ level of literacy in their first language. Teachers should be aware of the special needs of non-literate learners.

Pedagogical approaches to help ESL learners compensate for L1 illiteracy should be identified and evaluated.

Koda, K. (1999). Development of L2 intraword othrographic sensitivity and decoding skills. Modern Language Journal, 83, 51-64.

The differences in intraword structural sensitivity of adult ESL learners from alphabetic and non-alphabetic L1 backgrounds, the differences between native and non-native speakers of English with respect to intraword sensitivity, and the extent to which intraword sensitivity affects ESL decoding ability. Intraword structural sensitivity refers to a readerÕs ability to use both phonological and morphological information in processing words in print.

The learners were forty beginning level ESL adults enrolled in an intensive English institute. Twenty participants were native speakers of Korean (a non-roman alphabetic language); twenty were native speakers of Chinese (a non-alphabetic language). All the participants had studied English in high school in the home countries and all had been in the United States for less than six months at the beginning of the study. The two groups completed the reading and listening sections of the TOEFL at the beginning of the study; there were no significant differences in their mean scores on either test.

Six native speaker university students also completed the tasks as an L1 reader control group.

An orthographic acceptability judgment task was created using lists of frequent words. The letters of the words were rearranged to create non-words. Some of the non-words were legal (possible) letter strings in English, while others were illegal (impossible) letter strings in English. For example, a word like double could be rearranged to form the legal non-word boudel or the illegal non-word ebdluo. Non-words were divided into those that began with high frequency first letters in English (letters that often appear at the beginning of a word) and those that began with low frequency first letter in English. Thus, the non-words could be high-frequency legal, high-frequency illegal, low-frequency legal, or low-frequency illegal. Students were asked to judge whether the words were possible or not in English. Their answers would indicate their understanding of English intraword phonology and morphology.

Additionally, two decoding tasks were used. The first consisted of fifty legal non-words that the students were asked to read. The second was a homophone judgment test. The students were given thirty pairs of words. The first word of each pair was a high frequency English word. The second was a possible non-word (e.g., rain and rane). The students were asked to judge whether the non-words were homophones of the high frequency word.

Students from both groups were better able to discriminate between legal and illegal letter strings when they began with high-frequency letters. This indicates that L2 readers do develop sensitivity to the positional frequency of letters. The Korean speakers were better able to discriminate between legal and illegal sequences when the words began with low-frequency letters, which suggests that their L1 experience with an alphabetic language had helped them to refine their intraword structural sensitivity.

There were also differences between the native speakers of English and the ESL participants. The native speakers made similar judgments to those of the ESL speakers on the legal letter strings, but were far more consistently correct in judging illegal word strings. They made few errors in judging illegal word strings, while the ESL participants accepted around 45% of high-frequency and 40% of low-frequency illegal word sequences. This suggests that one factor distinguishing native and non-native readers is the ability to detect violations of orthographic constraints.

In terms of the affect of orthographic sensitivity on decoding, there were differences between the Chinese and Korean speakers. Whereas there was a significant correlation between intraword structural sensitivity (as measured by orthographic acceptability tests) and decoding performance (as measured by non-word and homophone tests) for the Korean speakers, there was no such correlation for the Chinese speakers. This suggests that Korean speakers utilize intraword sensitivity more than Chinese speakers in decoding. This could indicate that the Chinese speakers rely still on the L1 whole word processing strategies when reading in English, while Korean speakers rely on intraword processing strategies that they can transfer from their L1 reading experience.

The study is very well designed and controlled, and the analysis is careful and well explained. It should be remembered that this is a highly controlled laboratory experiment, and it is uncertain how the findings relate to reading in a more natural setting. Also, it is important to note that the experiment does not give any developmental information. We donÕt know if Chinese speakers eventually begin to process written English more like Korean speakers do, and we donÕt know if or when either group gains intraword sensitivity similar to that of the native speakers.Ê More information on the morphosyntactic structure of Korean would be helpful in determining the affect of this factor for Korean speakers learning to read English. The fact that spoken Chinese uses tones to transmit the semanticÊ aspects of language, whereas spoken English uses phonemes that are more or less mapped to graphemes in written English could also contribute to the difficulties Chinese speakers have with English intraword processing.

Intraword structural sensitivity affects text-processing abilities. Teachers should help beginning ESL students, especially those from non-alphabetic L1 backgrounds, to gain awareness of the phonological and morphological information embedded in English words.

Also, both L2 experience and L1 background affect studentsÕ ESL decoding abilities. Teachers should be aware of L1 orthographic differences when teaching decoding strategies.

This study could be replicated using learners at different proficiency levels, to see how the relationship between intraword structural sensitivity and decoding develops over time. This study could also be replicated keeping track of response times, to see if learners with alphabetic L1 backgrounds can make judgments more quickly than learners with non-alphabetic L1 backgrounds.

Lantigne, B. & Schwartzer, D. (1997). The progress of Rafael in English and family reading: A case study. Journal of Adolescent and Adult Literacy, 41, 36-45.

The progress of a Mexican immigrant student in a community-based adult ESL and family literacy program. While increasing individualsÕ English proficiency can have a positive impact on their lives, it is likely that adding a family literacy emphasis to adult ESL can have a positive impact on an entire family. This research sought to evaluate the impact of such a program on one individual and his family.

Rafael is a Mexican immigrant in his fifties who lives in a predominately Latino, low-income neighborhood. He and his wife are the parents of seven children, all under the age of fifteen. The researcher describes him as a caring, involved husband and father. Rafael works as a baker and attends ESL courses in a community ESL program run in a local library. He had limited formal education in his native language, but had achieved a reasonably high level of Spanish literacy. He wants to learn English to improve his familyÕs circumstances and his childrenÕs future opportunities. He is described as an inquisitive student with a real desire to learn. At the beginning of the study, he had been studying in this particular program for five months.

The primary researcher in this study is one of the teachers at the community English center. This research was completed as part of her masterÕs degree in TESOL (the secondary researcher was her academic advisor). She approached this research because of her interest in family literacy and in increasing opportunities for Latino-Americans. She introduced family literacy to the class by exposing them to interesting childrenÕs literature in the library, which they were encouraged to take home to share with their children.

The case study followed RafaelÕs progress over three months in the family literacy program. The researcher gathered information on his progress in English learning and in family literacy primarily through her observations of him in class, analysis of his writing samples, and personal interviews conducted in Spanish with a bilingual translator present. Data were triangulated with informal conversations with Rafael and his family, interviews with other teachers acquainted with Rafael, and RafaelÕs test scores on the programÕs placement test and on the Basic English Skills Test (BEST).

Rafael generally improved his English skills during the period of the case study. His writing samples progressively included fewer errors and more accurate self-corrections. Based on the interviews, it seemed that his listening comprehension skills also improved, and he gained several levels on the Basic English Skills Test. His comfort with reading in English and with reading to his children in English also improved. While at the beginning of the program, he expressed reluctance to read to his children in English, by the end of the program he and his children were engaged in reading together. During visits to his home, his children chose books for him to read to them, and continued to examine them after the reading ended. His children also accompanied him to the English classes at the library, choosing books that they wanted to take home and read together.

The study makes use of data gathered in multiple ways to give the reader a clear understanding of Rafael as a person and as an English learner. It contrasts with a fairly common stereotype of Latino immigrants as lazy or uninterested in learning English. It also gives clear examples of RafaelÕs progress in improving his oral and written English and the impact of family literacy on his family.

However, it is important, as with all case studies, to remember that the data was collected on one person only, and should be generalized with caution. Also, three months is really quite a short amount of time to gather case study data. We do not know if the progress Rafael made continued beyond the study.

Including family literacy in ESL adult literacy curricula can help increase the exposure to reading that individuals and families receive. It is possible that this could have a long-term impact on the entire family.

This study could be extended over a longer period of time, following the progress of an individual and their family in family literacy over time. Also, quantitative research could be done to determine whether these results are typical of participants in family literacy programs.

Laufer, B. (1997). The lexical plight in second language reading: Words you donÕt know, words you think you know, and words you canÕt guess. In J. Coady & T. Huckin (Eds.) Second language vocabulary acquisition (pp. 20-34). Cambridge: Cambridge University Press.

The possibility of guessing the meaning of words from context. While researchers who work in top down processing often list guessing as a strategy that learners can use to overcome a lack of vocabulary, and while extensive reading enthusiasts counsel beginning students to guess the meaning of unknown words and continue reading, Laufer points out that this strategy is both insufficient, and also often inappropriate. Basing her arguments on linguistic analysis and on previous studies of vocabulary and reading, Laufer presents reasons why bottom-up processing skills like attention to vocabulary cannot be ignored.

Vocabulary studies have indicated that there is a threshold of vocabulary for second language reading. (See, for example Alderson, 1984; Coady, 1997). In order to comprehend a text and in order for learners to be able to transfer their first language reading strategies to the second language, they must be able to comprehend 95 to 98% of the words in a text. Some researchers believe that vocabulary knowledge is the single most important predictor of reading comprehension. Top-down processing researchers have contended that vocabulary knowledge is not as important as good top-down processing skills, which include guessing word meanings from context. This study examines several conditions that make guessing inappropriate, because readers are unlikely to arrive at an accurate enough meaning of a given word to allow for comprehension.

The author divides her argument into three categories: words that are unknown to the reader (the issue of vocabulary threshold), words the reader mistakenly thinks s/he knows (false cognates and others), and words a reader cannot guess the meaning of (the insufficiency of context).

When too many words in a passage are unknown, it is impossible for the reader to apply L1 reading strategies. For example, a reader cannot apply a Òfind the main ideaÓ strategy if he does not understand the ideas in the text at all. Being able to apply such strategies is contingent on second language vocabulary knowledge. Additionally, human cognitive processing limitations restrict understanding of reading to texts in which most of the words are automatically understood. In other words, if a reader has to expend mental energy to understand many of the words, he will not have the cognitive capacity to pay attention to the overall meaning as well. This explains why learners who are dealing with difficult texts may be able to comprehend sentences without ever understanding the connection among sentences. In this case, a lack of vocabulary knowledge short-circuits reading comprehension.

Likewise, reading comprehension can short-circuit when learners think they know words that they really do not know. Examples of this sort of short circuit occur when students do not recognize false cognates or interpret idioms literally. They can also occur when students encounter words with deceptively transparent meaning or morphological structure. For example, a student might analyze shortcoming by analyzing the component words and arrive at a meaning like a short visit. Short-circuits may also occur when students do not recognize multiple meanings of a word, or when they confuse the meanings of words that are phonologically or orthographically similar. When students do not recognize that they do now know a word, they can invoke the wrong schemaÑthus, bottom-up processing skills can thwart the benefits of top-down processing.

Lastly, there are many conditions in which word meanings cannot be guessed. Laufer contends that words rarely occur in a sufficiently redundant context for meaning to be inferred from context. Also, if the student doesnÕt know the words that provide contextual clues, he is unable to use the context for guessing word meaning. Guessing also can only work if the student is able to predict the meaning of the passageÑif the student and the author have similar schema for the concept being described. If the authorÕs schema is unexpected by readers (possibly because of cultural differences), they are more likely to guess word meanings that fit with their own, rather than the authorÕs meaning.

LauferÕs typology is very useful for understanding why second language learners, especially at the beginning levels of literacy, are unable to apply their L1 reading strategies or the strategies their teachers teach. It is a clear argument against the complete reliance on extensive reading to help learners increase their vocabulary, and an reiteration of the importance of directing literacy learning, especially at the lower levels of proficiency, towards bottom-up processing skills like vocabulary learning and recognition.

While extensive reading is theoretically appealing, it may not be applicable especially for beginning students. Vocabulary development to the threshold level -- usually estimated at 3,000 words (Coady, 1997; Grabe & Stoller, 2002) -- is essential for students to be able to read in the second language.

