Richard Schmidt University of Hawai‘i at MÅnoa Yuichi Watanabe Kanda University of International Studies

MOTIVATION, STRATEGY USE, AND PEDAGOGICAL PREFERENCES IN FOREIGN LANGUAGE LEARNING Abstract This chapter reports the results of a survey of motivation, reported use of language learning strategies, and learner preferences for various kinds of pedagogical activities carried out with 2,089 learners of five different foreign languages (Mandarin Chinese, Filipino [Tagalog], French, Japanese, and Spanish) at the University of Hawai‘i. Questionnaire responses were factor analyzed, and a common factor structure was found for the sample as a whole, consisting of the factors of Value (a belief that studying the language is worthwhile for a wide variety of reasons), Expectancy (a combination of self-confidence, self-assessed aptitude for language learning, and lack of anxiety), Motivational Strength, Competitiveness, and Cooperativeness. The fact that many of these students are studying the language of their ethnic heritage also emerged as a distinct motivational factor. Scales based on the factor analyses were used to analyze similarities and differences among groups of students learning the different target languages, as well as relationships between the various components of motivation and those related to learning strategy use and pedagogical preferences. It was found that motivation does indeed affect strategy use and preferences for different types of classroom activities, but some associations are much stronger than others. The motivational factors of Value, Motivational Strength, and Cooperativeness affect strategy use and pedagogical preferences most strongly, while the Heritage Language factor appears to have little or no influence on these variables. Of the different types of learning strategies, the use of cognitive and metacognitive strategies is most affected by motivation, and among the types of pedagogical preferences that we investigated, approval of challenging activities was most affected by motivation.

INTRODUCTION The research reported in this chapter attempts to be integrative, bringing together a number of interests within the field of foreign language learning and teaching that have usually been investigated independently: the structure of foreign language learning motivation in a university setting, the kinds of learning strategies that foreign language learners use, and the kinds of classroom structures and types of activities to which foreign language students react positively. The research was stimulated by both practical and theoretical concerns.

Schmidt, R., & Watanabe, Y. (2001). Motivation, strategy use, and pedagogical preferences in foreign language learning. In Z. Dörnyei & R. Schmidt (Eds.), Motivation and second language acquisition (Technical Report #23, pp. 313–359). Honolulu: University of Hawai‘i, Second Language Teaching and Curriculum Center.

On the practical level, most teachers of second and foreign languages recognize the importance of student motivation, want to know more about it, and want to enhance their students’ motivation in whatever ways are possible. In particular, they hope that curriculum development (including textbook selection) both fosters learning and appeals to students sufficiently. Other issues in the particular learning and teaching context investigated in this study may be unusual but are by no means unique among American universities. The University of Hawai‘i is a large multi-campus university, of which the University of Hawai‘i at MÅnoa is the flagship campus. One of its major strengths is the diversity and depth of its language offerings. Over 30 languages are regularly offered, including not only the major European languages and the major East Asian ones (Chinese, Japanese, and Korean), but also a large number of languages that are offered at few other institutions in the United States, including Vietnamese, Indonesian, Thai, Filipino (Tagalog), Ilokano, and Khmer among the languages of Southeast Asia, and Hawaiian (an official language of the State of Hawai‘i, together with English), Samoan, Tahitian, Maori, and Chomorro among the languages of the Pacific. The offering of these specific clusters of languages is unique to this university, as is the fact that the three “biggest” languages (in terms of student enrollment) are Hawaiian, Japanese, and Spanish (instead of the usual trio of French, German, and Spanish), but some of the issues are not. A striking characteristic of the student population is that about half of all students are studying the language of their ethnic heritage,1 with this rising to close to 100% for some languages. This is true across the United States for the less commonly taught languages (normally defined as every language except French, German, and Spanish), and the only reason that most languages are taught at all in the U.S. is that there are students of those particular ethnicities anxious to study them. As described in other chapters of this volume (Kondo-Brown, Syed), the heritage factor is an extremely important source of language learning motivation, affecting both language choice and persistence in second language learning. A second salient issue at the University of Hawai‘i during the time of this research, not uncommon at U.S. universities, was the existence of a graduation requirement of two years of study of Hawaiian or a foreign language for all undergraduates. With respect to theoretical issues, the research-based literature on motivation for foreign language learning has concentrated almost exclusively on learning second and foreign languages in the more traditional sense, that is, learning a language associated with someone else’s culture rather than one associated with your own or one that you claim as part of your ethnic background. Integrative motivation, for example, has been viewed as an interest and willingness to get to know about someone else’s culture and to 1

