- l w y'1041" v .yv' w , M . lllllllllHUN/IIIIIHIIHIHIIll/Ill!ll/llil/lfll/lllllll 3 1293 10499 8541 This is to certify that the dissertation entitled Learning Style Preferences of ESL Students presented by Lorna LaVerne has been accepted towards fulfillment of the requirements for Ph.D. degree in English 1 .. ! flat/1 @/ AM'L'.:c«// . / . " Major professor Date August 12, 1981 MS U i: an Affirmative Action/Eq ual Opportunity Institution 0- 12771 ‘ MSU LlBRARlES ”3—- RETURNING MATERIAL§3 FTace in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. Vb- ? r m. ’3' 4‘2}; '2. 1. 2 m: , W .‘ UUN be? i 3.; Le 33 1 n ‘ \‘ D j ' ’ f I?!“ thatiit MflN 1 3 2009 grmf513:3859 LEARNING STYLE PREFERENCES OF ESL STUDENTS BY Lorna LaVerne A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of English 1981 ABSTRACT LEARNING STYLE PREFERENCES OF ESL STUDENTS BY Lorna LaVerne This study was designed to measure specific learning strategies of adult students of English as a second language (ESL) along the dimension of analytic or synthetic learning style. The purpose was to determine whether a relationship exists between learning style and the following variables: degree of proficiency in English, amount of progress being made, language/cultural background, age, sex, roommate's language, study plans, and the length of time the student had been studying at the English Language Center (ELC)- The Learning Preference Opinionnaire, an instrument designed for this study, listed twenty-two specific language learning strategies which reflected either analysis or Syn- thesis. The opinionnaire was presented to the 156 BLC stu- dents on all proficiency 1evels,aldng withzademographic data sheet. A measure of each student's English proficiency was also obtained. Data were gathered from a smaller group of eleven subjects by conducting three separate interviews with each student. Analysis of the large group data showed the following: (1) the reliability of the Learning Preference Opinionnaire was .48 overall; .43 for the analysis items; and .69 for the synthesis items; (2) there was a moderate correlation (r=.36) between learning style and level of proficiency; (3) there was no relationship between learning style and language/cultural background; (4) there were too few cases to allow for any testing of the relationship between learn- ing style and prior experience in foreign language learning; (5) there was no relationship between learning style and age, time in the intensive program, progress, sex, study plans, or roommate's language. The small group interviews tended to reflect these results and suggested that the following factors may have an influence on learning style preference: (1) previous contact with speakers of the target language; (2) exposure to the language as a means of communication; (3) the purpose for which the language is being learned; and (4) the context, formal or informal, for both the task and the learning process. Further research is needed, both to refine the Learning Preference Opinionnaire and to try to explain variances not accounted for by learning style, perhaps by means of tests for such factors as aptitude, attitude and motivation, and other dimensions of cognitive style. ACKNOWLEDGEMENTS Many people have contributed in one way or another to the development and completion of this dissertation, but the lion's share of the credit must go to Dr. Paul Munsell, my major professbr, without whose constant help and encouragement none of the accomplishments of the past year would have been possible. I am also grateful to the other members of my guidance committee, Dr. James Stalker, Dr. Stephen Judy and Dr. John Alford for their support and advice throughout my years of graduate study. Special thanks go to Ms. Gabriella Belli for guiding me through the maze of statistics, computers, and SPSS programs. A final note of appreciation must go to the Office of Educational Resources at the College of Osteopathic Medicine at MSU, for their patience with my many absences from work during the data-gathering portion of the study, and to all those who gave their support and perhaps are not mentioned formally here, but to whom I will always be grateful. ii TABLE OF CONTENTS CHAPTER I. REVIEW OF LITERATURE AND PURPOSE OF THE STUDY A. Introduction...................... ...... .. ........ l B. Major Variables................. ........ . ......... 3 l. Affective Variables... ........................ 3 2. Cognitive and Learning Styles. ................ 4 3. Learning Environment .................. . ....... 6 C. Learning Strategies........... .................... 8 D. Self-Reports of Learning Strategies .............. 11 E. Areas Needing Research. .............. . ........... 13 1. Learning Stage ........ . ............... . ....... l4 2. Cultural Differences ......................... l4 3. Foreign Language Learning Experience ......... 15 F. Special Considerations Regarding Learning Strategies ...................................... 15 G. Purpose of the Study... ....................... 17 H. Hypotheses ...... . . ............................. 18 CHAPTER II. METHODOLOGY A. Introduction and Overview ........................ 20 B. Definitions of Terms............... .............. 21 C. General Procedures............. ........... . ...... 26 D. Instrumentation .................................. 27 1. Large Group Instruments ...................... 27 2. Small Group Instruments ...................... 30 E. Selection of Subjects ............................ 31 F. Procedures for Administration of Instruments ..... 33 1. Large Group Procedures ....................... 33 2. Small Group Procedures ....................... 34 G. Analysis of Data ....... .. .......... ............. 35 1. Special Scoring Procedures ................... 36 2. Statistical Treatment of the Large Group Data ......................................... 37 3. Analysis of the Small Group Data. ............ 38 H. Summary ............... ........ ................... 39 CHAPTER III. ANALYSIS OF LARGE GROUP DATA A. Introduction and Overview. .... .................. 40 B. Results of the Administration of the 0pinionnaire............ ......... ........ .....40 l. Scoring of the Learning Preference Opinionnaire.. ... ..........................40 2. Reliability of the Learning Preference Opinionnaire.. ..............................42 iii C. Results of the Statistical Studies.............. .44 D. Summary..........................................53 CHAPTER IV. RESULTS OF SMALL GROUP INTERVIEWS A. Introduction......................... ........... .55 B. Results of Interviews.......... ........... . ...... 56 C. Interview Data............................ ....... 59 1. Language Background.......... ............... 59 2. Type of Language Input and Native Speaker Contact. ..... . ........... . ................ 61 3. Media Influence ....... ’ ..................... 64 4. Self- Evaluations of Language Learning Ability .................. . ..... . ........... 65 5. Approaches to Language Learning... ........... 70 6. Specific Learning Strategies ................. 73 7. Teacher Evaluations .......................... 74 8. The Reading Task. ....... ................ 75 D. Conclusions ........... ..... ............ 76 E. Relationship of Small Group and. Large Group Data .......... .a. ..... ................. .......... 77 CHAPTER V. SUMMARY, DISCUSSION AND IMPLICATIONS A. Introduction ........ . ..... . ...................... 80 B. Summary of Results ..... .. .... ................... 80 C. Discussion of Limitations. .......... ...... .81 1. Reliability of the Learning Preference Opinionnaire .............. ..... . ........... .82 2. Context of the Learning Situation ............ 84 3. Makeup of the ELC Population. ....... ......... 85 D. Implications for Teachers and Learners. . ..... ..85 E. Suggestions for Further Research. ......... 86 F. A Possible Analysis/Synthesis Model for Language Learning ............. . .................. 87 APPENDIX A. Learning Preference Opinionnaire ............. 89 APPENDIX B Language Background Questionnaire ............ 92 APPENDIX C Reading Task Checklist... .................... 95 APPENDIX D. Teacher Evaluation Checklist... ........... ...96 APPENDIX E ANOVA Tables................. ........ . ....... 97 BIBLIOGRAPHY... ........... ............. .................. 100 iv LIST OF TABLES Table Page 1. Analysis/Synthesis Average (ASAVE).. ............... 41 2. Correlation of ASAVE with Proficiency, Age, Time at the ELC, and Progress (Whole Group) ............. 46 3. Correlation of ASAVE and Proficiency According to Language/Culture Populations..... ........ . ........ .47 4. Correlation of ASAVE and Proficiency According to Sex, Roommate and Study ............................ 50 5. Demographic Data on Interviewees........ ........... 57 6. Rankings of Interviewees 6n Proficiency and ASAVE..58 7. ’Data.on Interviewees for All Variables ............. 60 8. Overall Learning Style Scores of Interviewees. ..... 78 9. Analysis of Covariance: ASAVE by Culture with Proficiency ...... . ................. . . ............ 97 10. Analysis of Variance of Proficiency by Culture Groups ................. . ........................... 97 11. Analysis of Variance of ASAVE by Culture Groups....98 12. Analysis of Covariance: ASAVE by Sex with Proficiency ..... .................. ...... . .. ....... 98 13. Analysis of Covariance: ASAVE by Roommate with Proficiency........ ..... . ..... .......... ........... 99 14. Analysis of Covariance: ASAVE by Study Plans with Proficiency........................ ........ ...99 LIST OF FIGURES Figure Page 1. Interaction of Learner and Environment .............. 8 2. Subjects' Self-Evaluations of English Ability ...... 68 3. A3's Self-Ranking of English and French Ability....70 vi CHAPTER I REVIEW OF LITERATURE AND PURPOSE OF THE STUDY A. Introduction In 1973 a book was published which, along with many others, was aimed at language teachers; but unlike the others, it did not attempt to instruct the reader on the "best" language teaching methods or offer easy solutions to teaching problems. Its title was Focus on the Learner-~and, as a sign of the coming shift in research attention, it em- phasized language learners, not teachers, and it looked at their capacities, attitudes and learning strategies, as well as at what they learn (Oller and Richards 1973, vi). Most of the previous research on foreign language learning had tended to focus on methodology. Carol Hosen- feld pointed out in 1979 that "a significant portion of the history of research in language teaching may be character- ized as a search for instructional techniques that would lead to skill improvement" (Hosenfeld 1979, 51). This attitude has changed significantly. In the past few years, the professional literature on second language learning has expanded its concern with teaching methods to l include concern with the learner, emphasizing that the learner plays an active role in language acquisition and language learning. As Gardner and Lambert (1972) ask, "How is it that some people can learn a foreign language quickly and expertly while others, given the same opportunities to learn, are utter failures?" (p. 1). It is probable that methodology, which is uniform for these learners, is less significant in answering this question than the individual differences among the students; in other words, the answer lies not in methodology but in the individual learner. One major influence in this concern with the learner has undoubtedly been the contributions of applied linguists such as Roger Brown, Jean Berke-Gleason and many others in the field of child language acquisition, Heidi Dulay and Marina Burt in second language acquisition, and Kenneth Goodman, Frank Smith and others in their research on reading. Their confirmation that the learner plays a very active and creative role in the learning process has forced researchers to ask what precisely the learner does to learn a language. Considerable recent research has also focused on the adult foreign language learner. After John and Francine Schumann kept journals of their language learning experi- ences, they concluded, "Our profession spends a good deal of time in teacher preparation, teacher training and teacher education. Perhaps second language learning might be improved by investing some time in learner preparation, learner training and learner education" (Schumann 3 and Schumann 1977, 249). Carol Hosenfeld agrees, and says that the essential question now is not "What must the teacher do to produce a given level of proficiency?" but rather "What must the student do to attain a given level of proficiency?" (Hosenfeld 1979, 52). B. Major Variables As pointed out in the introduction to this chapter, when different students are exposed to the same teaching methodology, some succeed while others do not. In attempt- ing to determine what learner-related variables may be responsible for this, researchers have tended to focus on the following three areas: (1) affective variables; (2) cognitive and learning styles; and (3) learning envi- ronment. l. Affective Variables One school of thought, represented by Schumann (1975), Brown (1973) and others, made the assumption that differences in learner performance would most likely be accounted for by attitude, motivation, emotion, and other personality or psychological variables. As Len Sperry puts it: "...the focus has been on the learner qua person. When confronted with an intellectually capable learner whose performance failed to measure up to his supposed potential, psychologists and educators have tended to attribute this failure to an emotional block, a personality conflict or to social class factors" (Sperry 1972, 2). Robert Gardner and Wallace Lambert believe that atti- tudes and motivation play an important role in the variations seen in second language learning. They point out that while factors such as aptitude and intelligence are relatively stablepredictors of success, the importance of attitudinal measures is variable, depending on school district and social class. Their studies with English-speaking students learning FrenCh in Montreal have suggested that students with an "in- tegrative" orientation are more successful in second language learning than those who are "instrumentally" oriented (Gard- ner and Lambert 1972, 4—5). Joan Rubin and H.H. Stern also see affective variables as contributing to success in second language learning. Rubin says: "The good language learner is often not inhib- ited. He is willing to appear foolish if reasonable communi- cation results" (Rubin 1975, 47). Stern agrees, and lists "a tolerant and outgoing approach to the target language and empathy with its speakers" among the features that he con- siders to mark good language learning (Stern 1975, 312). 