Researchers should examine the effectiveness of different vocabulary instructional methods and the possibility of incorporating these into adult ESL literacy instruction.

Learning Disabilities Association. (1994). Learning disabilities and the acquisition of English language skills in the adult ESL population: A demonstration project. Minneapolis, MN: Author.Ê (ERIC Document Reproduction Service No. ED 382 052)

The process of determining whether or not ESL adult literacy students have learning disabilities and the effectiveness of a pull-out tutoring program for students with learning disabilities. For this research, a learning disability was defined as a life-long condition with presumed neurological origin that interferes in some way with development or expression of verbal abilities. There is little understanding of the effects of learning disabilities on second language development or on possible treatments for students with learning disabilities. Because both low English language proficiency and learning disabilities are reflected in verbal skills, it is often difficult to determine whether or not a student has a learning disability.

The learners were adult ESL students from various ethnolinguistic backgrounds studying at a private English center in the United States. The majority of the students were middle-aged or older, and many of them had been forced to curtail their formal education in their native languages before coming to the United States. Two classes with a total of fifty students were chosen for the study. All of the students were assessed for possible learning disabilities, and thirteen were eventually selected for tutoring.

The classroom teachers were experienced ESL teachers who had been teaching in the English center for several years. The teachers were assisted by learning disabilities specialists from the Learning Disabilities Association, who helped design and implement the assessment measures and tutor the students presumed to have learning disabilities. The majority of the students spoke either Russian or Vietnamese, so translators from these languages were also brought in to assist assessment. A Hmong education assistant also participated as a translator for Hmong, Thai, and Laotian.

The classes were tested at the beginning and end of the project. All the students completed a basic English skills test and a phonics inventory. They were also observed by their teachers and by a learning disabilities specialist. One of the classes also submitted a native-language writing sample and completed a learning styles inventory and an intelligence test. These were used to determine which students possibly had learning disabilities.

The students identified as having learning disabilities were pulled out of their classes twice a week for tutoring. The tutoring was similar to that commonly used with native speakers of English with reading disabilities such as dyslexia. It included extensive phonics training designed to improve studentsÕ phonemic awareness and understanding of sound/symbol correspondences in English. The tutor used kinesthetic, auditory, and visual techniques to reinforce the teaching. The tutoring was sequential, and mastery was required to move to a new level.

At the end of the program, qualitative evaluations of the program were collected from the learning disabilities tutor, the teachers, and the students who had participated in the tutoring.

Some of the tests used in this study were deemed inappropriate by the English center staff for determining whether or not limited English proficient adults had learning disabilities. Tests that are normed with reference only to native speakers may not be able to discriminate between language difficulties and learning disabilities. One of the best measures for determining whether a student did or did not have a learning disability was the native-language writing sample and the educational histories given through interpreters. It was learned that some of the students, for example, had never had formal schooling. Their difficulties in the classroom were therefore more likely caused by difficulty adjusting to the academic structure than to any neurological disorder. Additional problems with identifying students with learning disabilities were prompted by cultural issues. Many students thought that having a learning disability was shameful and were reluctant to participate in tutoring. Some of the test results were also compromised because students shared answers with their friends to prevent their friends from failing and being stigmatized. The researchers found that the nature of learning disabilities and the system of testing needed to be explained carefully and sensitively in the studentsÕ native languages.

The teacher and tutor evaluations of the program made it clear that working together had allowed all involved to benefit from each othersÕ expertise. The teachers learned many techniques that they had not been aware of for teaching phonemic awareness, which were especially helpful for working with the semi-literate students in the program. The tutors learned to embed language instruction in cultural instructionÑthe norm in ESL but not in other fields of language development. It also became obvious that commercially developed resources for working with students with learning disabilities were often inappropriate for the ESL population. Lessons that used more familiar vocabulary or that built on common ESL themes would be more useful.

The teachers did indicate that the students who had received tutoring made noticeable improvements in reading and writing, and post-tests indicated that tutored students had increased their reading proficiency during the program.

The study is a good first step in understanding how to identify and help adult ESL literacy students with learning disabilities. A better description of the learners who were presumed to have learning disabilities, as well as the criteria used to separate language deficiencies and learning disabilities would have been helpful to those researching and working in this field.

ESL teachers and learning disabilities specialists need to work closely together to integrate classroom language instruction and learning disabilities tutoring. Teachers and tutors need to be sensitive to cultural issues surrounding the identification and treatment of learning disabilities.

ESL specific tests for identifying learning disabilities should be developed, as well as materials designed specifically for working with ESL students who have learning disabilities.

McLeod, B. & McLaughlin, B. (1986). Restructuring or automaticity? Reading in a second language. Language Learning, 36, 109-123.

The automaticity of decoding skills of ESL learners at different levels. The researchers wanted to know if more proficient readers would decode more automatically, and therefore be able to apply meaning-based strategies more than less proficient readers.

The participants in this study were twenty native speakers of English who were university students and forty-four ESL students enrolled in an intensive English program IEP). The ESL students were mostly Japanese and had all received secondary school and some college education in their native language.

The ESL participants were divided into two groups, beginning and advanced, based on their placement tests into the IEP. All of the participants were given two tasks: an oral reading and an oral cloze task. On the oral reading, each participant read two 225-word passages out loud. The native speakers were asked to count the number of lines in the passages as they read in order to increase the memory burden on them and thus to increase the number of errors they would produce. Errors were defined as anything that deviated from the written passage. Errors were classified as meaningful (errors that persevered the original meaning of the passage and conformed to English syntax) and unmeaningful (errors that changed the meaning of the passage or violated English syntax). Each of these categories was further divided into insertions, deletions, word-order changes, and substitutions, and all errors were coded as one of these.

For the cloze test, the participants were given the first two or three lines of a new passage and asked to read them out loud and then fill in the next word. The student was then given the correct response and the process was repeated for the following two or three lines. There were ten missing words; participant responses were coded as plausible or implausible based on the syntax and the meaning of the passage.

On the cloze test, the authorÕs prediction that more advanced readers would be better able to apply top down predicting strategies was affirmed. The native speakers performed significantly better on the cloze test than the advanced non-native speakers, who significantly outperformed the beginning non-native speakers.

However, on the oral reading miscue analysis a different pattern was found. The authors had predicted that more advanced readers would make a higher percentage of meaningful errors. In this, the native speakers had a significantly higher percentage of meaningful errors than did the non-native speakers. However, there was no significant difference between the beginning and advanced ESL students in the proportion of meaningful errors to total errors. While the advanced ESL students made fewer errors overall, they made the same percentage of unmeaningful errors. This is not the expected pattern if advanced speakers are making decoding more automatic, and therefore able to expend more resources toward interpretation.

The researchers conclude that the advanced level students were not restructuring the reading process to make second language reading more similar to first language reading. Like the beginning students, they were still focused on perfect phonological decoding. While their higher proficiency made this more possible, and they therefore made fewer overall errors, they were still not reading the way native speakers do. Native speakers make use of contextual clues and apply top-down processing. The researchers suggest that the advanced ESL readers need to restructure the reading process, taking their emphasis away from perfect decoding and increasing their application of top-down processing strategies.

The study presents interesting results and analysis of the differences between native and non-native speakers of English in their reading proficiency and behavior. It is an interesting application of reading models to actual data, and an interesting use of data to challenge and shape existing cognitive models of reading.Ê

For purposes of clarity, it would have been better for the researchers to give more information about the proficiency level of the students involved, especially to give the reader more understanding of the gap in proficiency between the beginning and advanced ESL learners. If the gap is relatively small, for example, one could simply claim that the advanced group had improved, but not yet automatized the decoding process.

Even more advanced ESL students may not be able to transfer meaning-based first language reading strategies into the second language. Teachers may need to encourage good decoders to incorporate predicting strategies into their reading process. Direct instruction of predicting strategies may be necessary.

This research should be replicated among readers of different cultural groups, since strategy use has been linked to culture. Also, longitudinal research could investigate when the restructuring is likely to occur.

Mikulecky, L. (1992). Workplace literacy programs: Variations of approach and limits of impact. San Antonio, TX: National Reading Conference. (ERIC Document Reproduction Service No. ED 353461)

The impact of different workplace literacy programs on learnersÕ literacy beliefs and practices, on learnerÕs families, and on learnersÕ productivity on the job. Some of the literacy program sites included an ESL component, which will be discussed here.

The learners were ESL adult workers at various job sites in the United States. The jobs sites, all blue-collar industries such as automotive or manufacturing plants, had instituted on-site literacy programs for both ESL and native speakers in order to increase the functional literacy of their employees. Employees could choose to take part in the programs. Many did so with the hope that increasing their literacy in English would improve their chances of receiving promotions into positions that demanded greater literacy skills.

In each workplace, a combination of group interviews and questionnaires were used to assess the impact of the program on the learners. A structured interview based on Lytle (1990) was used to investigate the impact of the literacy programs on learnersÕ beliefs, practices, processes, and plans. Job performance ratings and employer data on attendance and productivity were also considered.

After participation in workplace programs, ESL learners made qualitatively measured improvements on the quality of their reading and writing on the job. They also reported reading and writing more and with more confidence, as compared to ESL speakers who did not participate in the programs. They were observed to be more willing to ask questions at work as well, indicating that increasing their literacy skills increased their confidence in their spoken English as well. When questioned about their reading strategies, they reported using more sophisticated reading strategies than workers who did not participate in the program, and performed as well on a written test about job-related scenarios as did native English speakers enrolled in workplace GED classes. After participating in the classes, ESL workers were more likely to define themselves as literate, and were more likely than non-participants to include reading and continuing education as part of their future goals.

The study gathered an exhaustive amount of information about workplace literacy programs and showed positive results for workplace instruction for both native speakers and for ESL participants. It is not surprising that workplace literacy programs would be successfulÑon-site classes result in lower absenteeism and closely connecting language instruction with the workplace helps ensure that the instruction is immediately relevant to the studentsÕ lives. While the highly functional nature of the literacy courses described here makes it unlikely that workplace instruction will lead to higher levels of literacy, they seem to be a good first step in the literacy process for ESL speakers. We need more information including comparisons with those English language who didnÕt take the classes. We also need to know the English proficiency and literacy levels of the workers.

When possible, literacy instruction for ESL adults should have applications to their everyday lives. In this case, directing the instruction to the participantsÕ careers led to noticeable gains in literacy in limited time. Increasing participantsÕ levels of literacy and comfort in reading English, even in a limited frame like workplace literacy, may increase their dedication to their personal education.

This sort of research should be piloted in other industries, and among non-workplace related groups. For example, church-based ESL literacy programs might have similar success in introducing adult populations to reading. Future research should include pre- and post intervention assessment data on the learners L2 proficiency and literacy levels.

Packard, B. W-L. (2001). When your mother asks for another book: Fostering intergenerational exchange of culturally relevant books. The Journal of Adolescent and Adult Literacy, 44 (7), 626-633.

An adult childÕs experience in sharing literacy materials with her mother and fostering her motherÕs English reading ability.

The learner involved in this study was a Chinese immigrant in her late forties. She had lived in the United States for 30 years and was functionally literate in English. While she was able to read English newspapers, she did not consider herself a reader in English and did not engage in pleasure reading in English.

The ÒteacherÓ in this study was the learnerÕs adult daughter, a graduate student in educational psychology, a lifelong lover of reading who speaks limited Chinese. As a second generation Asian-American woman, she felt isolated from her Chinese heritage, and was drawn to books with Asian-American themes.