The term “heritage learner” has two definitions in current circulation. Some use it to refer to students who are studying their own first language in a school setting. Spanish classes for Spanish speakers fall within this definition. Others use the term more broadly, as we are using it here, to refer to members of ethnic groups who are studying that ethnic language. Some of the students we call heritage learners do speak that language as their mother tongue (with English as their second language), but many more in our sample are second, third, or fourth generation immigrants who in the Hawai‘i context still colloquially refer to their “nationality” (i.e., ethnicity) as Japanese, Chinese, etc.

312 ◆ MOTIVATION AND SECOND LANGUAGE ACQUISITION

interact with members of that group, coupled with a willingness to learn a language to do so. However, the attitudes, orientations, and learning goals of heritage learners, who are learning a language of “inheritance” rather than “affiliation” in Rampton’s (1990) terms might be different. A second theoretical issue concerns the components of motivation in tertiary foreign language education and how these relate to theoretical models in the field. The immediate precursor to this study was a study of foreign language learners of English in Egypt (Schmidt, Boriae, & Kassabgy, 1996) in which application of multidimensional scaling to responses to an extensive questionnaire found three basic dimensions to EFL motivation, labeled Goal Orientation, Expectancy, and Affect. Goal Orientation in that study incorporated items related to a number of different traditionally recognized components of motivation, including both instrumental and integrative orientations. For that population, it did not seem to matter whether learners held integrative or instrumental orientations towards English. What mattered was that some learners were highly oriented towards the goal of learning English for both types of reasons (and others besides), while others attributed little value of any kind to the study of English. The dimension of Expectancy also included a number of factors often considered conceptually distinct, including self-confidence, positive thinking, and determination. Affect was a dimension that included intrinsic motivation for language learning (enjoyment of the activity itself), anxiety (reverse correlated), and a liking for challenge. The major theoretical motivation for this study, however, was not to examine motivation in isolation but to see what links could be discovered between motivation on the one hand and learning strategies on the other. Looking for links between motivation and learning strategies is motivated by a concern with how motivation works. Although some view motivation as part of a general “affective filter” (Krashen, 1985), which if high keeps target language input from reaching the language acquisition device and if low allows input to simply “go in,” we follow Gardner (1985, 1988) in believing that motivated learners achieve higher levels of proficiency because they put more of themselves into learning. Our basic assumption is that motivated learners learn more because they seek out input, interaction, and instruction, and when they encounter target language input they pay attention to it and actively process it however they can (Crookes & Schmidt, 1991; Schmidt, in press; Tremblay & Gardner, 1995). In other words, they use learning strategies. Several previous studies have found links between motivation for language learning and reported strategy use. Oxford and Nyikos (1989) found that motivation was the best predictor of strategy use in a large-scale study of university students. More motivated students used learning strategies of all kinds more often than did less motivated students. In a study of 138 students in first year Italian and Spanish courses, MacIntyre and Noels (1996) found that three variables from the Gardner socio-educational model (Attitudes Toward the Learning Situation, Integrativeness, and Language Anxiety) correlated with three types of strategies: Cognitive, Metacognitive, and Social. A composite variable labeled Motivation correlated with the same three types of strategies, plus memory, compensation, and affective strategies. Schmidt et al. (1996) found that SCHMIDT & WATANABE ◆ MOTIVATION, STRATEGY USE,… ◆ 313