2. Cognitive and Learning Styles Other past research has focused on such cognitive fac- tors as aptitude and general intelligence in attempting to predict success in foreign language learning (HanCock 1975). The main instruments used to measure foreign language apti— tude for the past twenty years have been the Modern Language Aptitude Test (Carroll 1959) and the Pimsleur Language Apti- tude Battery (Pimsleur 1966). More recently, other researchers have explored aspects of what has usually been referred to as cognitive or learn- ing style. H. Douglas Brown refers to it as "a rather amor- phous link between personality and cognition" (Brown 1980, 89). Kagan, Moss and Siegel see it as a set of "stable in- dividual preferences in modes of perceptual organization and conceptual categorization of the external environment" (Ka- gan, Moss and Siegel 1963; in sperry 1972, 4). The term has almost as many definitions as researchers; most agree, how- ever, that it includes conceptual and perceptual tendencies that are both individual and consistent over time. Because of inconsistencies in the literature in the use of the terms cognitive style and learning style, this study has chosen the term "learning style" and will use it consistently to refer to both cognitive and learning styles. The term "learning strategies," as an interaction of learning style and learning environment, will be discussed later. Dozens of different learning styles have been identified by educatdrs and psychologists. Brown (1980) discusses several of these, including field dependence/independence; reflectivity/impulsivity; tolerance/intolerance of ambiguity; and broad/narrow category width. These four learning styles in particular have been er plored with regard to second language learning. In the Good Language Learner study, Naiman, gt 31. (1978) discovered the following: (1) field independence correlated positively and significantly with language "success" in the classroom; (2) tolerance of ambiguity was the other significant predictor of success in foreign language learning; and (3) there was no relationship between category width and success in French as a second language. In addition, a study by Doron in 1973 found that in reading, reflective students of English as a second language were slower, but more accurate than impulsive students (Brown 1980, 94). Some interesting work has been done by Gloria Kuchinskas (1979), who discusses an important variable neglected by other researchers--the teacher's cognitive style. According to Kuchinskas, this influences the learning environment more than any other factor. Stern (1975) suggests that while good learners can adapt to almost any condition, poor learners are hard-hit by teaching methods inappropriate for them. 3. Learning Environment Some of the most recent work on language learning has focused on the relationship between the learner and his/her environment. Bialystok (1979) distinguishes between char- acteristics of the learner (language learning aptitude, attitude, motivation, personality variables) and characteristics of the learning situation_(length of exposure to the language, teaching method employed). She states that variations in the rate of learning are greatly affected by the learner's ability to adapt efficiently to the learning situation. In a more general study of learning, Dunn and Dunn (1978) found four types of variables that affect both chil- dren and adult learners. Two of these describe character- istics of the learner: the learner's own emotionality (motivation, persistence, responsibility, and need for structure or flexibility) and the learner's sociolggical preferences (working alone, in a pair, with peers or adults, or in various patterns). The other two variables describe characteristics of the learning situation: immediate en- vironment (sound, light, temperature and physiCal design) and physical needs (perceptual strengths, intake, time of day and mobility). Strevens (1977) also lists physical factors, such as overcrowding and lack of books, among the many possible factors responsible for failure in language learning. Some of Stephen Krashen's work (1976) has focused on whether the linguistic environment is formal (classroom instruction) or informal (out-of—class contact with native speakers). He takes the point of view that formal and in- formal environments contribute to different aspects of second language competence, i.e., informal environments affect acquired (or "natural") competence, whereas formal environments affect learned competence of second language learners. C. Learning Strategies So far we have examined three major variables which appear to influence the language learning process: (1) affective variables; (2) cognitive and learning styles; and (3) learning environment. The first two of these are in- herent in the learner, who then must adapt to the third variable, the environment, by using various strategies. It may well be that language learning sucCess depends on the ability of the learner to make this adjustment. The model below shows the interaction of the learner and the learning environment; strategies are both consciously and unconsciously originated by the learner to cope with the environment. Strategies ’T N Learner Learning Environment R J v Linguistic and Sociocultural Input Figure 1. Interaction of Learner and Environment Like the term "learning style," the term "learning strategies" has a number of different definitions. Joan Rubin calls learning strategies "the techniques or devices which a learner may use to acquire knowledge " (Rubin 1975, 43). Bialystok describes them as "the optional methods for exploiting available information to increase the proficiency of second language learning" (Bialystok 1978, 76). Bialystok and Frolich (1978) have tested the effects of aptitude, field independence, attitude and strategy use on achievement in language learning. They found that apti- tude and field independence were related, as were strategy use and attitude, but that the two factors which affected achievement were aptitude and strategy use. Bialystok (1978) believes that the possible learning processes and strategies are the same for all learners, but that it is the efficiency with which they operate that accounts for differences between individual language learners. Strategy use as a major factor in language learning success became an important research question after the publication of an article by Joan Rubin (1975) on "What the 'Good Language Learner' Can Teach Us." After extensive classroom observation and interviewing, she came up with a list of strategies which characterize the "good language learner," who: (1) is a good guesser; (2) has a drive to communicate; (3) is not inhibited; (4) attends to form; (5) practices; (6) monitors his own and others' speech; (7) attends to meaning; and (8) isolates cues and features which give maximum intelligibility. H.H. Stern (1975) pro- poses a similar list of strategies, on the basis of which he contrasts the good and the poor language learner. 10 The most extensive research to date on the good language learner is the work of Naiman, Frolich, Stern and Todesco (1978). Their subjects were English speakers learning French as a second language in Canada; all were given various tests in cognitive or learning style, language aptitude, personal- ity, and so on, and several who had been identified as either "good" or "poor" language learners were interviewed in depth. The Good Language Learner study sought to identify "(1) the strategies and techniques the learner consciously develops and employs, and (2) certain learner characteris- tics, in particular personality and cognitive style factors, which are likely to influence the use of strategies and techniques and thereby, indirectly, learning outcome" (Nai- man, e£_al. 1978, 4). The authors modified a list of strat- egies suggested by Stern (1975) and analyzed their interview data according to five major strategies: (1) an active ap- proach to the language learning task; (2) realization of language as a system; (3) realization of language as a means of communication and interaction; (4) management of affective demands; and (5) monitoring of L2 performance. The researchers' hope was to identify differences be- tween good and poor language learners in order to develop ways for the poor learner to overcome his/her difficulties. Many of the other factors which affect language learning tend not to vary; aptitude and field dependence/independence, for example, are inherent in the cognitive makeup of the learner, but they generally cannot be taught or modified 11 (Bialystok 1978; Gardner and Lambert 1972). It is possible, however, that strategies are teachable or variable to some extent (Brown 1980, 93). Rubin (1975) says that by considering the strategies of good language learners, we may gain information on lin- guistic processes; ultimately, this information may form the basis for training other language learners to use suc- cessful strategies. Hosenfeld (1979) and Stern (1975) also see this as an important implication of "good language learner" studies. D. Self-Reports of Learning Strategies One technique used to obtain data on students' learning strategies has been to ask them to report on these strate- gies in an interview. Questions have been raised as to the accuracy of the information gained from such self-reports of learning. Naiman, gt al. used self—reports successfully in the Good Language Learner study. Evidence that the learner $223 consciously monitor his/her language learning strate- gies has been provided by Krashen (1976 and 1977) and Bialystok (1978 and 1979), and both Rubin and Stern have in- cluded "monitoring" on their lists of strategies used by the good language learner. .Furthermore, some research has suggested that the learner's own perceptions of how he/she learns can be quite accurate, and that the learner is often able to predict 12 his/her own success in various types of learning. Farr (1971) showed that learners are able to predict the modali- ties (oral vs. written) in which they learn and achieve best. Domino (1970) grouped students in accordance with their perceptions of how they learned, and found that those who learned as they preferred scored higher on controlled tests. Finally, there is the undeniable link between one's perceptions and their effect on one's emotions and attitudes. In a sense, this is a variation of the well-known “placebo effect" in medicine; patients who are given a placebo, but who believe they have been given a healing medicine, have often reported "miraculous" improvements in their condition. In language learning as well, the learner's perceptions of a situation may be as influential as the situation itself. For example, William Acton (1979) has provided evidence that perceived social distance may be a more important factor in language learning than actual social distance. In addition, both Francine and John Schumann (1977) reported that many of their language learning difficulties arose because of conflicts between what they perceived as good language learning techniques and what the teacher per- ceived as good ones. They refer to these perceptions and reactions to both the language class and the target culture as "personal variables" that affect second language acqui- sition and learning, and which can account for individual differences. 13 E. Areas Needing Research The literature cited thus far has significantly in- creased our knowledge of the foreign language learning pro- cess. Various researchers have stressed the importance of affective variables, cognitive and learning styles, the learning environment, and the learner's strategies. A key question being asked is whether a student's strategies will vary, and under what circumstances. Rubin suggests that learner strategies will vary with: (l) the task; (2) the learning stage; (3) the age of the learner; (4) the context; (5) individual styles; and (6) cultural differences in cognitive learning styles (Rubin 1975, 43). A similar suggestion is made by the Good Language Learner researchers: "Different classes of variables, identified in the present study--cognitive style, attitude, learning strategy--may be of varying importance at different stages of language learning and in different learning environments" (Naiman, gg‘gl. 1978, 101). On the basis of these and similar comments in the literature, three variables have been chosen which seem to be in particular need of further research. These are: (1) learning stage, as described above by Naiman e; 31. and by Rubin; (2) cultural differences, suggested above by Rubin; and (3) previous foreign language learning experience, mentioned by Naiman, 33 £1. and by Rubin. 