The practice of reading together began when the teacher gave her mother Amy TanÕs The Hundred Secret Senses to read. This novel is about a woman and her relationship with her Chinese half-sister. Over time, Packard kept a journal of their interactions surrounding this and the other books they read. Most of the conversations were held over the phone. This book, and the other books read, The Kitchen GodÕs Wife, Eat a Bowl of Tea, and Wild Swans prompted discussion on the characters in the books, on Chinese culture and history, and on the authorÕs motherÕs family history.Ê

After Packard read a novel or memoir about the Chinese-American experience she would suggest to her mother that she read the book. Then they would discuss the books through phone conversations. They brainstormed other possible endings to the books. Increasingly, PackardÕs mother took the opportunity during these book discussions to Òweave inÓ stories of her own life and that of PackardÕs grandmother. Packard has now compiled a list of 11 books by Asian Americans, recommended by friends, that Packard and her mother plan to read and discuss.

By self-report, the motherÕs reading improved as did her confidence in her reading. She no longer read in a choppy way but rather read in a smooth line-by-line fashion. She also said she read faster. Chinese characters are read one at a time but in English it is faster to read the whole sentence together. She felt that she could skip words in English that she did not know and Òstill get the meaning of the whole story.

Both the mother and the daughter felt that they were able to participate as learners and teachers in this exchange as the mother had knowledge of Chinese and family history and culture, and the daughter had knowledge of the English language.

The motherÕs English literacy was not pre- or post-tested to see if her vocabulary and fluency had improved and could be transferred to other reading genres. The mother was not formally interviewed before and after the study to document how her attitudes and practices around reading in English had changed. All gains were untested and self-reported.

Implications for Practice

One could structure a family literacy program that featured shared reading (and writing) experiences between adolescent children and their immigrant parents using culturally relevant texts.

Key Words

Family literacy
Pleasure reading

Areas for Further Research

A family literacy program like this that taps into the cultural heritage of the participants could be studied using pre- and post-testing to determine how literacy-sharing in the family affects language learning.

Pakenham, K.J. (1983). Developing expectations for text in adult beginning ESL readers. Toronto, Canada: On TESOL Õ83. (ERIC Document Reproduction Service No. 275148)

The effectiveness of building ESL studentsÕ semantic and grammatical predicting abilities for increasing their reading comprehension. Most reading textbooks prepare students for reading through the use of meaning-based introductory questions, and then follow the reading with comprehension questions. The researcher felt that reading comprehension could be facilitated by introducing the students to the rhetorical, grammatical, and semantic structures of the text before reading, to help them hone their predicting skills before beginning the reading.

The learners were beginning-level ESL students at an intensive English center in the United States. The students were participants in a comprehensive reading course. The number of students is unspecified, as are their ethnolinguisitic backgrounds.

The researcher created materials that help introduce students to semantic and syntactic forms before they read. Termed Òexpectation exercises,Ó these activities involved first explaining the target vocabulary items, syntactic structures, or rhetorical structures. Following the instruction, students worked individually on multiple-choice sentence completion questions. The point of the sentence completion exercises was to encourage the students to begin to predict the linguistic content of the reading based on their knowledge of the language. They were meant to serve as a bridge between language study and reading study. The article explains the expectation materials and their use in the classroom. The findings are based on the teacher researchersÕ impressions of the success of the activities.

The researcher indicated that the materials were generally successful in the classroom. Although no objective testing was included in the study, he concluded that these exercises successfully encouraged students to apply both their background knowledge and their linguistic knowledge to comprehend texts.

The idea behind this study is interesting and the researcherÕs observations about the limitations of traditional textbook formats is certainly relevant. Many studies have indicated that previewing content can help improve learner reading comprehension; it is very likely that previewing linguistic forms and rhetorical structures could have a similar effect.

However, it is difficult to see how learners could apply these skills to texts that are not accompanied by this sort of previewing materials. While they might perform better on the classroom task, it is not possible to determine based on this research whether this would transfer to actual gains in reading proficiency.

Previewing for reading activities need not be limited to previewing for content, but could also include activities that help students anticipate the linguistic structures in the texts they read. It is possible that this could lead to improved reading comprehension.

These materials should be systematically evaluated in an experimental design in which the learners are monitored over time to determine what effect this intervention has on their reading comprehension of the texts that are previewed and on their reading proficiency in general. Such research should include a control group for comparison.

Petrimoulx, J. (1988). Sustained silent reading in an ESL class: A study. Chicago, IL: TESOL. (ERIC Document Reproduction No. 301 068).

The effects of Sustained Silent Reading (SSR) on ESL learnersÕ vocabulary development and reading comprehension. Most of the literature on SSR has focused on first language reading and has generally found SSR to be beneficial in improving reading comprehension, vocabulary, writing skills, and attitudes towards reading. The purpose of this study was to determine whether these results applied to second language reading as well.

Sixteen students from ten different countries (native languages unspecified) participated in this study. All of the students were enrolled in an intensive English program, and had TOEFL scores ranging from 475 to 525. This put them at a solid intermediate level. (600+ is generally considered passing out of need for academic English language preparation, 540 is acceptable for visas for health care workers.

At the beginning of the study, all the participants completed an in-house reading comprehension test consisting of a one and one-half page reading followed by multiple-choice and true/false questions on the structure and content of the reading. The participants also completed the vocabulary section of the Comprehensive English Language Test (CELT).

The participants were divided into three groups: one experimental group and two control groups. The SSR consisted of reading from self-selected materials and then participating in a class discussion of vocabulary from their individual readings. The class examined about five words and submitted them to a word ÒbankÓ for the class. At the end of each week, the experimental group received a list of the weekÕs vocabulary words along with their dictionary definitions and a sample sentence for each item.

At the end of the15- week semester, all the students retook the original tests. Additionally, a close-ended survey was administered to the experimental group students. The study investigated the studentsÕ reactions to the SSR.

While the experimental group did improve both their reading comprehension scores and their vocabulary scores more than the two control groups, the differences between their performances were not significant. The questionnaire revealed, however, that at the end of the study the students who participated in SSR generally considered the experience very positive, indicating that they found it enjoyable and felt that they had improved their language proficiency from the reading experience.

While the study did not result in significant findings in favor of SSR in the classroom, it does give indications that it might be a beneficial experience for language learners in terms of vocabulary growth and increased reading comprehension. However, the size of the experimental group (five students) would make it impossible to extent the findings even if they had been significant. This effect would have been somewhat mitigated had two, rather than three groups been formed. Since there was no treatment for both control groups, it is unclear why two control groups were needed.

The researcher attributes the lack of significant results to the brevity of the study. One could also argue that the total treatment time was too restricted for significant differences to emerge (ten minutes of SSR a day for fifteen weeks is equivalent to twelve and one-half hours only of SSR). The study would have been perhaps stronger had more students been involved in SSR for a longer period of time each day over a longer study. While the findings indicate that SSR might be beneficial for the language development of ESL students, further research is needed.Ê It is also important to remember that the discussions and class work surrounding the word bank, may also have contributed to the language development.

The survey data showed overall very high satisfaction with the use of SSR in the classroom. (Again, the satisfaction also may be with the word bank and class discussion of vocabulary.)Ê However, with such a small number of students in the experimental group, it is unclear why open-ended questionnaires or interviews were not used to gather more qualitative data on the studentsÕ view of the benefits that SSR had afforded them.

Engaging students in SSR might help them to increase their vocabulary and beneficially impact their language acquisition. The students in this study also enjoyed the SSR experience: it could possibly promote positive reading attitudes among ESL students. But, given the limited class time allotted to reading, it may not be the best use of precious class time. Maybe reading at home and keeping learner logs would be a better way to access the benefits of SSR.

This study should be replicated using a larger group of students and carrying out the experiment over a longer period of time. The experiment could also include the collection of more in-depth qualitative data.

Qian, D.D. (1999). Assessing the roles of depth and breadth of vocabulary knowledge in reading comprehension. The Canadian Modern Language Journal, 56, 262-305.

The relationships between breadth of vocabulary knowledge (the size of a studentsÕ vocabulary), depth of vocabulary knowledge (how well a student knows a word) and reading comprehension. Most studies of vocabulary learning have focused on the relationship between the approximate size of a studenÕs vocabulary and reading comprehension. However, knowing a word is more than knowing its meaning or gloss in the studentsÕ native language. Depth of vocabulary refers to knowledge of pronunciation and spelling, morphological properties, syntactic properties, connotations, polysemy (multiple meanings for the same word), register (formal, informal), frequency, etc. This study attempted to determine what roles breadth and depth of vocabulary knowledge played in reading comprehension.

Seventy-four students from intensive English centers in Canada participated in this study. All students included had a vocabulary of at least 3,000 word-families (commonly considered the minimum for reading ability in a second language). Only students who spoke Chinese or Korean as their first language were included in order to minimize cognate affects, and only students who had completed high school were included, so all students had similar levels of L1 literacy.

The students were administered four tests. First, they took a test of their vocabulary size to determine the breadth of their vocabulary knowledge. This test was adapted from the Vocabulary Levels Test, which has been used extensively in vocabulary research. The second test was a reading comprehension test based on a shortened reading section of the TOEFL, consisting of four passages and twenty multiple-choice comprehension questions. This test was given to establish reading comprehension levels for the students, in order to determine whether there was any relationship between reading comprehension and vocabulary breadth and depth. The third test, for examining vocabulary depth, was a modified version of the Word Associates Format. On each item of this test, a word from the second and third thousand word lists is given as the stimulus. For each word, students are asked to select synonyms from a list of four words, and to select words that appear with the stimulus in collocations from a separate list of four words. For both synonyms and collocations there can be one to three correct answers in the list to minimize the effects of guessing. Finally, the students completed a test of morphological knowledge. The test consisted of ten stimulus words, each of which had either a prefix or a suffix. Students were asked to define the parts of the word, and to explain if and how the affix changed the part of speech of the word.

High, positive correlations were found for scores on all of the tests. Extremely high correlations were found between the depth of vocabulary measure and reading comprehension and between depth and breadth of vocabulary. Multiple regression analysis indicated that depth of vocabulary knowledge accounted for an additional 11% of reading comprehension scores beyond that which was accounted for by breadth of vocabulary. The findings demonstrate both that the depth of vocabulary is related to the breadth of vocabulary, and that both of these areas of vocabulary knowledge have an extensive impact on reading comprehension.

The study clearly explains the importance of vocabulary learning to reading in second language and of considering both the depth and breadth of vocabulary knowledge when examining the impact of vocabulary knowledge on second language learning. However, since all the learners have vocabularies above the 3,000 word-family level, it does not give evidence of the importance of these areas of vocabulary knowledge for lower-level learners, or if the relationship is the same. Also, there was not enough variation among learners in terms of proficiency on any of the measures to draw developmental conclusions from the data. We do not know how these kinds of vocabulary knowledge are acquired in the process of learning a second language.

The measure for vocabulary depth is also somewhat problematic. Since half of the test was based on student understanding of word meanings, the material tested overlapped with the material tested on the vocabulary size test. As such, it is not surprising that the two correlate highly. Since the research considers and explains many different aspects of depth of vocabulary knowledge, it is perplexing that more were not included in the measure of vocabulary depth.

Both the size of studentsÕ vocabularies and the depth of their vocabulary knowledge may affect reading comprehension. Teachers should help students become aware of different aspects of vocabulary knowledge. This could include pointing out polysemy or morphological aspects of the words, or explaining appropriate registers or frequency of use for new words.

The other areas of depth of vocabulary knowledge could be assessed to determine how knowledge of pragmatics, frequency, pronunciation, and discourse features influences reading comprehension.

Robson, B. (1982). Hmong literacy formal education, and their effects on performance in an ESL class. In B. T. Downing & D. P. Olney, (Eds.), The Hmong in the west: Observations and reports (pp. 201- 225). Minneapolis, MN: University of Minnesota.

The effects of Hmong literacy in a Roman alphabet on student performance in a three-month English as a second language and cultural orientation programs (ESL/CO) at a refugee camp in Thailand.

The learners involved in this study were adult Hmong English language learners at the Ban Vanai refugee camp in Loei Province, Thailand. Prior to the ESL/CO course they were judged to have no English proficiency. Hmong literacy was tested with the Roman alphabet, the alphabet used by Hmong refugees at Ban Vanai and in most western countries as well. The other languages the refugees might be literate in ø Thai or Lao ø do not use the Roman alphabet.