Egyptian EFL learners with high Expectancy (determination), those with strong Instrumental Motivation, and those high in Sociability all reported using active cognitive strategies and organizing strategies. One recent study of learners of French (Gardner, Tremblay, & Masgoret, 1997) found somewhat more perplexing results. While motivation and the use of learning strategies were associated with each other, reported strategy use was not significantly correlated with L2 achievement, and a path analysis suggested that the use of strategies was related to lower levels of achievement. Gardner et al. (1997) suggest that since most of these learners had at least 9 years of prior study of French, successful learners may have adopted a fairly narrow range of effective strategies, while less successful learners may continue to attempt to use a broader range of strategies. Links between motivation and students’ attitudes towards different aspects of language pedagogy have been left largely unresearched. It seems intuitively likely that depending on one’s motivation for learning a language one might prefer different types of pedagogical activities. For example, if one is integratively oriented towards language learning one might prefer communicatively oriented classes over traditional ones. Schmidt et al. (1996) reported that students who scored high on the affect dimension of motivation welcomed communicative classes while those low on that dimension tended to reject the communicative classroom and that determined learners preferred classes in which there was a balance among different skill emphases and between teacher control and learner centeredness. We are not aware of any other research that addresses possible links between motivational factors and how students react to specific aspects of foreign language pedagogy. In light of these concerns, the objectives of the research reported here were to identify the combinations of factors (intrinsic/integrative/instrumental orientations, expectations, anxiety, perceived language aptitude, etc.) that define “motivation” for a university population with many heritage learners and to identify relationships among these motivational factors, reported use of language learning strategies, and preferences for particular types of classroom activities. Because of our desire to gather data from a reasonably large number of students studying a number of foreign languages, survey instruments were developed and used for the research.

METHOD PARTICIPANTS A total of 2,089 participants completed our survey questionnaire, representing students of five different foreign languages: Chinese (Mandarin), Filipino (Tagalog), French, Japanese, and Spanish. In each case, the survey instrument was distributed by instructors who were interested in the project and willing to give up some class time and was completed by students on a voluntary basis. Students were promised confidentiality and assured that their grades would not by affected in any way because of their participation or non-participation. Partly because not all instructors participated, and partly because enrollment in these languages varies greatly, there were larger numbers of responses from 314 ◆ MOTIVATION AND SECOND LANGUAGE ACQUISITION

students of Japanese and Spanish and fewer from Chinese and Filipino. Table 1 shows the distribution of participants by language and year of study at the university level. Table 1: Number of subjects in each language sample language European

Japanese

Filipino

Chinese

language/level Spanish, 1st year Spanish, 2nd year French, 1st year

n 326 315 187

total

1st year 2nd year beyond

466 394 169

1029

1st year 2nd year

69 45

114

1st year 2nd year beyond

45 60 13

828

118 2089

Additional demographic data were gathered concerning these participants but are not being reported here because each participating language department collected different kinds of data as part of the survey. Some requested information concerning student ethnicity and some did not; some wanted to know if the language being studied was spoken by anyone in the student’s immediate family and some did not; some wanted details on learning at the primary and secondary levels and some did not. INSTRUMENTS In a study carried out in Egypt by Schmidt et al. (1996), data were collected from 1,464 adult learners of EFL using a 97-item questionnaire that asked about the students’ motivation, learning strategies, and preferences for classroom instructional activities. That instrument was used as the basis for the one used in this study, with some significant modifications. Items clearly relevant to Egyptian learners of English but not to American learners of other foreign languages were eliminated and additional items were added based on discussions with foreign language teachers at the University of Hawai‘i concerning what they thought were the primary issues concerning the motivation of their students. This resulted in a 123-item questionnaire with questions on motivation, preferences for instructional activities, and learning strategies. The questions concerning motivation were grouped into 11 subcategories, with different scales addressing such aspects of motivation as intrinsic motivation, integrative

SCHMIDT & WATANABE ◆ MOTIVATION, STRATEGY USE,… ◆ 315

orientation, anxiety, and perceived language aptitude. The questions in the second two parts were not initially grouped into subcategories. In the first administration, the questionnaire was administered to students in first and second year Spanish and French courses. After correlational analyses and factor analyses were carried out, 22 items were judged to be ineffective and were removed from the questionnaire. After these analyses, the items concerning motivation were reorganized into separate scales and the questions concerning learning strategies and pedagogical preferences were also grouped into subcategories. In the next administration, the revised 101-item questionnaire was administered to students in all levels of Japanese courses and to students in first semester and third semester Filipino courses. Again, correlational analyses and factor analyses were conducted and 10 items were judged to be ineffective and removed. In the final administration, the revised 91-item questionnaire was administered to 118 second semester, fourth semester, and third year students of Mandarin Chinese. To enable comparisons of the data from different administrations, all the analyses to be reported here are based on the final version of the questionnaire, which consists of 91 items in three parts: Part A, Motivation, 47 items; Part B, Preferences for Instructional Activities, 20 items; and Part C, Learning Strategies, 24 items. Part A contains a single item concerning the language requirement (“I mainly study this language to satisfy the university language requirement”) and questions grouped into 12 scales. For each item participants were asked to indicate their degree of agreement with the statement on a five point scale (1=strongly disagree, 2=disagree, 3=neutral or no opinion, 4=agree, 5=strongly agree): •