14 1. Learning Stage As Rubin has pointed out, "Language learners may in fact use different strategies at different points in time in the learning process" (1975, 48-9). Stern agrees, and adds, "Assuming the list identifies mainly the good learner at elementary and perhaps intermediate stages, is the ad- vanced learner adequately covered by these strategies?" A similar point of view, that strategies are of varying impor- tance at different stages of language learning, was stated by Naiman, §£_gl.and was mentioned previously. 2. Cultural Differences Another variable needing research, which to my know- ledge has not been described empirically, is what Rubin has called "cultural differences in learning styles." She notes that "in some societies, listening until the entire code is absorbed and one can speak perfectly is a reported form of learning; in others successive approximation to native speech is used as a learning strategy; while in still others rote learning is the most common learning strategy" (Rubin 1975, 49). 15 3. Foreign Language Learning Experience Foreign language learning experience, as suggested by Rubin (1975, 49), appears to be another important factor in language learning. It is noteworthy that all of the good language learners presented in case studies by Naiman, 33 31. had already had extensive foreign language learning experi- ence. It is possible that their learning styles 3213 af- fected by this previous experience, including the number and types of languages learned and the amount of time spent studying them. F. Special Considerations Regarding Learning Strategies As the learner tries to adapt to different learning environments, his/her strategies are likely to be affected by how he/she resolves two of what Stern has called "major problems of learning." One of these problems, according to Stern, is the "code-communication dilemma"; good language learners will at times have to attend to the code (formal language) and at other times to communication (functional language), whereas the poor learner avoids both. Another problem involves "the choice between rational and intuitive learning . . . whether the language learner should treat the language learning task intellectually, conceptually, and systematically as a mental problem, or whether he should avoid thinking about the language and absorb the language 16 more intuitively" (Stern 1975, 310mll). In the Good Language Learner study, 68 percent of the 34 interviewees considered their language learning a highly conscious and systematic process, while 26 percent felt it comprised both conscious and unconscious elements. The two interviewees who saw language learning as a totally intuitive process were the unsuccessful learners (Naiman, e; 31. 1978, 11). As Krashen (1976) has pointed out, formal and informal learning environments contribute to second language profi- ciency in different ways. This, combined with Stern's pre- occupation with the code-communication dilemma, parallels a general sentiment both in and out of the field of language learning that a major distinction among learners is in the dimensions of rational processes vs. intuitive processes, or deductive vs. inductive reasoning, or analytic vs. synthetic approaches to learning. It appears useful to limit the_dimensions of this study to include learning strategies along a continuum of analysis (mainly conscious; concentration on code; language seen as a system) and synthesis (mainly intuitive; concen- tration on the whole meaning; language seen as a means of communication). Undoubtedly there are many other dimensions that fall along such a continuum, but the analysis/synthesis distinction appears to be one which captures a major source Of the variability in language learning strategies. 17 G. Purpose of the Study Recent research has indicated that a focus on students' learning styles and learning strategies-~including their own feelings and beliefs about what helps them learn-~can give useful and valid insights into both the learning process in general and the foreign language learning process in partic- ular. Research needs which have been identified include: (1) learning stage (Rubin; Stern; Naiman, 33 21.); (2) learn- ing strategies (Bialystok; Rubin; Stern; Naiman, e: 31.); (3) cultural differences in learning styles (Rubin); and (4) solving the dilemmas of code vs. communication and rational processes vs. intuitive processes (Stern; Naiman, E£.El-)° To date, however, no studies are available which deal with the learning styles of students of English as a second language (ESL) and the specific, conscious learning strate- gies which exemplify these styles. Research on the learning strategies of the second language learner has been carried out primarily with native English speakers learning French (Bialystok; Hosenfeld; Naiman, e£_§l.). Krashen's work (1976) on formal and informal environments presents data on ESL learners, but not on the strategies they might use to cope with these different environments. We may be able to take the work which has been done so far on learners of French and apply it to learners of other foreign languages. Nevertheless, there is a need to 18 accumulate separate data on the strategies of ESL learners from various countries and language groups. The purpose of this dissertation is to play a part in fulfilling this need by exploring the ways in which ESL students in the United States perceive various study strategies as valuable in moving them toward their goal of learning English. These strategies are classified in terms of the analy- sis/synthesis continuum mentioned earlier, and the resulting composite of strategies is considered to be learning gpyle. It is this learning style around which the hypotheses for this study are structured. H. Hypotheses The following are the hypotheses being tested by this study: ' Hl--Learning styles of students of English as a second language, as measured by an instrument designed for this study, will vary along with differences in English proficiency. H2--Learning styles of ESL students, as measured by the same instrument, will vary according to language or cultural background. H3--Learning styles of ESL students, as measured by this instrument, will vary along with dif- ferences in previous foreign language learn- ing experience. H4--Learning styles of ESL students, as measured by this instrument, will vary along with age, amount of time spent at the English Language Center, and progress made in a five-week period (numerical variables). 19 HS--Learning styles of ESL students, as measured by this instrument, will vary along with dif- ferences in sex, roommate, and program of study (categorical variables). H6--Learning style variations among ESL students, as measured by the instrument designed for this study, will be reflected in case study data collected from a small group of students, using interviews, a miniature learning expe- rience and ratings by teachers. CHAPTER II METHODOLOGY A. Introduction and Overview The previous chapter identified three areas in the field of foreign language learning which appear to need fur- ther research. These are: (1) learning stage; (2) cultural differences in learning styles; and (3) previous foreign language learning experience. The first three hypotheses of this study, listed at the end of that chapter, stated that. the learning styles of ESL students, as measured by an in- strument designed for this study, will vary along with dif- ferences in the three variables listed above. Two additional hypotheses stated that learning styles would vary along with differences in age, amount of time at the English Language Center, and progress (H4); and with differences in sex, roommate, and study plans (H5). The last hypothesis (H6) stated that variations in learning styles would be reflected in case study data collected from a small group of students. To measure learning style, I developed an instrument asking students to indicate preferences for specific learn- ing strategies. This instrument, called the Learning 20 21 Preference Opinionnaire (LPO), was given to a group of 156 students at Michigan State University's English Language Center at about the same time that they were given a pro- ficiency test battery. In addition, to provide background information and to attempt to validate the Learning Preference Opinion- naire, in-depth information was obtained from eleven stu- dents selected to represent three proficiency levels and three language groups. B. Definitions of Terms The following are definitions of the major terms used as they are understood in the context of this study. Learning Strategy refers to the learner's attempt to cope with the learning environment. "Strategy" has been defined by Ellen Bialystok (1978, 71) as one of many "con- scious attempts made by language learners to improve pro- ficiency by a variety of means." In this study, the term "learning strategies" refers to the specific items on the Learning Preference Opinionnaire (e.g., "Memorizing gram- mar rules"). Learning Spyle is a composite of learning strategies. Although specific learning strategies may vary among indi- viduals, they can be classified in terms of the dimensions of learning they represent; in this case, the dimensions are considered to lie on a continuum with analysis at one 22 end and synthesis at the other. The composite of a learn- er's strategies, in terms of these dimensions of learning, is considered to be that individual's learning EEXEE: In this study, learning style is expressed numerically by the Analysis/Synthesis Average, whiCh will be referred to in this study as ASAVE. Analysis means acting on language input by "taking it apart," looking at it systematically. Analysis may be as- sociated with frequent monitoring of one's language output, and is characteristic of the conscious language learning by many adults. In this study, analysis is expressed numeri- cally by a low score (ASAVE) on the Learning Preference Opinionnaire. Synthesis means acting on language input by dealing with it holistically, ”putting it all together." A learn- er who has internalized a language, or a native speaker of that language, is more likely to say that the language "sounds" or "feels" right rather than make that decision on the basis of a rule. Synthesis may be associated with lower use of monitoring (accompanied by reasonable accuracy) and is characteristic of language acquisition, in the un- conscious, "natural" way described by Krashen. Synthesis is expressed numerically in this study by a high score (ASAVE) on the Learning Preference Opinionnaire. ESL refers to English as a second language. In this study, ESL students are those who are studying English as a second (i.e., foreign) language in an intensive program 23 in the United States, specifically at Michigan State Uni- versity's English Language Center. ELC refers to the English Language Center at Michigan State University, which offers intensive programs in English to adults-~both those enrolled at Michigan State University and those studying only the English language. The ELC of- fers classes in grammar, Speaking and listening, reading, composition, and language laboratory at various levels of proficiency. Proficiency refers to an ESL student's performance on a battery of tests developed and administered by Michigan State University, indicating whether he/she is a beginning level student or more advanced. In this study, proficiency is expressed numerically by the learner's position on a continuum from 0 to 100, with 100 representing native speaker proficiency. The test covers grammar, vocabulary, aural comprehension, reading comprehension, and composition. Lapguage/Cultural Background refers to a learner's membership in a group whose members share the same native language or country. There are times when such groupings reflect subjective decisions; for example, Spaniards and Mexicans might be grouped together on the basis of language although their cultures are quite different, and Mexicans and Brazilians might be grouped together on the basis of geography although they differ in both language and culture. The question of how to group people for purposes of data analysis was carefully considered in this study 24 and was resolved by running separate data analyses only for certain groups. The groups of interest in this study are speakers of Spanish, Arabic, and Japanese. Besides representing very different languages and cultures, these are the largest groups at the English Language Center, each containing 29 or more students, and are typical of such groups enrolled in intensive English programs. The Spanish-speaking stu- dents are from several different Western hemisphere areas,including Mexico, the Caribbean, and South and Cen- tral America, with similar cultural patterns. The Arabic speakers are all from countries in the Middle East and speak Arabic as their native language. The Japanese speakers are only from Japan, and with 29 members, the Japanese group represents the largest number of ELC stu- dents from any single country. These three groups were considered together with all others in analyses done on the group as a whole. Students who did not speak Arabic, Spanish or Japanese were consid~ ered only in analyses of the total population, and not separately, since there were not enough students from any one country to allow for any meaningful generalizations. Thus, in this study, "language/cultural background" will refer specifically only to Arabic, Spanish or Japanese speakers. Foreign Language LearningExperience refers to fluency or near-fluency of ELC students infa language other than 25 their native language or English. Time at the ELC is expressed specifically by the num- ber of quarters the student has spent at the English Lan- ‘ guage Center at Michigan State University. Students were also asked how many years they had studied English in their home countries, but since that is often represented by two classes of English per week in high school, with the teacher