The goal of this study was to determine the effects of Hmong literacy in a Roman alphabet on student performance in a three-month English as a second language (ESL) and cultural orientation (CO) program.

Students registered for the ESL/CO program by filling out an information card and by taking a very simple placement test to judge their knowledge of English and their ability to write in Lao, Hmong, Thai, or English.

Additional information was collected on students registering for this study. They were asked information about prior education and whether or not they could read Lao or Hmong. They were also asked to read short Hmong passages aloud and the reading was rated by interviewers. All the interviewing was done in Hmong, by native speakers who had been trained to collect personal data and to give and score the reading tests. One of the interviewers was also a Hmong reading teacher at the camp. His expertise and status facilitated the collecting of more accurate personal data than Hmong usually gave to outsiders.

One-hundred and fourteen subjects were chosen, using only those who had learned no English prior to the ESL/CO program they were going to enroll in. For purposes of the study, differences in English acquisition of the following four groups would be looked at:

The students were then divided into classes according to their oral English proficiency and received ESL/CO instruction from American teachers who had varying amounts of experience and formal training for 12-weeks.

1. The English Comprehension Test ø a multiple-choice test to assess listening comprehension of non-literate or newly literate ESL students. The test was the ÒAnn and BenÓ test, a test developed by the Oregon Indo-Chinese Refugee Program. In order to eliminate cross-cultural problems in interpreting visuals, the test utilizes line drawings, which are taught to the students in their native language first, as a part of the test. Then an English sentence is read out loud by the test giver and the student marks the one picture of three that best corresponds to the sentence.

2.The English Reading Test ø an adaptation of the Ann and Ben test. This test was given immediately after the comprehension test and differed from that test in that the student had to read the sentences rather than listen to them. The same English content was tested in the comprehension and reading tests, and the same visuals were used. However, the actual test items differed from the comprehension to the reading test.

3. The English Production Test øÊ a variation of the John Test, a standardized oral production test widely used in refugee programs in the U.S. at the time of this study. ÊThis test consisted of a pre-test in which students were asked questions they had been taught in class and the John Test ø a standardized oral production test widely used in refugee programs in the United States in the early 1980s.

4. The Hmong Reading Test. In this test, participants were asked to individually read out loud a passage written in Hmong, using the Roman alphabet. They were scored on how fluently they could decode the written material.

5. The Hmong Writing Test. This test had three parts: the first required the student to write his name in English and Hmong spelling, his refugee number, and his address in camp. The second part was a dictation of letters and numbers in Hmong. The third part was dictated sentences in Hmong.

Of the original 114 participants,those remaining to be post-tested were as follows:

Robson examined the effect of Hmong literacy education on each of the English test separately and on the overall English score (which was the average of the three tests) by means of a two-way analysis of variance. She found that Hmong literacy and formal education both significantly increased scores on the English tests. The scores of those in group A ø participants with neither Hmong literacy nor previous educational experience ø were significantly lower than those in groups B, C, and D in all three tests: listening comprehension, reading, and production. In addition, the scores of those in group D were significantly higher than those in group C in ÒproductionÓ or speaking. On a standard ESL scale the subjectsÕ English language proficiency at the end of the course was designated to range from still almost zero (for group A) to high beginning or low intermediate (for the others).

1. The ability to read, in either Hmong or English, appeared to have a major impact on the acquisition of English. Lack of formal education did not appear to have as great an effect as lack of literacy. In addition, the learners who had no literacy in Hmong and no formal education learned very little English.

Ê2. A second finding is that the subjects who could read Hmong but had not been to school did not score significantly lower than those who could read Hmong and who had been to school. Literacy in Hmong appeared to help them to learn English as much as previous educational experience did.

3. A third finding is among those who had formal education, those who could read in Hmong scored significantly better than those who could not.

Of special interest is that fact that some of the subjects could read Hmong, having learned it at the refugee camp, but had never had formal education ø which for this study meant school in Laos. Similarly, some who had had formal education, could not read in Hmong, their native language, but perhaps had some literacy in Lao, Thai, Hmong, or English. As literacy is virtually always a result of formal education, the situation with the Hmong at Ban Vanai was unique.

The researcher in this study states some limitations of the study very clearly. She stresses that it is not clear why those learners who had Hmong literacy learned more English than those who did not: whether it was the fact that the literacy was in the native language or that it was from a language using a Roman alphabet.

Robson also points out that individuals who had learned some English prior to this program were not evaluated for this study. Because the placement test given by the consortium that provided the ESL/CO instruction was not sufficiently detailed or controlled to be use as a before and after proficiency test, Robson chose to look only at those learners who had no English prior to the study. This means that individuals who might be motivated to pick up English on their own and from self-study were eliminated. The findings, then can only apply to the effects of literacy and education on attempts to learn English in a formal classroom environment.

Another limitation of the study, also discussed by Robson, is the small number of participants. Robson found statistically significant differences in the scores of the post-tests of those who had Hmong literacy prior to ESL instruction and those who did not. However, only 44 participants remained (due both to attrition and the failure of the original pre-test given by the ESL/CO program at the camp to weed out those students who knew some English before they entered the program) provide the data for the two-way and the one-way analysis of variance tests.

All that said, the finding of the study ø the boost in learning English received by those with Hmong literacy ø is exciting.Ê

Since learners who have literacy in their native language, or, at least in a language with a Roman alphabet, those learning English with this literacy will progress more rapidly than those who do not. Those without this literacy then, should be taught in separate classes where they can receive the literacy instruction. It seems that Hmong speakers would benefit from literacy instruction in Hmong prior to learning English.

Looking at non-literate learners with other native languages, it seems that literacy instruction in the native language would facilitate their learning English, provided that the native language also used the Roman alphabet.

To expand the findings of this study, it would need to be replicated with learners from language groups that do not use the Roman alphabet. It hardly needs to be said, however, that it would be hard to replicate the situation of the Hmong adult learners in a refugee camp in Thailand ø where they are not exposed to the target language in environmental print and where they have few of the constraints of time and other responsibilities adults living and working in U.S. society have. It would be hard to declare with as much authority that it is the effects of the native language literacy that improves English acquisition, rather than literacy in the Roman alphabet. When studies are done, pre and post test scores should be collected, and there should be a large number of participants taking the test.

Rumelhart, D.E. (1985). Toward an interactive model of reading. In H. Singer & R.B. Ruddell (Eds.) Theoretical models and processes of reading (pp. 722-750). Newark, DE: International Reading Association.

The nature of the reading process in an interactive processing model. Many previous researchers developed bottom-up processing models (explaining the reading process in terms of reading first letters, then words, then sentences, etc.) or top-down models (characterizing the reading process as a guessing process based on readersÕ language knowledge and background knowledge), Rumelhart sought to build a model that integrated the top-down and bottom-up processing simultaneously.

In the interactive model, guessing based on reader knowledge occurs neither before text processing nor after it. Rather, reader knowledge processing occurs simultaneously to text processing and the two interact constantly to illuminate meaning. This means that our perceptions of words depends on the syntactic and semantic environment in which they are found (the word and our syntactic/semantic knowledge are simultaneously processed), our perceptions of syntax depends on the semantic environment, and our interpretation of meaning depends on the general context of the text itself. In RumelhartÕs model, visual input from the text is brought into a processing center. At the same time, syntactic knowledge, semantic knowledge, orthographic knowledge, and lexical knowledge are brought into the processing center, which uses all this input to arrive at the most logical meaning for the passage.

RumelhartÕs model is able to explain several different reading phenomena. For example, when subjects are presented with words and asked to spell the words aloud, they are more likely to correctly spell the word when it contains only legal letter strings, and to change the order of letters in words containing illegal letter strings (for English, these could include words beginning with non-English clusters like rp or with a non-English diphthong like uo). This indicates that word-level knowledge of the language is processed along with the orthographic input from the text. For example, miscue analysis consistently shows that readers are much more likely to substitute a grammatically possible word than a grammatically impossible word, showing that syntactic knowledge was co-processed with the textual input. Other studies have found that miscues are also overwhelmingly semantically possible. Also, studies have shown that proficient readers normally only process one possible interpretation of syntactically ambiguous sentences and are most often unaware of other interpretations of the sentences because of semantic cues in the sentence. This indicates that the syntax from the text and semantic knowledge are processed simultaneously. Rumelhart gives extensive examples of other research findings that are explainable by an interactive model.

Rumelhart clearly explains the research that indicates the need for an interactive model of the reading process. While the research has focused to a large extent on top-down processing, Rumelhart explicates the importance of inclusion of both top-down and bottom-up processing. Since RumelhartÕs focus is reading in general, he does not explain how this process might be different for second language readers. For the normally developing L1 reader, the features of spoken language that undergird RÕs model are readily available-but not as readily or consistently available to L2 readers. However, his explanation of the reading process is complete, concise, and thought-provoking.

In reading instruction for native and non-native speakers, top-down and bottom-up processing need to be addressed. Reading will not be successful until the reader has both sufficient language and topic knowledge to process the text, and sufficient letter and word recognition skills to rapidly integrate these sources of input. Reading instruction should take the various interacting components of reading into consideration.

Much of the research Rumelhart cites could be replicated using ESL students to learn how second language reading differs from first language reading.

Ryan, A. & Meara, P. (1991). The case of the invisible vowels: Arabic speakers reading English words. Reading in a Foreign Language, 7(2), 531-540.

The sensitivity of Arabic learners of English to missing English vowels in words. Arabic has highly consonant-based orthographic and morphological systems. A basic Arabic morpheme is three consonants, between which various vowels can be inserted to form the members of various word families. In Arabic orthography, short vowels are not represented. In order to efficiently comprehend written Arabic, readers need to focus on consonants rather than vowels. The researchers, based on common reading errors made by low level Arabic speakers learning English, hypothesized that Arabic speakers transfer these reading strategies to English and sought to test this hypothesis.

Three groups of subjects participated in this study: ten Arabic speaking intermediate-level adult ESL students, ten non-Arabic intermediate-level adult ESL students, and ten adult native speakers of English. All were at University College of Swansea, in Wales.

The subjects all participated in a computer task in which they were presented with ten-letter frequent English words from the Thorndike and Lorge AA to 20 per million categories. 100 words in total were used in the task. Each word was presented on the computer screen for one second, then blanked out. After two seconds, the subjects were again shown the same word either spelled correctly or with one vowel removed. Forty of the words were shown spelled correctly; sixty were missing one vowel. The vowel could be deleted from the second position (the second grapheme in the correctly spelled word), the fourth position, the sixth position, or the eighth position. The subjects were asked to indicate whether the two presentations for each were identical or not by pressing a ÔyesÕ or a ÔnoÕ key on the keyboard.

The Arabic speakers were significantly less accurate than either of the other groups in recognizing that vowels had been deleted. They also performed the task significantly slower than the other groups. All the groups completed the tasks significantly faster and more accurately when the vowels were deleted from the second and fourth positions than when they were removed from later word positions.

The study is well designed and controlled, but it is difficult to see exactly how recognizing missing vowels is important in reading. If the study had sought a link between Arabic learners performance on this task and their reading comprehension, there would be evidence that this is an important factor in second language reading for Arabic speakers. As this was not done, the application of this research to reading is somewhat unclear.

It is likely that intraword processing of Arabic speakers is different than that of other ESL learners, and it is possible that this affects their ESL reading comprehension, but this link remains to be demonstrated. Teachers should be aware that L1 orthographic systems may influence ESL readersÕ lexical processing, but it is unclear what, if any, interventions should be employed.

Researchers should determine whether the processing phenomenon tested here impacts ESL reading comprehension of L1 Arabic students. This could be done by comparing the performance of mixed reading ability Arabic ESL students on this measure with their reading comprehension, to determine whether there is a relationship between missing vowel recognition (and the processing strategy it represents) and reading comprehension.