Intrinsic motivation, statements expressing enjoyment of language learning

•

Instrumental orientation, statements concerning the financial, social, or other benefits of learning a language

•

Integrative orientation, statements about being able to interact with members of another cultural group

•

Heritage language orientation, statements concerning the students’ attachment to the language as part of their own identity and cultural heritage

•

Interest in foreign languages and cultures, in general (not a specific language)

•

Task value, that is, the value of the language course

•

Expectancy, statements concerning a student’s belief that s/he will do well and receive a good grade in the course

•

Anxiety, statements concerning test and speaking anxiety

•

Language aptitude, the student’s own perception of her/his aptitude for grammar, pronunciation, and so forth

•

Competitiveness, statements about doing better than other students and getting good grades

316 ◆ MOTIVATION AND SECOND LANGUAGE ACQUISITION

•

Cooperativeness, statements concerning relationships with other students and the teacher and learning in a cooperative environment

•

Motivational strength, statements concerning one’s intention to put one’s best effort into learning the language, keep up with the course, etc.

Part B contains 20 items related to the pedagogical preferences of language learners, also rated using a 5-point scale from strongly disagree to strongly agree, grouped into five categories (to be discussed in the results section). Part C contains 24 items concerning the use of learning strategies, also rated on a 5-point scale, grouped into four categories (to be discussed under results). The questionnaire items appear in Appendix A, together with the names of the scales to which they belong and brief labels used for easy reference in the tables that follow. Data were collected by the teachers of the many different language course sections that participated in this study over a period of 2 years, for one target language group at a time, always between the second and fourth weeks of a given 14 week semester, that is, after class procedures had been established and students were beginning to get to know one another but before pressures started building in preparation for mid-term examinations. The data were machine scored, and negatively worded items were re-coded positively by the two authors of this paper. Factor analyses, correlational analyses, and ANOVAs were performed using SAS 6.11 (1989). and StatView 4.5 (1994). in order to address the research questions.

RESULTS DIMENSIONS IN THE DATA Responses to each of the three parts of the questionnaire were factor analyzed to determine relevant dimensions in the data. Because factor analysis requires a large sample size, the data from French and Spanish classes were combined into a category of European languages (n=828). The data from students of Japanese (n=1029) were factor analyzed separately. Factor analysis was not conducted with the data from students of Chinese or Filipino because of their relatively small sample size (n=114 and n=118 respectively). Instead a final factor analysis on the data from all five language groups (n=2089) was conducted. All the factor analyses described below were based on a common factor analysis called Iterated Principle Factor Analysis used in SAS (version 6.11) with varimax rotation. Common factor analysis, unlike principal component analysis, uses only the common variance between items. It excludes the variance that is unique to each item, or error variance. Common factor analysis was used because we were interested in finding or validating psychological traits that are manifested in multiple items in our questionnaire. Both oblique rotations (by promax) and orthogonal rotations (by varimax) were carried out. Oblique rotations allow correlation between factors, which usually produces more easily interpretable results, a “simple structure.” However, orthogonal rotations have an advantage of obtaining “clean” factors uncontaminated by the overlap between factors. Since the results were

SCHMIDT & WATANABE ◆ MOTIVATION, STRATEGY USE,… ◆ 317

practically identical between the two rotation methods in terms of extracted factors, results from varimax rotations will be reported. The number of factors was determined by the eigenvalue (minimum of one), the scree plot, and the interpretability of rotated factors. In order for the dimensions in the data (i.e., categories) to be deemed valid, they need to demonstrate certain properties: •

Convergent validity. All the items in each category need to measure one and the same construct. Therefore, the category needs to be internally consistent. We can check this by Cronbach alpha, which is a way of measuring the correlations between items.