speaking primarily in the native language, I felt that using the amount of time spent at the ELC as a "standard measure- ment" was the only alternative to the sometimes meaningless measure of "years of study." Progress refers to the increase or decrease in ELC students' scores between the test given at the beginning of the term and the midterm exam given five weeks later. Formal Setting refers to an in-class ESL setting, generally with the teacher exerting a large degree of con- trol over the learning activities. While this study deals with both strategies of analysis and of synthesis, and while the questionnaire itself, for convenience, was ad- ministered in a formal setting (a classroom), an important premise of the study is that a student's learning style may vary depending on whether the setting is formal or informal. Informal Setting refers to an out-of-class setting, in which an ESL student may pursue his/her language learning activities in any way he/she chooses, without the control or guidance of the teacher, and in a largely self-directed man- ner. In this study, students are asked to answer the 26 questions on the Learning Preference Opinionnaire while as- suming that they would be pursuing their learning activities outside of class. C. General Procedures In order to test the hypotheses of this study, the following general procedureswere used: (1) A large group (N=156) of adult learners of ESL was selected from all levels of proficiency at the English Language Center. A small group of learners (N=ll) was selected to represent three language/cultural groups and three levels of proficiency. (2) The Learning Preference Opinionnaire was devel- oped, then administered to the whole group; the results were computed for each student. Demographic information was also gathered along with the Learning Preference Opinion- naire. The LPO was administered to the large group just before the students took their midterm examination, in order to have an estimate of their language proficiency at about the same time as the measurement of their prefer- ences for learning strategies. The proficiency measure used was a battery of tests developed by the English Lan- guage Center and included sections on grammar, vocabulary, aural comprehension, reading comprehension, and composition. (3) In addition to being given the above instruments, the small group subjects participated in an interview 27 based on the Adult Interview Questionnaire from the Good Language Learner study (Naiman, 33 El. 1978, 106-9),which gathered information on their language learning backgrounds. They were also given a miniature lesson that required them to read a passage and comment on it. In addition, an attempt was made to gather information on each student's learning be- havior by interviewing at least two of his/her teachers. D. Instrumentation The following are descriptions of: (l) instruments used in the large group; and (Z) instruments used in the small group. All but the proficiency test are in the appendices. 1. Large Group Instruments The Learnipg Preference Opinidnnaire was developed for this study in order to measure a student's learning style in terms of his/her average on a continuum of analysis or syn- thesis. There were four stages in its development: (1) A pilot form of the instrument was used with 68 ELC students who were not in the main study. (2) The items which appeared to offer a spread of re- sponses were then selected, and additional items were added to the list. (3) This list was then offered to the professional staff of the English Language Center with the request that 28 they comment on the items or edit them. (4) The revised list was then re-edited by faculty members of the English Language Center and the English Department. Throughout the development process, the analysis/ synthesis scale was seen as the integral dimension along which the responses would probably range. Therefore, half of the items were structured so that a positive response would reflect the analysis dimension, and the other half so that a positive response would reflect the synthesisdimension. Thus there would not be a bias toward one type of response or the other, and both types of learning would be equally represented. All items, however, were envisioned as being part of the same total scale, and the only difference between the analysis and synthesis items was considered to be the di- rection in which the response lay, not in the content of the item; on a five-point scale, "1" was considered a posi- tive response and "5" was considered a negative response for all items.‘ For example, if a student made a negative response on an item such as "Reading to get the general meaning," which reflects the synthesis dimension, that nega- tive response would indicate a positive analytical strategy. In other words, even though an item may reflect one dimen- sion or the other, it would still give an analysis/synthesis IEEBE: that is, a negative response on one dimension would be equivalent to a positive response on the other. 29 Reliability of the LPO was unknown at the time of its administration. A later reliability analysis using Cron- bach's Alpha reliability coefficients yielded coefficients of .48 for the analysis items, .69 for the synthesis items, and an overall coefficient of .49. The significance of these figures will be discussed further in the chapter on analysis of the large group data. The proficiency test battery, developed by the English Language Center, tests students in grammar and vocabulary, listening comprehension, reading, and composition, with the grammar and vocabulary portion being given a combined score. The Kuder—Richardson 20 reliability of the grammar-vocabulary test is .93; for the listening test, .88; for the reading test, .87; and for the composition test, .80. The estimated overall reliability for the test battery is over .95. Validity studies for this test-are indirect, based largely on (1) the accuracy with which the test appears to place students in groups at the English Language Center; (2) the correlation of this test with other nationally normed proficiency tests; and (3) the success with which it predicts which students will be able to function linguistically in their academic programs at Michigan State University. The scores on this test range from 35 to 95 out of 100 possible points, with 80 points or above usually considered the ac- ceptable criterion level for proficiency. A demographic information sheet was given to the large group along with the Learning Preference Opinionnaire. The 30 subjects were asked to complete this form, which requested a variety of demographic data to serve as variables in Hy- potheses 2 (language/cultural background), 3 (foreign lan- guage learning experience), 4 (age and amount of time at the ELC) and 5 (sex, roommate, and study plans). 2. Small Group Instruments The proficiency test, the LPG and the demographic data sheet were also used with the small group. However, four additional instruments were used in gathering small group data. All but the reading selection are in the appendices. The language background Questionnaire was a modified version of the Adult Interview Questionnaire used by Naiman, 33 21‘ in the Good Language Learner study (1978). It asked students a series of questions designed to shed light on their language learning experiences, and also elicited their ideas about learning a new language. The language of certain questions on the Adult Inter- view Questionnaire was found to be somewhat idiomatic, and most of the changes involved simplification. It was also found that in all but one case, questions which suggested that the learner may be multilingual were irrelevant to the subjects, nearly all of whom were monolingual before they came to the United States. Naiman E£.El- emphasize the "possible advantage of the interview as an alternative means of obtaining the same information as standardized tests of 31 personality and attitude, but at greater depth and without the disadvantage of test procedures" (Naiman 35 21° 1978, 101). The readipg task Checklist was used during a mini-lesson in which students were asked to read an article and comment on it. The checklist was used to aid in evaluating whether their behavior and level of comprehension reflected analytic or synthetic dimensions of learning. The reading task selection, on which the reading task was based, was a short article from the student newspaper News for You dealing with the results of.an earthquake in Italy. The newspaper, which uses somewhat simplified lan- guage, is used in adult literacy classes in the U.S. The teacher evaluation checklist was used while inter- viewing at least two teachers for each case study subject. The teachers were asked to characterize their students as learners, using criteria such as attention to form vs. attention to meaning. I then used the checklist as an aid in rating the analysis or synthesis tendencies of each student. E. Selection of Subjects The subjects who participated in this study consisted of both a large and a small group; the large group was made up of 156 adult foreign students enrolled in all levels at the English Language Center of Michigan State University, 32 and the small group contained eleven of these students who had been selected for in'depth interviews. The 156 subjects in the large group represented 38 dif- ferent countries and collectively spoke over 28 languages. There were 111 males and 45 females, whose levels of profi- ciency ranged from elementary to very advanced. Because a major assumption of this study was that learning style would change as a student became more proficient in English, it was necessary to include the elementary level students in the large group even though their English was limited. Special administration procedures are discussed in the ap- propriate section of this study. Subjects for the small group were chosen to represent the three largest native language groups at the ELC: Arabic (44), Spanish (32) and Japanese (29). It was also desired to represent various levels of English proficiency within each group, so nine interviewees were selected, rep- resenting low, intermediate and high proficiency Arabic speakers; low, intermediate and high proficiency Spanish speakers; and low, intermediate and high proficiency Japanese speakers” The subjects, chosen from a list provided by the foreign student counselor, were described by several teachers as likely to be enthusiastic and cooperative. In two cases--intermediate proficiency Spanish speaker and high proficiency Spanish speaker~-it appeared for a time that the subjects would be unable to continue their partici- pation, and new subjects were recruited. However, all 33 subjects completed their interviews, and the study was finished with two extra Spanish~speaking subjects for a total of eleven interviewees. Each language group was also represented by at least one female and two males. F. Procedures for Administration of Instruments 1. Large Group Procedures The Learning Preference Opinionnaire was administered as follows: (1) the LPO was administered only in English; (2) an overhead projector was used to display a large copy of the LPO for purposes of explanation; (3) the individual classroom teacher and I provided as much individual help as possible; (4) if necessary, students could make comments on the questionnaire in their native language; and (5) stu- dents were asked to imagine themselves in an out-of-class situation and to respond to the questionnaire as if they were working on their own to learn English. The proficiency test battery was administered by the staff of the English Language Center in accordance with their regular procedures in the middle of the term, after which the scoring and distribution of scores were handled by the ELC testing office and the results made available for this research. Demographic information was gathered by asking students to fill out a data sheet which was attached to the 34 Opinionnaire and administered along with it. No particular instructions were provided other than those given during the general administration of the LPO. 2. Small Group Procedures In addition to using the LPO, the proficiency test bat- tery and the demographic data sheet as described for the large group, I gathered small group data as described below. The LPO-based interview was spent discussing each strategy on the LPO in some detail. Students were asked to explain HEX they considered certain strategies to be useful or not useful, and were encouraged to elaborate. The language background questionnaire, used during the second interview, contained a series of questions on lan- guage learning background based on the Adult Interview Questionnaire of the Good Language Learner study. Unlike the Learning Preference Opinionnaire, which asked about specific strategies, the language background questionnaire asked subjects directly about their learning styles (e.g., preference for a systematic vs. an intuitive approach). _The reading task provided the basis for the third and last interview. Each subject was given a simplified news item about a recent major earthquake in Italy and was asked to read it and do whatever was necessary to understand it, such as guessing at new words in context, or using either an English-only or a bilingual dictionary. The atmosphere 35 was intentionally informal, since the LPO had already been considered in an out-of«class context. The students were not overtly instructed to read for facts or recall, but just to get the general meaning. They were then asked to relate as much as they could remember about the story, and the reading task checklist was used to note my observations of things such as their comprehension and their care in reading. The teacher evaluation checklist was used while gather- ing information during informal conversations with several of the interviewees' classroom teachers. An attempt was made to contact at least two teachers for each small group subject. The teachers were asked to characterize their stu- dents as language learners in general, and to comment on such aspects of learning as fluency vs. accuracy,attention to form vs. attention to meaning, and the student's depend- ence on his/her native language. G. Analysis of Data The two types of treatment used in the analysis of data for.the large group were: (1) special scoring proce- dures;and (2) statistical analyses, as described below. A third type of treatment was applied to the data from the small group interviews and will be described later. 