Shanahan, T., Mulhern, M., & Rodriguez-Brown, F. (1995). Project FLAME: Lessons learned from a family literacy program for linguistic minority families. The Reading Teacher, 48, 586-93.

The effectiveness of a community-based ESL/family literacy program in encouraging Latino/a parents to share literacy with their children. There is evidence that parental involvement in childrenÕs literacy can influence childrenÕs scholastic achievement. These researchers created a program to help the parents of potentially at-risk children share literacy in their homes. The report details their program and successes.

The learners were all parents of children who attended school in one of the six Chicago elementary schools where the program was conducted over the course of five years. The schools were all located in heavily Latino neighborhoods in inner-city Chicago. Because of fear of gangs, the classes met in the mornings, so most of the participants were mothers who did not work during the day. The participants (number unspecified) were primarily Mexican and had varying levels of L1 Spanish literacy. All had very limited English literacy.

The teacher participants were a professor and graduate students involved in bilingual education and literacy studies. This research was supported by grants from the U.S. Department of Education and by the Kraft Foundation.

The researchers initially opened the program in one school and later moved to other schools in the area. In each program, parents of children between the ages of three and nine were recruited through the schools. They attended ESL classes twice a week, and attended additional Parents as Teachers classes twice a month. In the ESL classes, they were exposed to communicative language teaching to help them increase their English fluency; in the Parents as Teachers classes they were exposed to ideas for sharing literacy with their children. Some of the activities included finding books in a library, creating a home literacy center, and observing literacy instruction in their childÕs school.

Before each Parents as Teachers class, the students were prepared for the content in their ESL class. The Parents as Teachers classes were conducted in Spanish. In each class of these classes, the teacher researchers presented information on the literacy building techniques, demonstrated activities, practiced them with the parents, and discussed how the parents could implement the activities at home. Then, the parents were assigned to try out the activities they had learned at home with their children. Follow-up on the activities was conducted weekly in the ESL classes.

The researchers gathered qualitative data on the success of the project through student feedback, feedback from the teachers and administrators at the elementary schools, and through their own observations of the program.

The parents involved in the courses became very excited about sharing literacy activities with their children. They felt more comfortable accessing literacy materials from libraries, book sales, and the elementary schools and felt more confident in selecting books for their children in both Spanish and English. Many parents reported initiating patterns of daily reading with their children, and reported that their children were more interested in books and reading as a result of this sharing.

Teachers from the elementary schools felt that the students whose parents participated in the family literacy program progressed more quickly in reading at school and had more positive attitudes towards literacy. Administrators from those schools also noted that parents who participated in the program were more likely to become involved in their childrenÕs school activities. They reported that parents seemed to be more comfortable in the school environment.

The report gives a clear explanation of the program and its goals and some good indications that it was perceived as being very positive. While there is no systematic data collection or any indication of the impact of the program on the parentsÕ ESL literacy, it does raise the possibility that the inclusion of intensive family literacy efforts in adult ESL education could have a positive influence on the literacy development for both parents and children.

It cannot be assumed that simply teaching parents to read or showing them childrenÕs books will have a positive impact on their childrenÕs literacy development. For adult literacy to filter down to children, parents must be exposed to activities and techniques for sharing literacy with their children. It is possible that learning experiences that help parents understand how to share literacy with their children could increase successful family literacy.

An in-depth case study of this type of program could be conducted, employing pre- and post-treatment assessment measures for the literacy development of both parents and children. Such a study could also systematically document changed family reading behaviors during the program.

Strucker, J. (1997) What silent reading tests alone canÕt tell you. Focus on Basics, 1 (B), 13-17.

The reading strengths and needs of different types of readers who test at the same level on silent reading comprehension tests. For this study, a native speaker and a non-native speaker of English who tested at the same intermediate level on an English reading comprehension test were compared to exemplify the different learner profiles in adult literacy programs.

The native speaker was a male in his twenties named Richard. Because his family had moved often in his childhood, RichardÕs reading difficulties were not diagnosed until he was in late middle school. He eventually dropped out of high school in order to help his family financially. After working in a number of jobs, he wished to earn a high school degree to join the military.

The ESL student was a female in her early twenties named Vanessa, a Peruvian immigrant who had dropped out of high school. She had completed beauty school, but did not posses a high enough level of English literacy to pass the licensing test. She wished to enter a job-training program.

The teacher was a researcher for a national reading foundation who had more than ten years experience working with native speakers and ESL learners in adult basic education literacy programs. He completed this research as part of a larger study of the profiles of adult basic education learners in the United States.

The learners completed several testing measures with the researcher, including a silent comprehension test, an oral reading, an oral vocabulary assessment, a spelling test, and tests of word analysis and word recognition. The researcher also noted their placement into classes in the adult education center and followed their progress.

As was mentioned, based on the silent reading comprehension test, the two learners were at the same level and would have been placed in the same class. The other measures, however, revealed significant differences in their reading behavior and needs. Richard had excellent meaning-based skills. He was able to resolve oral miscues, for example, based on context, and had strong oral vocabulary skills. However, his print-based skills were very weak. He had difficulty decoding, as evidenced by low word analysis and word recognition scores, and his low scores in spelling. While his silent reading comprehension score alone would have placed him in an intermediate class, he really needed more basic decoding and spelling practice to develop reading fluency and accuracy. In this setting, Richard began to progress in both meaning-based and print-based skills.

Vanessa had an opposite balance of skills. Her print-based skills, specifically spelling and word recognition were quite high, and the majority of her errors seemed to be prompted by her native language, Spanish. However, her comprehension scores were impaired by low vocabulary. Her need for increased vocabulary was also underscored by lower oral vocabulary scores. Thus, while VanessaÕs print-based skills were quite good, a limited vocabulary impeded her reading comprehension. She needed different instruction than Richard; her profile indicated that she required extensive reading and vocabulary building. She received this in an intermediate class and in additional ESL classes offered at the center.

The researcher relates these finding to those of his larger study, indicating that VanessaÕs profile is very typical of certain adult ESL students, while RichardÕs is typical of certain native speakers.

The study is reported with rich detail and makes a convincing case for the use of multiple assessments, rather than just silent reading comprehension tests, for the placement of ESL and native speaker adults in reading programs, as well as explaining the need for specialized instruction for different learner types and for not putting them in the same class just because they test the same on a silent reading comprehension test. It effectively illustrates the fact that not all learners at a similar comprehension level have the same needs, and that silent reading comprehension tests alone will not be able to uncover the differences between learners of different profiles.

Multiple assessments should be employed when placing students in literacy programs. The use of oral reading and vocabulary assessments are especially helpful in determining which students only have difficulties with the language, and which also have difficulties with reading.

While the researcher of this article is currently extending the research to a large population of adult basic education literacy learners, it could also be extended to ESL adult literacy students. The effects of educational background, native language literacy, and cultural background on literacy profiles could be considered.

Strucker, J. & Davidson, R. (in press) Adult reading components study (ARCS). Boston: National Center for the Study of Adult Learning.

Reading profiles of adult education students (both native and non-native speakers of English) in the United States.

Six øhundred seventy-seven adult basic education learners and 278 ESL learners at learning centers in seven states. 219 of the ESOL learners spoke Spanish as their L1. Of the Spanish speaking ESOL learners, 80% had an average of 12-13 years of education, 14% had an average of 8 years of education, and 5% had an average of 6.5 years of education in their native language, and appeared reading disabled in Spanish.

All the participants completed a battery of reading proficiency tests. These included measures of silent reading comprehension, oral reading, spelling, word recognition, word analysis, phonemic awareness, automaticity, and short-term memory. The Spanish-speaking participants also completed Spanish word recognition, oral vocabulary, and reading comprehension tests.

The highest cluster of Spanish-speakers in terms of prior education in their native language had high levels of vocabulary, decoding, and comprehension in Spanish. The Spanish-speakers with less educational background tended to have lower Spanish vocabulary and reading comprehension. The Spanish-speakers with the least educational background had very weak decoding skills in Spanish and appeared to be reading disabled. They also tended to have not acquired English reading skills, vocabulary, and listening comprehension.

However, even the Spanish-speakers with high literacy levels in their L1 had difficulties in some aspects of English decoding, especially letter sounds in isolation, despite having, on average, several years of ESOL classes.

This research looks at reading proficiency both in the L1 and L2. The numbers of students being studied is large enough to have importance. Of great interest is the fact that it assesses and provides data on a fact that many ESL teachers have known anecdotally:Ê unlike the native speaker adult basic education (ABE) population, the percentage of adult English language learners who are learning disabled is probably no greater than that of the general population.

Even highly literate students may need instruction and practice in some basic English decoding skills, especially at the phoneme level. Learning English letter sounds could help them attend to small differences that affect meaning, especially morphosyntactic information at the ends of English words, and it might help them to be more successful at English pronunciation. Also, given that 80% of the Spanish speakers tested had high school or above levels of Spanish reading, some of these students could perhaps benefit from a faster-paced track in ESL, one more geared to take advantage of their L1 literacy skills and relatively high levels of L1 education.

Different methods of decoding instruction could be designed to help previously literate students transfer decoding skills from their first language to the second. The effectiveness of such methods should be empirically tested.

Tan, A., Moore, D.W., Dixon, R.S., Nicholson, T. (1994). Effects of training in rapid decoding on the reading comprehension of adult ESL learners. Journal of Behavioral Education, 4, 177-189.

The effect on reading comprehension and reading speed of training readers to rapidly recognize words in isolation. Much research has indicated that increasing the automaticity of reading processes like word recognition can improve both reading speed and reading comprehension. The researchers wanted to empirically evaluate this claim.

Three learners participated in this study. All the subjects were recent immigrants to New Zealand who were studying ESL. Three different L1 language backgrounds were represented as the subjects came from Japan, China, and Cambodia. At the beginning of the study, their scores on the Neale Analysis of Reading Behavior indicated that they read at an age-equivalent of 7-9 for reading rate, 6-7 for accuracy, and 6-7 for comprehension.

The study was designed as a multiple-baseline study, in which all the participants completed multiple assessments of their reading behavior over time. The assessments and treatments were centered on reading passages that the researchers prepared. Each of the fifteen passages was rated at the seventh- grade level. For each passage, the researchers prepared a word list of 25 words from the passage that the researchers thought the subjects might find difficult. Comprehension tests consisting of ten text-explicit and two text-implicit questions were prepared by the researchers.

All participants attended three treatment/experimental sessions a week for five weeks. The sessions were conducted by the first author in a one-on-one setting.

In the initial baseline sessions, the participants were given the word list and asked to read it out loud. Errors in accuracy and elapsed time were recorded. Then the readers were given the passage and asked to read it aloud. They were informed that comprehension questions would follow. Errors in accuracy and elapsed time for the oral reading were also recorded. Following the reading, the subjects orally answered the orally-administered comprehension questions.

For the treatment sessions, words from the word lists were printed on note cards. The participants practiced with the cards in flashcard style until they could recognize all the words in one second or less. They were then given the words in list form and were required to read the list accurately at a rate of one and one-half second or less per word. If they were unable to do this, they reviewed the flashcards again. They then read the text aloud and answered the comprehension questions.

For the final week of the treatment sessions, the researchers initiated the reversal stage, which was an exact return to the baseline procedures.

The participants all decreased the mean time per word for reading words in the vocabulary lists during the treatment phase of the study, but during the reversal phase returned to the baseline speeds, indicating that the treatment did not affect their reading speed for reading new words in isolation. However, their reading speed overall for reading the passages decreased during the intervention stage of the study and remained lower during the reversal stage, indicating that practicing reading at faster speeds did help the participants to increase their overall reading speed.