•

Divergent validity. Each category needs to be distinct from other categories. Categories need to be distinct enough from each other to be called separate categories. Factor analysis addresses this problem.

•

Content validity. Meeting the above criteria by statistical analysis is not enough. The naming of each category depends on judgments as to what it represents. Item selection and writing need to be based on theory. This is also necessary to obtain categories that are more likely to meet the criteria of convergent and divergent validity.

The selection of items and preliminary categories for Part A, Motivation, were based on relevant theories and past research findings, addressing the content validity criterion. The results of factor analyses and reliability estimates determine the extent to which the questionnaire satisfies the requirements of convergent validity and divergent validity. For Part A, factor solutions were very similar across European, Japanese, and the combined data sets. Tables 2, 3 and 4 summarize factor solutions obtained and the total variance accounted for. In each case, two solutions are presented with different numbers of factors specified. The number of subjects indicated are those remaining after list-wise deletions for missing data.

318 ◆ MOTIVATION AND SECOND LANGUAGE ACQUISITION

Table 2: Factors for Part A – European languages (N=790) number of factors=6 (total variance accounted for=41%) f1. Value components f2. Expectancy, Anxiety f3. Motivational strength f4. Aptitude, Competitiveness f5. Heritage f6. Cooperativeness

number of factors=5 (total variance accounted for=39%) f1. Value components f2. Expectancy components f3. Motivational strength f4. Competitiveness f5. Heritage

Table 3: Factors for Part A – Japanese (N=1,007) number of factors=7 (total variance accounted for=40%) f1. Value components f2. Expectancy, Anxiety f3. Aptitude f4. Motivational strength f5. Competitiveness f6. Heritage f7. Cooperativeness

number of factors=5 (total variance accounted for=37%) f1. Value components f2. Expectancy, Anxiety, (Aptitude) f3. Motivational strength, Cooperativeness f4. Competitiveness, (Aptitude) f5. Heritage

Table 4: Factors for Part A – All language groups (N=2,023) number of factors=7 (total variance accounted for=41%) f1. Value components f2. Anxiety f3. Expectancy, Aptitude f4. Motivational strength f5. Heritage f6. Competitiveness f7. Cooperativeness

number of factors=5 (total variance accounted for=38%) f1. Value components f2. Expectancy components f3. Motivational strength f4. Competitiveness f5. Heritage

Table 5 shows the factor solution for the entire sample when five factors are specified for motivation. The most striking fact about Table 5 is that a large number of items that are conceptually rather different (such as instrumental orientation and integrative orientation) do not constitute separate factors but emerge as part of a single factor which we have labeled Value. A total of 20 questionnaire items load on this factor, including all items from our scales for intrinsic motivation, instrumental orientation, integrative orientation, task value, and interest in foreign languages and cultures. That is, it is not the case that some of the learners in our sample are instrumentally oriented towards language study while others are integratively oriented, others have a general interest in SCHMIDT & WATANABE ◆ MOTIVATION, STRATEGY USE,… ◆ 319

languages and cultures, and yet others just enjoy language learning. Instead, we find that our learners either see value in learning the foreign language they are studying for all of those reasons or for none of them. Table 5: Factor solution for Part A — All language groups (N=2,023) f1 Value intrinsic 1 intrinsic 1 intrinsic 3 intrinsic 4 intrinsic 5 intrinsic 6 lang req instrumental 1 instrumental 2 instrumental 3 integrative 1 integrative 2 integrative 3 interest 1 interest 2 interest 3 interest 4 task value 1 task value 2 task value 3 heritage 1 heritage 2 expectancy 1 expectancy 2 expectancy 3 anxiety 1 anxiety 2 anxiety 3 anxiety 4 anxiety 5 anxiety 6 aptitude 1 aptitude 2 aptitude 3

f2 Expectancy

f3 f4 Mot str Competitive

f5 Heritage

68* 58* 46* 55* 63* 73* –68* 43* 49* 49* 70* 57* 58* 57* 36 69* 62* 59* 46* 54* 20 38 24 8 2 –14 –6 –8 –11 –8 –15 18