36 1. Special Scoring Procedures Since the fiveepoint answer scale on the Learning Preference Opinionnaire indicated whether the subject felt each strategy was useful (#1) or not useful (#5), a positive response (a low score) for eleven of the items would indi- cate a high preference for synthesis, and a positive response (a low score) for the other eleven items would indicate a high preference for analysis (i.e., a low preference for synthesis). As a result, the eleven items reflecting a preference for synthesis had to be transformed so that the students would receive a high score for all items showing a preference for synthesis, even if the response was #1 or #2. This adjustment was done by computer, and in all statis- tical analyses 3 lower score indicated preference for analy- sis and a high score indicated preference for synthesis. The responses on the twenty—two items were then added together into an overall total score. In addition, to determine a smaller working number (the Analysis/Synthesis Average) that represented the learning style, I divided this total by the number of valid responses for each student (i.e., questions not answered were not included when com- puting the average). In other words, the total score was divided by the number of questions the student answered to give each student's Analysis/Synthesis Average, or ASAVE. The proficiency test battery was scored according to the normal procedures of the ELC Testing Office; at the 37 time of the study, both initial and midterm scores for each student were made available for this research. A progress score was then computed for each student by subtracting the average score for the first test from the average score for the second (midterm) test. For the convenience of the sta- tistical procedures that were to be used in the analysis of the data, the progress scores were grouped into five ranges of progress, from low to high. Demographic data obtained along with the responses on the LPO were numerically coded for use in the statistical analyses, and will be discussed in more detail in the following chapters. 2. Statistical Treatment of the Large Group Data Responses from the Learning Preference Opinionnaire and the demographic data sheet were coded onto computer data sheets, after which data cards were punched by the Test Scoring Office at Michigan State University. All com- puter programs used were from the Statistical Package for the Social Sciences (SPSS). The following statistical analyses were applied to the data from the LPO: (l) Cronbach's Alpha reliability coefficients were cal- culated for the analysis items alone, for the synthesis items alone, and for the LPO as a whole. (2) Pearson Product«Moment correlation coefficients 38 were calculated to show the relationship between ASAVE and (a) proficiency, (b) age, (c) time at the ELC, and (d) pro- gress, in order to test Hypotheses l and 4. (3) Regression analyses were run for both the whole population and for the following subpopulations: language group (Arabic, Japanese or Spanish); sex (male or female); roommate's language (English, same native language, or no roommate); and study plans (English only, or academic pro- gram); these regression analyses permitted the testing of Hypotheses 2 and 5. (4) Analyses of variance and covariance were used as an additional means of testing Hypotheses l, 2, 4 and 5. However, it turned out to be impossible to test Hypo- thesis 3 in this study. Hypothesis 3 states that learning style as measured by the LPO will vary according to differ- ences in previous foreign language learning experience. A frequency count on the demographic data showed that there simply were not enough cases to allow this factor to be investigated. 3. Analysis of Small Group Data An attempt was made to measure formally the responses of the small group interviewees during the investigation by assigning a score to all four aSpects of it (the ASAVE, the LPO interview, the language background interview and the reading task). A chart was prepared evaluating data from 39 these four areas in terms of a tendency toward analysis or synthesis. This was done with an eye toward determining the validity of the Learning Preference Opinionnaire by attempting to calculate an independent learning style score that could be compared with the Analysis/Synthesis Average. H. Summary This chapter has defined the principal terms used in this study, and has described the general procedures.folr lowed and the instruments used. A discussion of the selec- tion of subjects was followed by a description of adminis- tration procedures and analytical treatment of the data. CHAPTER III ANALYSIS OF LARGE GROUP DATA A. Introduction and Overview This chapter presents the results of analyses of the data gathered during the administration of the Learning Preference Opinionnaire to the large group. The first part of the chapter presents a description of how the LPO was scored; the means and the standard deviations for the whole group and for the language/culture groups used in this study; and the estimates of reliability for the Opinionnaire and for the separate analysis and synthesis scales. In the second part of the chapter, the results of various statis- tical analyses are reviewed in terms of whether or not they appear to confirm the first five hypotheses of this study as listed at the end of Chapter I. B. Results of the Administration of the Opinionnaire 1. Scoring of the Learning Preference Opinionnaire As discussed in Chapter II,_the LPO was administered to a total of 156 students in the regular classes at the 40 41 English Language Center. The twentyatwo responses of each student, which were made directly on the Opinionnaire, were then transferred to data sheets and scored by machine. As mentioned in the preceding chapter, on eleven of the items, a positive response was associated with a high pref- erence for synthesis; on the other eleven items, a positive response was asSociated with a high preference for analysis. Therefore, in order to have a continuous scale from analysis to synthesis, with the low end of the scale consistently representing analysis and the high end of the scale consis- tently representing synthesis, the synthesis-items were transformed as follows: (1) was assigned a value of S (2) was assigned a value of 4 (3) was assigned a value of 3 (4) was assigned a value of 2 (5) was assigned a value of l The total score and the Analysis/Synthesis Average are not shown separately for each of the 156 subjects, but were included in all of the statistical procedures used in the study. The mean Analysis/Synthesis Averages and the standard deviations are recorded in Table 1, both for the whole group and for the three language/cultural groups of interest. Table l. Analysis/Synthesis Average (ASAVE) Population N Variable Mean 'SD Whole Group 156 ASAVE 3.19 .33 Arabic 44~ ASAVE 3.14 .36 Spanish ‘32 .ASAVB 3.17 .36 Japanese 29 ASAVE 3.13 .29 Others 51 ASAVE 3.29 .30 fl 42 The interpretation of these figures will be included in the discussion of the hypotheses, but it should be pointed out that while the ASAVE is somewhat higher for the miscellaneous "other languages," no meaningful conclusions can be drawn since this group contains no more than ten members from any one language group; most of the languages represented in this group have only three or four speakers at the ELC, not enough to justify any separate implications regarding ASAVE. It is readily apparent that the mean response was in the center of the range, and showed little tendency toward either synthesis or analysis. One poSsible explanation for this result is that students appeared to have a strong ten- dency to choose the positive or "yes" response to the items. Since in half of the items this represented a score of five, and in the other half of the items it represented a score of one, the average tended to fall in the middle. As evidence of this interpretation, it was common for students to say during the administration of the LPO, "All of these are useful!" In other words, students tended to say that the strategies listed were "useful," whether they represented analytic or synthetic learning strategies. 2. Reliability of the Learning Preference Opinionnaire In order to evaluate the usefulness of the Learning Preference Opinionnaire as a measuring instrument, a 43 Cronbach’s Alpha reliability test was run, yielding alpha reliability coefficients. This particular formula was chosen because it could be applied to a single administra- tion of the LPO. Reliability coefficients were obtained for three sepa- rate scales: the analysis items alone, the synthesis items alone, and the complete opinionnaire with analysis and syn- thesis items together. The Cronbach's Alpha reliability coefficient for the instrument as a whole was .48--a fairly respectable coefficient for a new, untested instrument, but not high enough to allow for the type of definitive state- ments about learning style we would like to be able to make. Overall, analysis and synthesis items correlated nega- tively with each other, as expected. However, the point of interest here lies in the reliability coefficients for the separate analysis and synthesis scales. The reliability for the analysis items alone was .43; on the other hand, the synthesis items alone yielded a coefficient of .69. These results were surprising in view of the fact that all twenty-two items were initially thought of as measuring the same variable of learning. As it turned out, however, the two halves of the instrument behaved differently. The analysis items generally did not correlate well with each other even within that half of the LPO, whereas only one synthesis item (#16) correlated negatively with the other synthesis items. Further consideration of the items revealed that if 44 item #16 were eliminated from the synthesis scale, the re- maining items on that scale represented the ten best items on the questionnaire. It was then found that if these ten items were used exclusively, a reliability coefficient of .72 could be obtained with this instrument. C. Results of Statistical Studies To test the hypotheses, the following statistical pro- cedures were used: (1) regression equations were used to determine the Pearson Product-Moment correlation coefficients used to describe the interrelationships among variables; and (2) analysis of variance and analysis of covariance pro- cedures were used when the data were not linear. Since the variables in a regression equation must be continuous (numerical), the Analysis/Synthesis Average (ASAVE) was chosen as the dependent variable in the regres- sion analyses used for this study; proficiency, age, amount of time at the ELC, and progress functioned as the indepen- dent variables. These analyses tested Hypotheses l and 4. The other type of variable, one for which a numerical value would have no meaning, is called a categorical vari- able. The categorical variables in this study (language/ cultural group, sex, roommate's language, and study plans) were handled in two ways: (1) separate regression analyses .were run for each 0f the subpopulations of interest; and (2) analyses of variance were used to determine the effects 45 of categorical variables on continuous ones, e.g., the effect of language/cultural group on ASAVE. These two approaches were used to test Hypotheses 2 and S. The first hypothesis to be reported on, however, will be Hypothesis 3, which states that learning style prefer- ences of ESL students, as measured by an instrument (the LPO) designed for this study, will vary along with differ- ences in previous foreign language learning experience. Results of a frequency count showed that only 20 of the 156 students in the total population spoke another foreign lan- guage; nearly all of these were Europeans, a very small per- centage of the ELC population. Since such a small sample would not permit justification of any speculations which might be made about it, I concluded thatit would not be possible to test Hypothesis 3 within the context of this study. As previously mentioned, regression analyses were used to test Hypothesis 1, which states that ASAVE “all. vary along with differences in proficiency; and Hypothesis 4, which states that ASAVE MdJJ. vary along with differences in age, amount of time at the ELC, and progress. From these analyses, Pearson Product-Moment correlation coeffi- cients were obtained showing the relationship between ASAVE and the variables of proficiency, age, time at the ELC, and progress. These are displayed in Table 2, along with an indication of statistical significance (p values). 