In terms of accuracy, the participants were most accurate in their reading of the words both on the lists and in the passages during the treatment stage of the study. During the reversal stage, their accuracy scores returned to the baseline levels. The comprehension scores followed a similar patternÑthe participants significantly improved their comprehension scores during the treatment phase of the study, but during the reversal phase their comprehension scores returned to near-baseline levels.

The authors conclude that the lasting benefit of the treatment was in the increased reading speed, the only improvement that the readers generalized to readings when the treatment was not given. They attribute the gains in reading speed to the volume of the English reading practice that the students engaged in, not specifically to the treatment itself. They conclude overall that the study of word lists before reading only improves reading speed, accuracy, and comprehension on the passages associated with the lists, and that these benefits are not generalized to reading other texts.

The study had some good design features, the statistics wer well explained, and the findings responsibly interpreted. However, an experimental study of only three subjects can hardly yield significant, generalizable results.

The lack of results for the treatment could possibly be attributed to the lack of recycling in the vocabulary use. It is possible that students could generalize the reading comprehension gains from improving rapid word decoding if the words they practiced with were recycled in subsequent readings. The researchers established that these skills do not generalize to new words. It would interesting to see if the improved decoding time would be maintained with the same words in new contexts.

The study indicates that the benefits of improving vocabulary recognition before reading are not generalized to reading overall. This indicates that, at least without recycling, work with word lists may not be a productive activity for improving reading comprehension.

A similar study, with more subjects, could recycle the words studied in the treatment phase in the reversal phase, to see if accuracy, speed, and comprehension gains from studying word lists are maintained when the same words appear in new contexts.

Thuy, V.G. (1992). High-tech for effective ESL/family literacy instruction. Harrisburg, VA: Pennsylvania State Department of Education. (ERIC Document Reproduction Service No. ED 356 683).

The effects of utilizing CD-ROM software in a community ESL literacy program for limited English proficient adults. The study was conducted to determine what the benefits and drawbacks of using this technology were measured in terms of language acquisition, retention, and attendance. The purpose of the study was to aid future teachers, tutors, and lab instructors in the use of technology in ESL literacy programs.

Thirty refugee and immigrant ESL students from Southeast Asia participated in the study. All of the students completed pre-tests of vocabulary and comprehension for each of the stories used in the computer programs. The researchers attempted to divide the students into groups with equivalent mean scores, but they allowed students in the end to self-group. The result was an intermediate experimental group (pre-test mean=56.67), and roughly equivalent beginning level experimental (pre-test mean=3.19) and control (pre-test mean=3.13) groups.

All of the teachers participating in the study were teachers from the Indochinese-American Council. They all had extensive experience in ESL literacy instruction in multilingual classes.

The pre-tests and division into experimental and control groups occurred at the beginning of the fifteen week semester. The classes met for two hours three days a week throughout the semester. The first hour of every class for both groups was spent on general ESL instruction. The experimental group spent the second hour of each class in the computer lab using CD-ROM software designed to build ESL literacy skills. For the control group, the second hour of each class was spent working on literacy using paper versions of the CD-ROM content. Post-tests, exit interviews, and questionnaires were completed at the end of the semester.

To report the findings, the experimental group was split into two groups: beginning and intermediate. No similar split was made with the control group, which by the end of the report is characterized as beginning level. The beginning experimental group improved their scores on vocabulary and comprehension more than either the intermediate experimental group or the control group. The intermediate experimental group improved their vocabulary scores more than the control group as well. These effects could have been due to the level of the instructional material. If the material was more suited to the beginners, the intermediates wouldnÕt have had that much to learn from it, and so would not have showed gains.

From the questionnaires and the exit interviews, it was found that the learners who participated in the experimental group viewed the experience very positively. They all expressed interest in continuing to study using technology. The researcher also noted that the technology had been effective in eliciting speaking from otherwise shy learners. He also noted that the use of technology was a positive experience because most of the learners had little access to computer equipment, and viewed the program as a chance to gain both language and computer skills.Ê

There is no explanation given for splitting the experimental group into two groups by proficiency but not doing the same for the control group. If the groups were really formed at the beginning of the study such that the distribution of scores was similar between groups, there should have been no reason for dividing them. If the control group really was a beginning level group, then this group can only act as a control for the beginning level experimental group, and there is thus no adequate control for the intermediate experimental group. It is impossible to determine whether their gains were the result of the exposure to computer technology or the result of the pre-program proficiency level.

Dividing the experimental group also meant that there were two fewer members in each group to perform inferential statistics. We cannot tell from this study whether the improvements are significant, nor whether the differences among the performance of the different groups were significant.

The qualitative review of the participantsÕ interviews, while not reported with very rich detail, did give interesting insights into some of the perceived advantages of integrating technology into an adult ESL literacy setting. Predictably, the subjects themselves liked using the computers.

Supplementing ESL literacy with technology can increase student motivation in literacy learning. It can also add to literacy learning an increased confidence in using technology, which is increasingly important for progression in the workplace. Possibly, the incorporation of technology into ESL literacy classes can contribute to accelerated learning.

This sort of study could be replicated using more modern technology, such as some of the web-based programs directed toward ESL learners. A replication could either employ more rigorous pre- and post-test measures and statistics or could gather more in-depth detail on student reactions to working with technology.

Tse, L. (1996a). When an ESL adult becomes a reader. Reading Horizons, 37, 16-29.

The personal and behavioral changes an ESL adult student experiences when introduced to pleasure reading. Helping studentsÕ change their reading behavior by encouraging them to read more and to focus on meaning is hypothesized to increase their language development by massively increasing their exposure to input. The researcher wanted to know what specific motivational and behavioral changes a learner would experience during this experience.

This is a case study of one student participant in a reading class in an intensive ESL program. The participant was an Indonesian woman who had come to the United States with her husband and children to study English in order to eventually pursue a masterÕs degree at an American university. Although she had studied English for many years, she had never read books in English and had never engaged in English pleasure reading at all. She also rarely read for pleasure in her native language.

The teacher researcher of this study was the teacher of the extensive reading class mentioned and had designed the syllabus for the class. Her teaching philosophy for literacy centered on the benefits of extensive reading with a focus on meaning. She instructed her students to read quickly and guess the meanings of new words, rather than using dictionaries, in order to increase their exposure to input. She also acted as an individual tutor for the participant in this study, meeting with her twice a week for two months following the end of the four-month course.

Data were collected through the participant observations of the practitioner researcher during classes and through interviews with the participant over the six months of the course and tutoring. At the first of the course, the researcher asked the participant about her reading habits and attitudes toward reading in her language and in English. During several of the tutoring sessions following the course, the researcher and the participant returned to this theme and to the ways in which the participantsÕ reading behavior and attitudes had changed. Data from interviews and observation were recorded in notes taken by the researcher.

At the beginning of the study, the participant felt uncomfortable reading and unsure of her ability to understand the readings. She indicated that she had liked reading, and initially did not participate much in the in-class discussions of the reading. When she began to participate in class, her comments were mostly language-focused questions. After a few weeks, however, she began to engage more in discussions of the issues raised in the readings.

The interview data were analyzed according to three categories of comments: opinion and value, feeling, and knowledge. The participant indicated that throughout her participation in the class, her opinion of reading as a part of language learning had changed. While before the class she had considered reading to be secondary to grammar and vocabulary study in language learning, by the end of the course she considered reading to be the best way for her to learn English and to improve her vocabulary, spelling, and writing skills. In terms of feeling, or her emotional response to the reading experience, she felt that she had overcome unproductive strategies such as the overuse of the dictionary and that she had become confident in her ability to read in English. She also felt that the books had increased her knowledge of American culture and beliefs, which gave her new insights into her own culture. The researcher also noted that the participant had significantly increased her reading ability. Whereas she began the semester struggling with third-grade-level books, she ended the semester reading seventh-grade-level books comfortably. That is gain of four grade levels in less than six months.

This is a one-participant case study. Care should be taken in generalizing findings beyond this participant. However, the researcher supplies rich detail on the participantÕs attitudes and beliefs about reading, as well as her classroom behavior and reported reading behavior. It is a good window into the personal and behavior are changes a learner may experience when engaging in reading in a second language. Having adult ESL learners read and talk about literature, poetry, and art (to supplement acquiring functional English) can make the process of being a newcomer and a new language learner less ÒinfantilizingÓ and more adult.

This study indicates that introducing ESL adult students to pleasure reading can help them to alter their reading attitudes, reading behavior, and reading proficiency in a positive manner. It indicates that students as well as teachers can come to see pleasure reading as a legitimate language learning experience.

This research could be replicated with a larger number of participants. Also, language-based pre- and post-testing could also be introduced to quantify the effects of pleasure reading on language development.

Tse, L. (1996b). If you lead horses to water they will drink: Introducing second language adults to books in English. California Reader, 29 (2), 14-17.

The effectiveness of an approach to introducing second language adults to reading in English. The approach involved explaining the importance of reading in language development to the students, helping students comprehend popular novels, and helping students to read more effectively.

The learners were ten adult intermediate and advanced ESL students in a community language course that met for thirty hours over ten weeks. The ten students chose to take the course because they wanted to improve their reading ability. None of the students had ever read a book in English.

In the first stage of the class, the teacher discussed with them some research on second language reading that indicates that reading facilitates language development, and then explained the notion of implicit learning (learning from comprehensible input) to the students to help motivate them throughout the project. Then, the teacher shared some ideas with them to help make reading more enjoyable. These suggestions were all directed to reading more quickly, fluently, and frequently. The class read six novels during the course. The teacher chose the first two and then allowed the students examine other books and choose as a class the remaining four novels they would read. During the course, the class participated in activities and discussions centered on the issues and themes of the stories they were reading. Each week, the students were asked to write briefly about their experiences with the reading and the difficulties they encountered.

The students reported progress in their level of comfort while reading in English as the semester progressed. They reported that they had become interested in the reading and were relying less on the dictionary. While all of the students had good sentence processing skills at the beginning of the study, some noted that their ability to comprehend the text as a whole had increased. Their increased fluency decreased the frustration of reading for them. Students also reported that reading was easier when they were interested in the topic. Many indicated that they planned to continue reading in English after the semester ended.

The study indicates that exposure to English novels and an understanding of the importance of developing literacy to developing language skills can motivate students to change their reading practices and their attitudes about reading. The study does not indicate that this program influences studentsÕ language development, as no language assessment measure was used.

StudentsÕ motivation to read, confidence in themselves as readers, and reading habits can be affected by extensive reading of books that students find interesting. Reading instructors should find books that appeal to their studentsÕ interests. Instructors should, as did the teacher in the study, explain the rationale for our instruction to our students, thus taking advantage of their adult metacognitive abilities and making them more self-conscious learners.

This study should be replicated to include assessment measures, so the effect of the reading experience on the studentÕs language development can be better assessed.

Wade-Woolley, L. (1999). First language influences on second language word reading: All roads lead to Rome. Language Learning, 49(3), 447-471.

The influence of learnersÕ L1 on their processing of second language written words. The author hypothesized that different L1 orthographic systems could differentially influence the way that learners access words from their lexicons due to different orthographic and phonological processing patterns.

The learners involved in this study came from two different groups: Japanese L1 ESL students in an intensive English program (IEP) at a Canadian university and Russian L1 learners of English and Hebrew at an IEP/Intensive Hebrew Program (IHP) at an Israeli university. All the learners were successful readers in their L1. Prior to the beginning of the study, none of the learners had lived in an area where English was primarily spoken (the Japanese students were all tested at the beginning of their first semester of study in Canada and had been in Canada for less than three weeks). They were of similar ages, similar gender distribution, and had similar backgrounds in English study.

Of this pool of subjects, sixteen from each group was selected based on having completed all the instruments and on being matched to a learner from the other group. Therefore, there were sixteen pairs of students in the study; each pair was equivalent based on their scores on a test of word recognition. A comparison of the participantsÕ scores on reading and vocabulary sections of the TOEFL indicated that there were no significant differences in reading ability between the two groups.