46 Table 2. Correlation of ASAVE with Proficiency, Age, Time at the ELC, and Progress (whole group) Variable Simple r with ASAVE Proficiency .36** Age ..05 Time at ELC .09 Progress -.03 **p (.01 (highly significant) The simple r-correlations show that there is a moderate but statistically significant correlation between ASAVE and proficiency for the population of 156 ELC students, and no meaningful correlation between ASAVE and the other variables. In other words, two conclusions can be drawn from these data: (1) students who scored higher in English proficiency also tended to have higher Analysis/Synthesis Averages (i.e., a greater tendency toward synthesis); and (2) proficiency was the only variable which had any significant correlation with ASAVE. Since age, time at the ELC, and progress show no significant correlation with analysis or synthesis, they will not be included in future tables showing results of the statistical analyses, and will only be mentioned where relevant. In summary, Hypothesis 1, stating that ASAVE will vary along with differences in proficiency, is moderately confirmed in this study. Hypothesis 4, stating that ASAVE will vary along with differences in age, time at the ELC, and progress, is not confirmed. Hypothesis 2, stating that learning style as expressed 47 by ASAVE *will vary along with differences in language/ cultural background, was first tested by running separate regression analyses for the three subpopulations of interest, i.e., speakers of Arabic, Spanish or Japanese. Again, pro- ficiency proved to be the only variable which had signifi- cance. Table 3 illustrates both the simple r-correlations for these subpopulations and the percentage of the variance in ASAVE that is accounted for by proficiency, using the generally accepted calculation that the percentage of the variance is expressed by the statistic multiple r square. Table 3. Correlations of ASAVE with Proficiency According to Language/Culture Populations Population _fl_ Variable Simple r ” Multiple RSQ Whole Group 156 Proficiency .36** .13 Arabic 44 Proficiency .301 .09 Spanish 32 Proficiency .54* .29 Japanese 29 Proficiency .48* .23 Other 51 Proficiency .16 .03 **p<.01 *p<.05 1.05=1nese and Portuguese as two examples. 71 Nearly all of the students said they would prefer to learn the new language intensively; J2 was the only one who wanted to proceed "little by little." He also said that the first thing he would do to learn French would be to take a French course in Japan, three times a week for a long period of time, to gain elementary knowledge. The only other sub- j’ ects who said they would take private lessons in their own countries were 81 and J3. Most of the interviewees were very much in favor of the method of going to live in a country where the target language is spoken and taking a course there--very much the same thing they have done by coming to the United States to learn English. In some cases, though, they would go beyond that. Al and A2 specified that they would want to live with a family in the target culture, in addition to taking a course. 83A said that he would study part time, but would want to live in a small town and get most of his language practice "on the street." Only one subject said that he would "go to the country and immerse himself in the language." Here, the word "im- rue‘I‘se" is understood to refer to an out-of—class context, such as people describe when they spend some time in a Country and "pick up" the language without formal study. I hterestingly, this Student (Jl) rated himself only "elemen- tary" in all aspects of his English ability. He was also the most difficult to interview, due to his limited abili- ties in both understanding and expressing himself in English. 72 One subject, SZB, found he could not give an easy an- swer to whether he would take private lessons, visit the country, or use other methods to learn a language, saying that it depended on the purpose. If he were interested in the language in order to do business with its speakers, he would take an intensive course such as that in the ELC; but if he just wanted to travel and meet people, he would be likely to take a more intuitive approach, such as going to Japan and immersing himself in the language. SZB was the only interviewee to raise the issue of purpose in language learning; as will be seen later, a Student's purpose for doing a language task is crucial and makes a significant ‘11 fferencein the way he/She approaches that task. When asked to describe their final gOal in learning a new language, the interviewees' responses Showedhow p111‘I>ose can influence language learning approach. All three of the Japanese Students, as well as 81, said that their final goal in learning a new language would be just to Speak and understand it, usually for travel purposes. These four people were the same ones who would not choose to take a language course in the target country, but would rather learn it initially at home (51, J2, J3) or by an "immersion" method (J1). Of the remaining interviewees, Al Said that his goal was to read and understand the lan- 81lage; all the rest wanted to gain total communicative s .. 1(1115 in speaking, understanding, reading and writing. 73 6. Specific Learning Strategies Both the Learning Preference Opinionnaire and the modified Adult Interview Questionnaire asked students to describe their specific learning strategies for dealing with another language. The LPO served as an initial point of discussion in the first interview, and both the subjects' choices on the Analysis/Synthesis scale and the comments they made about specific items generally confirmed the overall pattern, in terms of proficiency level and in terms 0f culture, Shown in the large group data. The other more open-ended questionnaire asked students to describe what they thought was appropriate at various 8":ages of language learning, especially at the beginning 3 tage. As expected, "understanding" was the need most fre- quently mentioned. The next most important item was "simple C()l'lVersation"; obviously, a student's main concern when just lz’eginning to learn a new language is to gain basic skills for "survival" purposes. Other items, chosen from a list, iliCluded reading, grammar and pronunciation. On the whole, the interviewees insisted that there was no _o_n_g skill which had to come first, but that they needed various combinations of skills; A3 and SZB were particularly :8 irm about this. It was also interesting to note that these two students, along with SZA and J1, said that a new language lgarner should speak and be active right from the beginning. The rest said that in the early stages they would rather be 74 more passive and develop their receptive abilities before trying to write and speak in the new language. Finally, the majority of the interviewees agreed that language learning is an active, conscious process. A few students qualified this Statement; S3A said that it was partially conscious, but somewhat more intuitive than con- scious; S3B said that it was a conscious process, but that motivation was important; Jl, who is lower in proficiency and is a Slower learner, said that languages could be learned unconsciously (i.e., "acquired") if the learner had a long enough time in which to do it, but that the process w(31.11d be more conscious if the learner had only a short time for it. Only 51 said that the language learning pro- Cess was primarily intuitive. 7 ‘ Teacher Evaluat ions Brief interviews were held with at least two of each S‘11533'ect's classroom teachers, during which I asked them to c1liiracterize their students as learners on the basis of their classroom behavior. The teachers' characterizations g eherally coincided with the impressions I had formed of each student after interviewing him or her. This was evi- dent even when the teacher and the student were at cross 131“”«l‘poses; for example, the student who said his teachers were only there to do a job, not to help him, was described 75 13y his teachers as passive, bored, and. generally resistant to class activities. 8 - The Reading Task For the reading task, students were given a simplified newspaper item and asked to read it and do whatever they needed to do to understand it. They were then asked to summarize the item, and were asked a few questions if the summary wasn't clear to me. The most interesting result of this was that even Students who tended to be analytic and said they looked up eVery word tended to take a rather casual approach to the task. Some of them, including A2 and J2, admitted that they did not try to read it carefully, but just to get the general meaning, as I had told them to do. These results tend to support previous statements that the purpose of a language task may be an important determi- nant of a student's approach to the task. In a few cases, they were given a second article to read after being ques- t ioned on the details of the first article. Immediately, they changed their attitudes, anticipating that they would a8a-in be asked questions about details. Even highly syn- thetic learners like J3 and S3B began to use their diction- a. - 1+ les and take notes. 76 D. Conclus ions Results of the interviews with the case Study subjects 1F<>llow the general pattern shown by the large group data; rlzalnely, that lower proficiency students tend to be more analytic and higher proficiency Students tend to be more synthetic in their learning styles. The interviews also tend to confirm that these variations occur across all cul- tures; that is, with respect to an analytic or synthetic approach to language learning, a high proficiency Japanese Student would tend to have more in common with a high Preficiency Arab, for example, than with a low proficiency Japanese. The most important factor in a student's language learning background appears to be outside contact with SSJF’€?£1kers of the target language. Here, the subjects who were highest in synthesis were also the ones who had had C’ol‘ltact with Americans in their home countries or who saw E11glish as a means of communication. Many of the other interview questions brought out dif- ferences which seemed more related to personal preference At:]hl€1n to proficiency or language group as a rule. However, One point on which many of the high synthesis and high pro- .jE.jL<:iency learners tended to agree was that "ability" was J‘ess a factor in language learning success than ordinary s thdy and hard work. Another conclusion drawn from the interview study was 77 that learning styles and learning strategies vary according to both the purpose of the task and the context of language learning. Students could sometimes be quite analytical, as when they thought they might be quizzed, but when the pres- sure was off, they tended to be more concerned with communi- cation than with cracking the language code. B - Relationship of Small Group and Large Group Data As Stated in previous chapters, one of the reasons for COllecting data from a small group as well as a large group Was to see if the interview data would assist in determining tlie validity of the Learning Preference Opinionnaire as a measuring instrument. This is the test of the final hypo- thesis of the study, as stated below: H6--Learning style variations among ESL students, as measured by an instrument designed for this Study, will be reflected in case study data collected from a small group of students, using interviews, a miniature learning expe- rience and ratings by teachers. In order to translate the small group data into some 8 c37ft of independent numerical measure for comparison with the Learning Preference Opinionnaire, I weighted each of the components of the small group investigation as follows: the teacher evaluation, 1;, the reading. task, 2; the language background interview, 3; and the LPO~based inter- V - ' 16w, 4. These values represent an ascending order of 78 importance of the components; the LPOebased interview, of course, ranks highest, since weare attempting to assess the validity of the LPO for the large group. Each interviewee received a score from one to five for each component, with "one" representing the analysis end of the scale and "five" representing the synthesis end of the scale. The score for each component was then multiplied by the number representing the weight ranking for it; these products were then totaled and divided by ten, to come up with an overall score for each interviewee that could be compared to the Learning Preference Opinionnaire. Table 8. Overall Learning Style Scores of Interviewees* a Sub (1) (2) (3) (4) J ect Teacher Reading Language LPO-Based Overall ASAVE Eval. ' Task Background Interview Interview A1 3. 5 4 3. 5 ' 4 3.8 3. 63 A2 3 3 2. 5 2 2. 45 2. 19 A3 3 4 4 3 3. 5 3. 23 $1 3 2 3 2.5 2.6 2. 59 SZA 5 4 4 3. 5 3. 9 3. 27 SZB 4 4 4 3 3.6 3. 33 83A 4 4 4 3 3. 6 3. 36 $33 3 4 3 4 3. 6 3. 72 J1 2 2 3 2. 5 2. 5 3. 09 J2 3 3 3 2. 5 2.8 3. 04 J3 3 4 4 4 3.9 3. 77 \ *D-.86 correlation between overall interview score and ASAVE 79 While based on subjective criteria, these scores do appear to provide some evidence to support the Analysis/ Synthesis Averages produced by the Learning Preference Opinionnaire. Although most of the overall interview scores are somewhat higher-wand in a few cases, lower--than the Analysis/Synthesis Averages, it is readily apparent that the independent interview scores follow the same gen- eral patterns for the same students as the ASAVE scores. Ca 1culation of a Spearman rank-difference correlation co efficient for the interview scores and the ASAVES resulted in a correlation of .86. This confirms Hypothesis 6, that 1 e arning style variations Shown by the Learning Preference Op inionnaire in the large groUp would be reflected by the Gas e study interview data. CHAPTER V SUMMARY, DISCUSSION AND IMPLICATIONS A - Introduction This chapter will first present a summary of the results 0 f the study. Some of the findings were unusual or unex- Pe cted, and these are discussed in the first section of the chapter, along with some speculations on the possible causes. The remainder of the chapter discusses implications of the pr esent research, with possible classroom applications and snggestions for further research. E3 “ Summary of Results Data were gathered on student learning styles in terms c:.:£; zinalysis or synthesis by asking them to respond to a list Of Strategies on the Learning Preference Opinionnaire. Re- 8 ~. 2. 1 ts of the administration of the LPO to a group of 156 a . . (11.111; ESL learners and data gathered during interv1ews w1th gr('rup of eleven learners indicated the follow1ng with r _ espect to the hypotheses of this study: Hypothesis 1, that learning style as expressed by the 80 81 Analysis/Synthesis Average (ASAVE) on the Learning Preference Opinionnaire would vary along with the student‘s level of proficiency in English, was confirmed. Hypothesis 2, that learning style as expressed by ASAVE would vary along with differences in a student's language/ cultural background, was not confirmed. Hypothesis 3, that learning style as expressed by ASAVE would vary along with differences in previous foreign lan- gu age learning experience, could not be tested in this study due to an insufficient number of cases. Hypothesis 4, that learning style as expressed by ASAVE would vary along with differences in age, amount of time s12>ent at the ELC, and progress, was not confirmed. Hypothesis 5, that learning style as expressed by ASAVE would vary along with differences in sex, roommate's language, and study plans, was not confirmed. Hypothesis 6, that learning style preferences expressed on the Learning Preference Opinionnaire would be reflected in case study data gathered from a small group of eleven 8 tudents selected to represent a cross section of the large group, was confirmed. Q ‘ Discussion of Limitations In view of statements by previous researchers that suc- Q . esSful language learners take a systematic approach to learning (Naiman, gt ’_a__1_. 1978) or else see it as a 82 problemasolving task (Hosenfeld 1976), the emergence of syn- thesis as a major factor was somewhat unexpected. Three Possible reasons for this finding were explored: (.1) the Learning Preference Opinionnaire may have somehow favored those who had a preference for synthesis, or else the synthesis items might have been more reliable; (2) the context in which students were asked to respond to the items, :3. - e. , the assumption of an "out-of-class" situation, may have influenced their answers; and (3) the particular makeup of the ELC population may have contained more learners who We re slightly oriented toward synthesis. 1 - Reliability of the Learning Preference Opinionnaire As stated in the results chapter, the analysis and syn- thesis dimensions of learning were originally envisioned as b e ing at opposite ends of the same continuum. However, the wide difference in the reliability coefficients for the two 3 Qa~1es, analysis and synthesis, suggests that the two halves 0 Is the Opinionnaire may in fact have been measuring two dif— ferent types of learning processes, and not two aspects of the same dimension of learning. To review briefly, the Cronbach's.Alpha reliability Qoefficient for the analysis scale was .43, with‘the relia~ b ility of the synthesis scale at .69, and- the overall re? *1 iability of the instrument at .48. This low reliability 'suggests a need for further refinement of the Opinionnaire 83 before definitive conclusions can be drawn regarding learn- ing styles of the type investigated in this study. In fact, with the present instrument, one apparent improvement was easily indicated. The eleven items for which a positive response indicated a preference for synthesis had a reliability coefficient of - 6 9. One of the items (#16) appeared to be negatively cor- re lated with the other synthesis responses, and eliminating it raised the reliability to .72. While still low, this figure represents surprising stability for so few items. Even so, in view of the relatively low- reliability of .48 for the whole instrument, one can probably assume that the p°sitive and significant correlations between ASAVE and pro- fi ciency would increase withhigher reliability. Other cor- I“5'31ations with age, time at the ELC, and progress are so 1 0W that they are unlikely ever to be of major importance, though of course further studies need to be conducted with a refined instrument, perhaps with synthesis items only. The interview studies also offered some data to support the previous suggestion that the instrument may have been measuring two different types of learning styles. For ex- a‘:l"11[>le, the two students with the highest synthesis scores were J3 (3.77) and 5313 (3.72). However, J3's high score was the result of positive responses to the synthesis items, while 838's high score was the result of negative responses to the analysis items. It seems that being pro-synthesis 18 not necessarily the same as being anti-analysis. Thus, 84 the Learning Preference Opinionnaire may well be measuring two different dimensions of learning, rather than two as- pects of a single dimension of learning. 2- Context of the Learning Situation The interview data showed that students may take a more analytical approach to a task if they are asked to meet more formal expectations. In anticipation of the possibility that students might take approaches in the (:1 assroom which they might not otherwise choose when learn- ing a language outside of class, the LPO asked the large group respondents to assume an out-of—class, informal S ituation while answering the questionnaire. A student who would "read for meaning" at home might "read for detail" in class. If students were indeed assuming an informal context a 1iile answering the questionnaire, it would make sense for them to. favor the synthesis items. In research such as the Gt>c>d Language Learner study, where learners indicated a high 1:) a1‘eference for analytical methods of language learning, the, : eSults may likewise have been influenced by the possibility O f students assuming an ineclass situation. It is possible that if the Leraning Preference Opinionnaire were adminis— tered again and students were asked to indicate what they found useful for in-class learning, the mean responses might well lean more toward the analysis end of the scale. 85 3. Makeup of the ELC Population In view of the positive correlation between ASAVE and proficiency, the ASAVE may have leaned slightly toward syn- thesis when a high number of the ELC students were at the advanced proficiency level. This in fact was found to be the case, with the mean proficiency score for the population be :ing 71 points out of 100. - Implications for Teachers and Learners The major conclusion that a tendency to synthesize is Correlated with higher levels of proficiency in English has 8 everal implications for ESL learning. First, it supports the idea that language learning is different in the elemen- tary, intermediate and advanced stages. all, Research by Naiman, (1978) Stern (1975) and Rubin (1975) suggested this may be the case. Second, these "stages" of language learning are more 1 ikely to be part of a continuum of language learning pro- gl‘ess, rather than three separate stages such as beginning, lhtermediate and advanced. Evidence for this may be seen in the ASAVE rankings and the proficiency rankings for the Stuall group subjects (Table 6), where lower and middle pro- fiCiency students' 'scores tend to overlap. Furthermore, S ince "proficiency" has been defined in this study as a student's position on a continuum of ability in English, 86 there is ample reason to describe the language learning pro- cess itself in the same way. Another implication is that in view of the conclusion that we cannot predict a student's learning style on the basis of what country he/she comes from or what language he/she speaks, teachers should be prepared for students to use 119;}; analytic and synthetic approaches to learning, according to the context and the student's purpose, and should not assume, for example, that Spanish-speaking stu- dents will use a more synthetic approach than Arabic-speaking s tudents. Finally, it may prove useful for ESL teachers to use a simple Opinionnaire, such as a refined version of the LPO, to learn what strategies their students are bringing into the classroom with them. John Nelson (1981) stresses that the ESL instructor needsto learn as much as possible about his/her students' learning strategies-~those learned from previous experience-~in order to help them develop b etter ones. E - Suggestions for Further Research It must be stressed that this has been a preliminary s tudy, and that much more needs to be done in the area of l earming styles and strategies of. ESL students. One sug- geStion, obviously, is the refinement of the Learning Preference Opinionnaire. It would be desirable to further 87 refine the synthesis items so that a more accurate measure of the dimension of synthesis may be obtained. Furthermore, it may be a good idea to develop two different forms of the opinionnaire--one for analysis, one for synthesis-~rather than considering analysis and synthesis as part of a continuum. Another suggestion is to find other ways to estimate the validity of the Learning Preference Opinionnaire, perhaps by comparison with other instruments that measure analytic and synthetic learning dimensions. Third, since for the whole group and for one subpopula- tion, the percentage of the variance accounted for by the ASAVE-proficiency correlation is quite small, it is necessary to explore what other variables may further explain the variance. In particular, it would be desirable to give standardized tests such as Carroll's or Pimsleur's aptitude battery and Gardner and Lambert's battery of tests on atti- tude and motivation, as well as cognitive style measures such as field dependence/independence and tolerance/intoler- ance of ambiguity, especially since these measures have been shown to predict language success (Naiman, 33.213 1978). .F. A Possible Analysis/Synthesis Model for Language Learning The preference of a student for analysis or synthesis whet: learning a language may well be related to what Stephen Krashen (1976) has called language acquisition and language 88 learning. Krashen has said that when a language is inter« nalized "naturally," as children learn, this may be called "acquisition," as opposed to the quite conscious language "learning" of the adult. The question here is, just how well does the adult second language learner internalize the target language? In the beginning, not well at all. The learner usually studies the rules, analyzes them, tries to generalize, and monitors his/her own speech carefully when trying to pro- duce the language. The learner constantly asks, "Why?" when deciding how to handle a complex situation, and may depend on the dictionary or translation. But as the L2 learner progresses toward L2 competence, behaviors such as analysis and monitoring tend to become less and less conscious. They may still be there, but they are gradually internalized and become automatic, i.e., they have become "acquired." Just as a native speaker of a lan- guage may "feel" certain uses of language to be correct but may not be able to say why, so a highly proficient non-native Speaker may "know" something is right or wrong without know- ing exactly why. The suggestion is that the more advanced an L2 speaker is, the more his/her Second language has in common with the. native speaker's "acquired" language; and that his/ her Llearning style becomes more synthetic or intuitive becanise the target language has become more internalized. APPENDICES APPENDIX A LearninggPreference Opinionnaire To the student: You are now taking classes in the English Language Center. Imagine that during the next month you can study English in any way you want, and to do what you, perSonally, know works best for you. How useful would each of the following study activities be? You may write comments or explanations if you wish, and you may use either English or your own language for this,‘ (Note for the reader: Analysis items are #1,3,4,S,8,10,11,14,17,18,20; Synthesis items are #2,6,7,9,12,13,15,16,19,21,22) 1‘21 1195 Useful Useful 1. Memorizing grammar rules 1 2 3 4 5 (comment) 2. Guessing new words by looking at the 1 2 3 4 5 whole sentence or paragraph (comment) 3. Doing homework assignments l 2 3 4 5 (comment) 4. Writing new words 2 or 3 times 1 2 3 4 5 (comment) 5. Translating English sentences to my 1 2 3 4 5 own language (comment) 6. Trying to think in English 1 2 3 4 5 (comment) 7. Practicing during language lab 1 2 3 4 5 "Open hours" (comment) \ 89 10. 11. 12. 13. 14. 15. 16. 90 Looking up words in a bilingual dictionary (with my language) (coment) Using English word games (examples: "Scrabble" or crossword puzzles) (comment) Reading to get all the words right' (comment) Making lists or charts of new words (coment) Watching television and movies ( comment) Speaking English without worrying about mistakes (comment) Looking up words in an English- English dictionary (comment) Reading to get the general meaning (comment) Looking up only the most important words I don't know~ (comment) Useful 17. 18. 19. 20. 21. 22. 91 Translating my own language to English (comment) Looking up all the words I don't know (comment) Speaking English with as many people as possible (comment) Repeating new words 2 or 3 times (comment) Reading things like advertisements, restaurant menus, comic strips, etc. (comment) Reading books in my major or profession (comment) Not "Useful 5 ***********************************************************************i: 23. What other things, not mentioned here, do you find useful? 24. What other things, not mentioned here, do you find NOT useful? 10. 11. 12. 13. 14. APPENDIX B Language Background_guestionnaire (based on Naiman, 25 31. 