Since reading ability and exposure to English were held roughly constant, the main difference between the groups was the first language. Both of these languages differ substantially from English; neither of them uses Roman orthography. Japanese and Russian writing systems are of different types; Russian uses a phonologically-based alphabet, while Japanese uses a syllabary (kana) and a logographic system (kanji). Research on L1 reading has indicated that readers of different alphabet types use different word-level decoding processes; the researcher sought to determine what effect these would have on second language reading.

All of the students completed several measures of word-level reading and decoding:

1. TOEFL reading and vocabulary subsections. This test was used to estimate the similarity between groups of reading level.

2. Woodcock Reading Mastery TestÑWord reading subtest. This is a test of isolated word recognition. The subjects were presented with words printed on cards and asked to read them out loud. Performance on this test was used to match student pairs across groups to ensure group equality.

3. Woodcock Reading Mastery TestÑword attack subtest. This tests learnersÕ ability to apply structural analysis of words and phonics strategies to pronounce isolated unfamiliar words. Learners were presented with unfamiliar words printed on cards and required to produce a natural reading within 5 seconds.

4. Peabody Individual Achievement TestÑSpelling recognition subtest. On this test, participants are presented with sets of four visually and phonologically similar words, only one of which is a correctly spelled word. A word is pronounced aloud and the participant is required to point to the item matching that word.

5. Orthographic knowledge. Students were presented with pairs of pseudowords (e.g., filv-filk) and asked to indicate which was more English-like.

6. Phoneme deletion. Students were presented with a psuedoword and asked to pronounce it. They were then instructed to delete a target phoneme from the syllable onset or coda (e.g., if given the stimulus smeck, learners could be asked to say it without the /s/).

ÊAs was mentioned before, there were no significant differences between groups on the TOEFL scores and the word recognition scores were used to match students in the groups to further ensure equality. There was also no difference between the groups on their pseudo-word decoding and on their working memory as measured by pseudo-word recall.

The Japanese L1 speakers were faster and more accurate on tasks involving recognition of orthographic patterns, in both real English words and in pseudowords. The Russian L1 speakers, however, were faster and more accurate in deleting phonemes from words. The researcher suggests that these differences are likely due to differences in their L1 literacy experience. The Russian L1 speakers are used to a phonologically-based alphabetic system. Like English speakers, they are likely to access words from writing through the phonological system, because they are used to having a relationship between the pronunciation and writing of words. The close relationship between the phonological system and their reading ability is demonstrated in their ability to put non-words through phonological processes like segment deletion.

In Japanese, the writing does not reflect the phonology. It is more likely that Japanese readers do not access words solely from phonology, but also from their orthography knowledge. Therefore, these learners are not used to focusing on phoneme to sound mapping in reading, but rather on sight recognition of letter sequences. This emphasis could give them an advantage in performing English orthography tasks.

The study seems very carefully done and well controlled. It does seem reasonable that different learners would focus on the text-processing strategies that are most important in their native language literacy as well. The study does not, however, include any developmental data or any indication of the effects of the use of different strategies on reading development in a second language.

Teachers and researchers should recognize that reading tasks require learners to use their reading knowledge in specific ways that might be influenced by their native language literacy rather than by their reading ability in the second language. Scores from any one task type may not be indicative of overall student ability, but rather of preferred processing strategies. Teachers should be aware of the native language literacy they students have so as to build upon the strategies the students are likely to have in L1.

The issues raised in this paper (of L2 learners using different processing strategies for word retrieval in reading) could be corroborated with data from other language groups and from learners at different levels of second language proficiency.

Wilson, M. (1983). Developmental patterns of reading proficiency in adult ESL students: Implications for ESL classrooms. Toronto, Canada: TESOL. (ERIC Document Reproduction Service No. 248 709)

The changes in learnersÕ miscue behaviors as they improved their reading comprehension over time. The researchers were interested in how learnersÕ total miscues, semantically acceptable miscues, and syntactically acceptable miscues change over time and how these related to learnersÕ reading comprehension and views of reading.

The learners were seven ESL speakers from mixed ethnolinguistic backgrounds (native languages not specified) who were all studying ESL at a community ESL program. All of the learners at the beginning of the study were ranked as beginning/low intermediate level speakers according to the Michigan Placement Test. No information is given about their ESL vocabulary or L1 educational background. All of the students had studied English for a 15 months or more.

The researcher audio-taped the learners reading a story and then asked them to retell the story. The selected stories were written at a level of difficulty at which learners were likely to make errors. The learners continued to study at the ESLÊ program from four months. At the end of the four months, the researcher audio-taped them again reading a different story at a similar level of difficulty to the first story. They were again asked to retell the story.

All the learners but one decreased their overall miscues per one hundred words of text. However, there were mixed patterns on their production of syntactic and semantic errors. All of the learners made more syntactically than semantically possible errors, and all learners increased their percentage of syntactically possible errors. However, some learners decreased their percentage of semantically possible errors from the first reading to the second. And while all learners had fewer semantically than syntactically possible errors, for some the difference was very pronounced. For example, learner 2 made many more syntactically possible miscues than semantically possible (75% vs. 17%) while learner 3Õs percentage of syntactically possible miscues (68%) was much more similar to his percentage of semantically possible miscues (53%).

Interestingly, the readers who made the fewest oral miscues were not necessarily the ones who comprehended the stories best, as illustrated by their oral retellings of the story. To illustrate this point, the researcher describes in detail the performance of two learners. One, a Chinese woman, a lower overall use of miscues than the other learner, a Spanish woman. However, her miscues were much more likely to be non-words, showed a much higher degree of graphic similarity to the printed word, and were much less likely to be semantically possible in the context. Her oral retellings indicated that, while she had read the stories very accurately, she had not comprehended them. She indicated that she had been paying attention to the graphic information and had not thought about meaning. The Spanish learner, on the other had, indicated that she read to understand ideas. While she made more errors, she was able to adequately retell both of the stories. This section of the data analysis does not contain information on how these learnersÕ comprehension of the data improved over time, nor on how the retellings were evaluated.

The findings illustrate the possibility that readers may be able to accurately read text that they do not comprehend, an important fact for teachers to acknowledge.Ê However, it is not clear whether vocabulary differences were actually producing this result, since the study designers didnÕt track this critical variable. Because mixed L1 backgrounds are always a problem, especially in a small ÒNÓ study such as this, one wonders why the author chose to do the study with a heterogeneous group.

The fact that the Spanish womanÕs miscues were much more likely to be non-words, and showed a much higher degree of graphic similarity to the printed word may be indicate that she was reading with her Spanish L1 phonetic system, rather than with English.

The nature of oral retelling may actually prompt learner over-attention to graphic information. It is hard to imagine that readers, regardless of their orientation to reading, would pay more attention to pronunciation than to meaning when reading silently. While this study claims to illustrate the harm to comprehension of reading for graphic accuracy, it is difficult to see how this translates to non-oral reading situations. Because of the difficulty experienced in answering questions after oral reading ø due to the fact they have been concentrating on reading perfectly, and in ESL learnersÕ case, pronouncing properly-- at the very least,Ê in this study.Ê the researcherÊ should have allowed the people to re-read the stories silently before asking them to re-tell.

Likewise, it is difficult to follow the researcherÕs claims about the learnersÕ comprehension of the passages without any information on how the retellings were evaluated.

Readers who pay excessive attention to graphic information in oral reading may not be taking in meaning. Essentially, this means that learners who do not read for meaning may need practice reading in situations that push them to focus on the meaning of the text. Silent reading, rather than oral reading, might be more suited to helping students process for learning. Teachers should be careful to balance silent and oral reading for both pedagogical and assessment purposes. Learners who do not read for meaning may need practice reading in situations that push them to focus on the meaning of the text.

This study could be replicated examining the reading comprehension of learners reading aloud and silently to determine if the reading orientations examined in this study affect silent reading comprehension as well.

Zimmerman, C. B. (1997). Do reading and interactive vocabulary instruction make a difference? An empirical study. TESOL Quarterly, 3(1), 121-140.

The effect on reading and vocabulary knowledge of specific, interactive vocabulary instruction integrated into an ESL literacy curriculum.

The learners were recruited from an intensive, pre-academic English program connected with an American university. Four intact classes from the advanced intermediate (level 5 of 6) were used in the study. Two classes (18 students) served as the experimental group; two classes (17 students) served as the control group. The students came from mixed ethnolinguistic backgrounds. Many planned on earning a degree in the United States.

The teachers in the study were all experienced ESL teachers. One teacher taught both of the experimental group classes; two teachers each taught one of the control group classes.

At the beginning of the study, all the students completed a vocabulary checklist test to estimate vocabulary size. They also completed a questionnaire that asked them to rate the effectiveness of various types of vocabulary instruction. The questionnaire was designed to investigate studentsÕ beliefs about vocabulary learning. During the next ten weeks, they all participated in regular ESL courses. Their teachers instructed them all to read a minimum of five hours a week of self-selected texts. They were instructed to hand in weekly reading records of the amount and source of their reading each week. These records were not graded to encourage honest reporting. Teachers also reported the amount of required reading for their classes that the students were assigned to do during these ten weeks.

During the ten-weeks of the experiment, the students in the experimental group participated in three hours per week of vocabulary instruction, using a vocabulary teaching text, Lexis: Academic Vocabulary Study (Burgmeier, Eldred, & Zimmerman, 1991). The instruction consisted of approximately 80% group interactive work, during which the teacher was available for assistance. At the end of the study, the students again completed the vocabulary checklist test, and again completed the questionnaire about preferred reading instruction.

There were no significant differences in the amount of self-selected reading between the experimental and control groups in the study. Students who participated in vocabulary instruction improved their vocabulary knowledge as measured by the pre- and post-treatment checklist test significantly more than the students who did not participated in this instruction. Since these students participated in approximately the same amount of reading during the study, the researcher attributes this difference to the vocabulary instruction, rather than incidental vocabulary learning through extensive reading.

At the beginning of the study, the students had roughly similar vocabulary instruction preferences. By the end of the study, students in the experimental group indicated that they preferred learning vocabulary in interactive contexts and through reading. The researcher notes that the difference in the perceived effectiveness of vocabulary learning through reading could be motivated by the greater amount of reading assigned to the experimental group (50% more), assuming that engaging in more reading helped these learners to perceive the instruction type as more effective.

The study investigates an important questionÑthe usefulness of vocabulary instruction in the L2 classroom. Many teachers avoid direct vocabulary instruction, assuming that students will be able to learn vocabulary from reading. This study challenges the assumption that reading instruction without vocabulary instruction can be sufficient for language learning.

However, the use of intact classes in this research complicates the experimental design. There were two, rather than one, differences between the treatments experienced by the experimental group and the control group. The experimental group engaged in both direct vocabulary instruction and approximately twice as much required reading. It is impossible to determine how much of the difference in their post-treatment performance can be attributed to vocabulary instruction, and how much can be attributed to increased input through required reading assignments.

Additionally, it should be remembered that all the measures of the effectiveness of vocabulary instruction are self-reported. The checklist tests require students to rate their familiarity with a word. It is possible that students who have engaged in significant amounts of vocabulary instruction are more likely to rate their own vocabulary knowledge generally more positively, in order to assert that the instruction time was not wasted.

Finally, it is unclear from the description of the research design of this study whether the three hours per week of vocabulary instruction were part of the regular instructional time for the experimental group, or if this instruction represented additional instructional time. If the latter is the case, the improvement on vocabulary might be attributable to more instructional time and input, rather than the vocabulary instruction itself.

It is possible that extensive reading alone is insufficient for helping ESL students achieve the level of vocabulary knowledge they need to be communicatively successful in English. Teachers and researchers should investigate the effectiveness of vocabulary instruction as a part of literacy instruction.

This study could be replicated in a more controlled manner, to ensure the effectiveness of vocabulary instruction along with reading over reading alone. The various methods of vocabulary instruction could be compared to determine what is most effective.