1978, 106-9) You were born in and your native language is Were any other languages spoken in your neighborhood? What was the first foreign language you learned? When you learned , what did you study--mainly grammar, for example? Do you remember what kind of textbooks you used? Did the teacher speak in English or in your language? Did the students speak in in English, or did you read and translate? Do you remember what kind of homework you had? Was there anything that was really difficult for you? Did you have any chance to meet Americans, or people who speak English, outside the classroom? Did you hear radio programs or see movies in English? Have you studied any other foreign languages besides English? Which ones? How long? Please tell me whether you think your English is elementary, average or advanced, according to these descriptions (Naiman, p. 6) Elementary, Average Advanced Understanding Speaking Reading Writing Do you think you have a talent for learning languages? that is, are you strong or weak-(not strong) in learning languages? 92 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 93 Do you think you have a good ear for languages?--~ Do you have a good memory? Do you like to look at language yourself and analyze it, figure out the rules? Or have the teacher explain it? Have you been able to learn English because of the teacher? or the English program? or your own study efforts? or your own personality or attitude? If you had the time and the opportunity to learn another language besides English, how would you feel about it? (1) I hate the thought of it (2) It scares me (3) I feel okay (so-so) about it (4) I would like to do it (look forward to it) (5) I am very excited about it. Which language would you choose? What would be your final goal in learning ? (1) To speak and understand (2) To read and write (3) To speak, under- stand and read (4) To speak, understand, read and write. What is the first thing you would do to learn ? (1) Travel to and just walk around and absorb the language (2) Travel to and take a language course there (3) Buy a course-and study by yourself (4) Go to a teacher or language school for private lessons '(5) Take a language class, in the U.S. or in your own country (6) Other If you could take as much or as little time as you wanted, would you prefer to learn intensively in a short time--as at the ELC- or just a few times a week for a longer time? Do you think language learning is different at the beginning, mid- dle and advanced levels? If you were learning , which of the following things would you mainly like to do in the beginning? (1) Understand the spoken language (2) Learn to read (3) Learn pronunciation (4) Learn simple conversations (5) Learn to write (composition) (6) Learn grammar (7) Learn about the cultural background In the beginning of learning a language, would you prefer to be firmly guided by the teacher, or to learn in your own way? In the beginning of learning a language, do you prefer to be more active and speak immediately, or to be more passive and just read, listen, etc.? We've talked about what you like to do in the beginning. Is there anything that's especially useful to do at the advanced level? 28. 29. 30. 31. 32. 33. 34. 94 Do you feel that learning a language is conscious-« that you think about how you do it? or is it unconscious—-you don't learn by studying, but by absorbing it, letting it sink in? Do you feel that when learning English, you have to "forget about" your native language? or does it help if you compare your language and English? ' ' Is it useful to use a bilingual dictionary? ' ' or to translate? Do you feel that it's possible to learn to think in English? How can you learn to do this? Have you developed any special study methods that you would find useful in learning another language, as follows: (1) in learning the sounds (pronunciation) (2) in learning the grammar (3) in learning the vocabulary (4) in improving listening comprehension (5) in learning to speak (conversation) (6) in learning to read (7) in learning to write (composition) Have you ever had any bad or negative feelings about language learning, for example, feeling frustrated or impatient or discouraged or confused ? Did you ever feel strange or silly, or as if you sounded ridiculous to yourself when speaking the new language (English)? Have you ever felt helpless or inhibited while learning English-- as if you were a child and couldn't do anything? APPENDIX C Reading Task Checklist Oral instructions to student: "Read this story and try to understand it as well as possible. If you don't know a word, you may guess, or use an English dictionary, or use your bilingual dictionary. Just do whatever you need to, in order to understand." Evaluation (based on short retelling or summary) None Much 1. Use of dictionary 1 2 3 4 5 Comments: Slow Fast 2. Speed of reading 1 2 3 4 5 Comments: Casual Careful 3. Care in reading 1 2 3 4 5 - Comments: Easy . Hard 4. Difficulty of selection 1 2 3 4 5 Comments: Partial Whole 5. Grasp of meaning ' l 2 3 4 5 Comments: Random Meaningful 6. Grasp of details 1 2 3 4 5 Comments: Low High 7. Level of observed 1 2 3 4 5 comprehension Comments: Summagy: 95 APPENDIX D Teacher Evaluation Checklist How would you characterize this student in general as a learner (e.g., personality, cooperation, motivation, anxiety, etc.)? What kinds of questions does the student ask (e.g., the "why" type vs. the "content" or "what means this" type)? How would you characterize the student's class participation? None Much I 2 3 4 5 Is the student's primary concern in English with accuracy or fluency? Accuracy Fluency 1 2 3 4 5 Does the student attend primarily to form or to meaning in class lessons (especially reading and composition)? Form Meanigg 1 2 3 4 5 How would you characterize the student's dependence on his/her native language? (thinks Low High (translates . in l 2 3 4 5 almost English) constantly) How does the student approach the learning of new vocabulary? a) bilingual dictionary b) English dictionary c) guessing from context 96 APPENDIX E ANOVA Tables Table 9. Analysis of Covariance: ASAVE by Culture with Proficiency _— Source of Degrees of Mean Significance Variation Freedom Sguare F of F Covariate: Proficiency 1 2.364 24.565 .001 Main Effects: Culture 3 .128 1.332 .266 Explained 4 .687 7.140 .001 Residual 150 .096 Total 154 .112 Table 10. Analysis of Variance.of Proficiency by Culture Groups Source of Degrees of Mean Significance Variation Freedom Sguare F of F Main Effects: Culture 3 1046.502 7.157 .001 Residual 151 146.216 Total 154 163.755 98 Table 11. Analysis of Variance of ASAVE by Culture Groups Source of Degrees of Mean Significance Variation Freedom Sguare F of F Main Effects: Culture 3 .268 2.470 .064 Residual 151 .108 Total 154 .112 Table 12. Analvsis of Covariance: ASAVE by Sex with Proficiency Source of Variation Covariate: Proficiency Main Effects: Sex Explained Residual Total Degrees of Mean Significance Freedom Sguare F of F 1 2.374 24.787 .001 1 .172 1.796 .182 2 1.273 13.292 .001 153 .096 155 .111 99 Table 13. Analysis of Covariance: ASAVE by Roommate with Proficiency Source of Variation Covariate: Proficiency Main Effects: Roommate Explained Residual Total Degrees of Mean Freedom Sguare 1 2.374 2 .002 3 .793 152 .097 155 .111 24.347 .022 8.131 Significance of F .001 .978 .001 Table 14. Analysis of Covariance: ASAVE by Study Plans with Proficiency Source of Variation Covariate: Proficiency Main Effects: Study Explained Residual Total Degrees of Mean Freedom Sguare 1 2.374 1 .078 2 1.226 153 .096 155 .111 24.629 0 806 12.717 Significance of F .001 .371 .001 BIBLIOGRAPHY BIBLIOGRAPHY Acton, William. "Second language learning and perception of differences in attitude)‘ Unpublished doctoral disser- tation,University of Michigan, 1979. Bialystok, Ellen. "A theoretical model of second language learning." Language Learning, 28 (1978), 69-83. "The role of conscious strategies in second language proficiency." Canadian Modern Language Review, 35:3 (March 1979), 372-394} and Maria Frolich. "Aspects of second language learning in classroom settings." Working Papers in Bilingualism, 13 (1977) and Maria Frolich. "Variables of classroom achievement in second language learning." Modern Language Journal, 62 (1978), 327-336. Brown, H. Douglas. "Affective variables in second language acquisition." Language Learning, 23 (1973), 231-244. . Principles of Lan ua e Learning and Teachin . 'Englewood Cliffs, New Jersey: Prentice- Hall I , nc., 1980 ; Yorio, Carlos.Alfredo; and Crymes, Ruth 8., eds. 0n TESOL '77: Teaching and LearningEnglish as a Second Lan uage: Trends in’Research an 'Practice. Washington, .C.: Teachers ofEnglish to Speakers of Other Languages, 1977. Carroll, John B. and Sapon, Stanley M. Modern Lan ua e Aptitude Test. New York: The Psychblogical Corpora- tion, 1959. ’ Carroll, John B. "Learning theory for the classroom teach- er." in Jarvis, ed. (1974), 113-149. Chastain, Kenneth.‘ Development of Modern Foreign Languages: Theory to Practice. Philadelphia: 'The Center for Curriculum Development, Inc., 1971. 100 101 Domino, George. "Interactive effects of achievement orien- tation and teaching style on academic achievement." ACT Research Report 39 (1970), 1-9. Dulay, Heidi C.; Burt, Marina; and Finnochiaro, eds. View- points on English as a Second Language. New York: Regents, 1977. Dunn, Rita and Dunn, Kenneth. 'Teaching Students Through Their Individual LearningTSfyles: A_Practical Approach. Reston, Va.: ’Reston Pub ishing Company, Inc., 1978. Farr, Beatrice J. "Individual.differences in learning: Predicting one's more effective modality? Ph.D. dis- sertation, Catholic University of America, 1971. Ann Arbor, Michigan: University Microfilms, July 1971, p. 1332A. Gagne, R., ed. Learnin and Individual Differences. Columbus, Ohio: C arles E. Merrill Publishing Co., 1967. Gagne, R. "Instruction and the conditions of learning." in L. Siegel, ed., Instruction: Some Contemporary View- points. San Francisco: Chandler Publishing Company, Gardner, Robert and Lambert, Wallace E. Attitudes and Motivation in Second Langua e Learning. Rowley, Mass.: NeWburngouse Publishers, 1 721 Hancock, Charles R. "Student aptitude, attitude and moti- vation." in Lange and James, eds. (1972), 127-156. Hosenfeld, Carol. "The new student role: Individual dif- ferences and implications for instruction." in Jarvis, ed. (1975), 129-167. . "Learning about learning: Discovering our students' strategies." Foreign Language Annals, 9:2 (1976), 117—129. "A learning-teaching view of second ’lénguage instruction." Foreign Language Annals, 12:1 (1979), 51‘54. Jarvis, Gilbert A., ed.‘ The Challenge of Communication. (ACTFL Review of Foreign Language Education, Vol. 6). Skokie, Illinois: National Textbook Company, 1974. 4, ed. 'Perspective: ’A New Freedom (ACTFL Reviewof.F6reign Language Education, Vol. 7). Skokie, Illinois: National Textbook Company, 1975. 102 Joiner, Elizabeth G. "Tailoring language instruction to student needs." in Jarvis, ed. (1974), 151'184. Krashen, Stephen. "Formal and informal linguistic environ- ments in language acquisition and language learning." TESOL quarterly, 10 (1976), 157.168. "The monitor model for adult second lan- guage performance." in Dulay, Burt and Finocchiaro, eds. (1977). . "Some issues relating to the monitor model." in Brown, Yorio and Crymes, eds. (1977),l44-158. and H. Seliger. "The role of formal and infbrmal environments in second language learning: a pilot study." International Journal of Psycholinguis- tics, (1976), 15-21. Lange, Dale L. and James, Charles J. Foreign Language Edu- cation: A Reappraisal (ACTFL Rev1ew o oreign an- guage Education, Vol} 4). Skokie, Illinois: National Textbook Company, 1972. Naiman, N.; Frolich, N.; Stern, H.H.; and Todesco, A. The Good Lan ua e Learner. Toronto, Ontario: The Ontario Institute for Studies in Education, 1978. Nelson, John. "Integrating the adult's L2 learning and communication strategies." Paper presented at the 15th International TESOL Conference, Detroit, Michigan, March 1981. New Readers Press. "Tent City in Italy." News for You. Syracuse, N.Y.: December 17, 1980. Oller, John and Richards, Jack C., eds. Focus on the Learner: Pragmatic Perspectives for the’Ian ua e Teacher. Rowley, Mass.: Newbfiry House, 197 . Pimsleur, Paul. Pimsleur Language Aptitude Battery. New York: Harcourt, Brace andIWorld, 1966. and Quinn, T. The PsychOlogy of Second Language Learning. Cambridge: CambridgeUniversity ress, 971. Parker, Orin D. Cultural Clues to the Middle Eastern ' Student. Washington, D.C}: American Friends of the Middle Ease, Inc., 1976. Rubin, Joan. "What the 'good.1anguage learner‘ can teach us." ‘TESO ‘Quarterly 9:1 (1975), 41-51. 103 SavilleaTroike, Muriel.‘ Foundations of Teaching English as a Second Lan ua e. Englewood Cliffs, New Jersey: Prentice-Hall, 76. ' Schumann, Francine and Schumann, John H. "Diary of a lan- guage learner: An introspective study of second lan- guage learning." in Brown, Yorio and Crymes, eds. (1977), 241-249. Siegel, L. Instruction: Some Contemporary Viewpoints. San Francisco: Chandler Publishing Company, 1967. Sperry, Len. Learning_Performance and Individual Differences. Glenview, Illinois: Scott, Foresman andCo., 1972. Strevens, Peter. "Causes of failure and conditions for success in learning and teaching of foreign languages." in Brown, Yorio and Crymes, eds. (1977), 266-277. Stern, H.H. "What can we learn from the good language learner?" The Canadian Modern Language Review, 31 (1975), 304-318. Wesche, Marjorie B. "The good adult language learner: A study of learning strategies and personality factors in an intensive course." Unpublished doctoral disser- tation, University of Toronto, 1975. Wesche, Marjorie B. "Learning behaviors of successful adult students on intensive language training." Canadian Modern Language Review, 35 (1979), 415-427. Zampogna, Joseph; Gentile, Ronald J.; Papilia, Anthony; and Silber, Gordon R. "Relationships between learning styles and learning environments in selected secondary modern language classes." Modern Language Journal, 62 (1978), 443-448.