Note: There has been little research performed on ESL adults learning to read in English. For this reason, we have included all we found, whether or not it was research that was well conducted. We wanted to show the universe of it exists. Had we insisted only on studies that were rigorously done, we would have severely limited the base.

Alphabetical Listing of Articles

Adams, M. J. & Collins, A. (1985). A schema-theoretic view of reading. In H. Singer & R.B. Ruddell (Eds.), Theoretical models and processes of reading (pp. 404-425). Newark, DE: International Reading Association.

Alderson, J.C. (1984). Reading in a foreign language: a reading problem or a language problem? In J.C. Alderson & A.H. Urquhart (Eds.), Reading in a foreign language (pp. 1-27). New York: Longman.

Brown, C. (1993). Factors affecting the acquisition of vocabulary: Frequency and saliency of words. In T. Huckin, M. Haynes, & J. Coady (Eds.), Second language reading and vocabulary learning, (pp. 263-86). Norwood, NJ: Ablex Publishing Co.

Carrell, P.L. (1989). Metacognitive awareness and second language reading. Modern Language Journal, 73 (2), 121-134.

Carrell, P.L. (1991). Second language reading: Reading ability or language proficiency. Applied Linguistics, 12, 159-179.

Carrell, P.L., Pharis, B.G., & Liberto, J.C. (1989). Metacognitive strategy training for ESL reading. TESOL Quarterly, 23 (4), 647-678.

Carroll, S.D. (1999). Storytelling for literacy. Grand Rapids, MI: Annual Meeting of the Michigan Reading Association. (ERIC Document Reproduction Service No. ED 430234)

Carson, J.E., Carrell, P.L., Silberstein, S., & Kuehn, P. (1990). Reading-writing relationships in first and second language. TESOL Quarterly, 24, 245-266.

Chervenick, E.C. (1992). Schema theory: Teaching U.S. history to beginning amnesty students. Unpublished masterÕs thesis, Biola University, La Miranda, CA. ERIC Document Reproduction Service No. ED 354 787.

Cho, K.S. & Krashen, S.D. (1994). Acquisition of vocabulary from the Sweet Valley Kids series: Adult ESL acquisition. Journal of Reading, 37, 662-667.

Coady, J., Mgoto, J., Hubbard, P., Graney, J. & Mokhtari, K. (1993). High frequency vocabulary and reading proficiency in ESL readers. In T. Huckin, M. Haynes, & J. Coady (Eds.) Second language reading and vocabulary learning (pp. 217-228). Norwood, NJ: Ablex Publishing Corporation.

Coady, J. (1997). L2 vocabulary acquisition through extensive reading. In J. Coady & T. Huckin (Eds.) Second language vocabulary acquisition (pp. 225-237). Cambridge: Cambridge University Press.

Constantino, R. (1995). Reading in a second language doesnÕt have to hurt: The effect of pleasure reading. Journal of Adolescent and Adult Reading, 39, 68-69.

Cotterall, S. (1991). Meiko: Case study of a second language reader. Reading in a Foreign Language, 8(1), 611- 627.

Devine, J. (1988). A case study of two learners: Models of reading and reading performance. In P.L. Carrell, J. Devine, & D.E. Eskey (Eds.) Interactive approaches to second language reading (pp. 127- 151). Cambridge: Cambridge University Press.

Eskey, D.E. (1997). Models of reading and the ESOL student: Implications and limitations. Focus on Basics, 1(B), 9-11.

Fitzgerald, N.B. & Young, M.B. (1997). The influence of persistence on literacy learning in adult education. Adult Education Quarterly, 47(2), 78-91.

Garreton, R. & Terdy, D. (1991). Correlation study of adult English as a Second Language (ESL) and Adult Basic Education (ABE) reading tests. Boston: Adult Learning Resource Center. (ERIC Document Reproduction Service No. ED 352 849)

Goldberg, R. (1997). Deconstructing the great wall of print. Connections: A Journal of Adult Literacy,1, 8-13.

Goodman, K. (1988). The reading process. In P.L. Carrell, J. Devine, & D.E. Eskey (Eds.) Interactive approaches to second language reading, (pp. 11-21). Cambridge: Cambridge University Press.

Gough, P.B. (1985). One second of reading. In H. Singer & R.B. Ruddell (Eds.) Theoretical models and processes of reading (pp. 661-86). Newark, DE: International Reading Association.

Grabe, W., & Stoller, F. L. (2002). Comparing L1 and L2 reading. In Teaching and researching reading (pp. 40ø63). Harlow, England: Longman.

Griffin, S. M S. (1990). The effect of media choice on English literacy acquisition for preliterate Hmong adult learners. Unpublished doctoral dissertation, University of Washington, Seattle, WA.

Hardman, J.C. (1999). A community of learners: Cambodians in an adult ESL classroom. Language Teaching Research, 3, 145-66.

Haynes, M. (1993). Patterns and perils of guessing second language reading. In T. Huckin, M. Haynes, & J. Coady (Eds.) Second language reading and vocabulary learning (pp. 46-64).

Hilferty, A.G. (1996). Coding decoding: Predicting the reading comprehension of Latino adults learning English. Unpublished doctoral dissertation, Harvard University, Cambridge, MA.

Hsueh-chao, M.H. & Nation, P. (2000). Unknown vocabulary density and reading comprehension. Reading in a Foreign Language, 13(1), 403-430.

Hudson, T. (1982). The effects of induced schemata on the Òshort circuitÓ in L2 reading: Non-decoding factors in L2 reading performance. Language Learning, 32, 3-31.

Huntley, H.S. (1992). The new illiteracy: A study of the pedagogic principles of teaching English as a second language to non-literate adults. Unpublished manuscript. (ERIC Document Reproduction Service No. ED 356685)

Joe, A. (1998). What effects do text-based tasks promoting generation have on incidental vocabulary acquisition? Applied Linguistics, 19 (3), 357-377.

Johnson, K.E. (1992). The relationship between teachersÕ beliefs and practices during literacy instruction for non-native speakers of English. Journal of Reading Behavior, 14, 83-108.

Jones, M.L. (1996). Phonics in ESL literacy instruction: Functional or not? Philadelphia, PA: World Conference on Literacy. (ERIC Document Reproduction Service No. ED 436 104)

Kessler, C., Cohen, B., & Walsh, R. (1996). Classroom interaction for adult literacy. Netanya, Israel: The Second International Conference on Teacher Education. (ERIC Document Reproduction Service No. ED 405743)

Klassen, C. & Burnaby, B. (1993). ÒThose who know:Ó Views on literacy among adult immigrants in Canada. TESOL Quarterly, 27, Ê377-397.

Koda, K. (1999). Development of L2 intraword othrographic sensitivity and decoding skills. Modern Language Journal, 83, 51-64.

Lantigne, B. & Schwartzer, D. (1997). The progress of Rafael in English and family reading: A case study. Journal of Adolescent and Adult Literacy, 41, 36-45.

Laufer, B. (1997). The lexical plight in second language reading: Words you donÕt know, words you think you know, and words you canÕt guess. In J. Coady & T. Huckin (Eds.) Second language vocabulary acquisition (pp. 20-34). Cambridge: Cambridge University Press.

Learning Disabilities Association. (1994). Learning disabilities and the acquisition of English language skills in the adult ESL population: A demonstration project. Minneapolis, MN: Author.Ê (ERIC Document Reproduction Service No. ED 382 052)

McLeod, B. & McLaughlin, B. (1986). Restructuring or automaticity? Reading in a second language. Language Learning, 36, 109-123.

Mikulecky, L. (1992). Workplace literacy programs: Variations of approach and limits of impact. San Antonio, TX: National Reading Conference. (ERIC Document Reproduction Service No. ED 353461)

Packard, B. W-L. (2001). When your mother asks for another book: Fostering intergenerational exchange of culturally relevant books. The Journal of Adolescent and Adult Literacy, 44 (7), 626-633.

Pakenham, K.J. (1983). Developing expectations for text in adult beginning ESL readers. Toronto, Canada: On TESOL Õ83. (ERIC Document Reproduction Service No. 275148)

Petrimoulx, J. (1988). Sustained silent reading in an ESL class: A study. Chicago, IL: TESOL. (ERIC Document Reproduction No. 301 068).

Qian, D.D. (1999). Assessing the roles of depth and breadth of vocabulary knowledge in reading comprehension. The Canadian Modern Language Journal, 56, 262-305.

Robson, B. (1982). Hmong literacy formal education, and their effects on performance in an ESL class. In B. T. Downing & D. P. Olney, (Eds.), The Hmong in the west: Observations and reports (pp. 201- 225). Minneapolis, MN: University of Minnesota.

Rumelhart, D.E. (1985). Toward an interactive model of reading. In H. Singer & R.B. Ruddell (Eds.) Theoretical models and processes of reading (pp. 722-750). Newark, DE: International Reading Association.

Ryan, A. & Meara, P. (1991). The case of the invisible vowels: Arabic speakers reading English words. Reading in a Foreign Language, 7(2), 531-540.

Shanahan, T., Mulhern, M., & Rodriguez-Brown, F. (1995). Project FLAME: Lessons learned from a family literacy program for linguistic minority families. The Reading Teacher, 48, 586-93.

Strucker, J. (1997) What silent reading tests alone canÕt tell you. Focus on Basics, 1 (B), 13-17.

Strucker, J. & Davidson, R. (in press) Adult reading components study (ARCS). Boston: National Center for the Study of Adult Learning.

Tan, A., Moore, D.W., Dixon, R.S., Nicholson, T. (1994). Effects of training in rapid decoding on the reading comprehension of adult ESL learners. Journal of Behavioral Education, 4, 177-189.

Thuy, V.G. (1992). High-tech for effective ESL/family literacy instruction. Harrisburg, VA: Pennsylvania State Department of Education. (ERIC Document Reproduction Service No. ED 356 683).

Tse, L. (1996a). When an ESL adult becomes a reader. Reading Horizons, 37, 16-29.

Tse, L. (1996b). If you lead horses to water they will drink: Introducing second language adults to books in English. California Reader, 29 (2), 14-17.

Wade-Woolley, L. (1999). First language influences on second language word reading: All roads lead to Rome. Language Learning, 49(3), 447-471.

Wilson, M. (1983). Developmental patterns of reading proficiency in adult ESL students: Implications for ESL classrooms. Toronto, Canada: TESOL. (ERIC Document Reproduction Service No. 248 709)

Zimmerman, C. B. (1997). Do reading and interactive vocabulary instruction make a difference? An empirical study. TESOL Quarterly, 3(1), 121-140.

**Carrell, P.L. (1989). Metacognitive awareness and second language reading. Modern Language Journal, 73 (2), 121-134.**

**Carrell, P.L., Pharis, B.G., & Liberto, J.C. (1989). Metacognitive strategy training for ESL reading. TESOL Quarterly, 23 (4), 647-678.**

**Cotterall, S. (1991). Meiko: Case study of a second language reader. Reading in a Foreign Language, 8(1), 611- 627.**

**Hilferty, A.G. (1996). Coding decoding: Predicting the reading comprehension of Latino adults learning English. Unpublished doctoral dissertation, Harvard University, Cambridge, MA.**

**Hsueh-chao, M.H. & Nation, P. (2000). Unknown vocabulary density and reading comprehension. Reading in a Foreign Language, 13(1), 403-430.**

**Ryan, A. & Meara, P. (1991). The case of the invisible vowels: Arabic speakers reading English words. Reading in a Foreign Language, 7(2), 531-540.**

**Wade-Woolley, L. (1999). First language influences on second language word reading: All roads lead to Rome. Language Learning, 49(3), 447-471.**

**Zimmerman, C. B. (1997). Do reading and interactive vocabulary instruction make a difference? An empirical study. TESOL Quarterly, 3(1), 121-140.**