31 36 20 19 37 21 –23 2 –3 6 6 9 5 15 5 17 5 26 2 10 –5 6 53* 59* 72* –45* –64* –63* –44* –47* –56* 37

25 34 19 14 17 10 –6 2 8 –3 13 2 9 5 17 13 17 30 48* 8 10 10 16 20 –7 –2 –15 5 –1 –3 –3 6

8 6 3 18 1 –10 16 30 14 14 12 18 14 9 9 2 18 3 6 8 2 14 20 29 –1 –1 8 7 –2 15 2 24

11 3 4 9 3 14 –13 12 10 21 13 24 19 –9 –13 4 0 5 2 –4 82* 74* 2 –1 2 –2 1 0 –8 6 6 8

11 7

39 40*

0 3

29 21

–8 –6

320 ◆ MOTIVATION AND SECOND LANGUAGE ACQUISITION

aptitude 4 compete 1 compete 2 compete 3 compete 4 cooperate 1 cooperate 2 cooperate 3 motiv’l str 1 motiv’l str 2 motiv’l str 3 motiv’l str 4 motiv’l str 5

27 –1 32 7 10 17 15 16 28 0 15 4 29

46* –19 6 21 10 –1 –15 –5 21 –7 29 14 24

10 30 3 –6 19 29 25 24 43* 52* 40* 54* 58*

32 38 43* 42* 45* 11 32 20 –12 7 –7 11 10

2 6 11 4 0 –1 –4 6 6 2 –2 4 12

note: Lines denote divisions between major categories. Values are multiplied by 100 and rounded to the nearest integer. Values greater than 0.4 have been flagged by an “*”. Variance explained by each factor: factor1 factor2 factor3 factor4 factor5 7.507308 4.567245 2.419648 1.721466 1.558590

A factor labeled Value shows up in each of the analyses shown in tables 3, 4, and 5, and the specific items that load on this factor are stable across all the analyses. The same is true of Factor 5, which we have labeled the Heritage factor and which consists of only the two items on our heritage language learning scale. That is, although an integrative orientation towards a language associated with another social group and culture did not emerge as a distinct motivational factor for this population, an orientation towards learning the language of one’s own cultural heritage did. In each analysis, factors that we have labeled Expectancy, Motivational Strength, and Competitiveness also emerged, but there are some differences in which items load on these factors, depending on the target language sample and how many factors are specified for a solution. For example, items concerning anxiety appear to constitute a separate factor if a seven factor solution is chosen for the whole sample, but items related to anxiety load (negatively) on the Expectancy factor in all other solutions. A separate factor of Cooperativeness was formed only when the number of factors was specified as seven (for Japanese and all languages combined) or six (for pooled French and Spanish data). For Part B (pedagogical preferences), unlike for Part A, no predetermined categories were established, and all of the categories were derived from factor analysis. Factor solutions were almost identical across the three data sets, as shown in tables 6, 7, and 8. Table 9 shows the factor solution for the entire sample when five factors are specified for pedagogical preferences. Table 6: Part B – All language groups (N=2023) number of factors=5 (total variance accounted for=40%)

SCHMIDT & WATANABE ◆ MOTIVATION, STRATEGY USE,… ◆ 321

f1. Practical f2. Cooperative f3. Challenging f4. Traditional f5. Innovative Table 7: Part B – Japanese (N=1007) number of factors=5 (total variance accounted for=41%) f1. Practical f2. Cooperative f3. Challenging f4. Traditional f5. Innovative Table 8: Part B – European (N=790) number of factors=5 (total variance accounted for=40%) f1. Practical f2. Challenging f3. Cooperative f4. Innovative f5. Traditional Table 9: Factor solution for Part B — All language groups (N=2023) f1 f2 f3 f4 f5 Practical Cooperative Challenge Innovative Traditional grammar read/write vocabulary relevant ask Q’s pronunciation

23 26 43* 39 58* 37

6 6 4 7 15 2

16 7 7 –2 10 22

9 16 12 30 5 25

69* 52* 38 22 13 21

listen/speak communicate feedback everyday lg no English

53* 59* 64* 51* –7

6 12 12 9 –14

38 22 1 –2 41*

21 21 5 22 19

16 14 16 9 12

322 ◆ MOTIVATION AND SECOND LANGUAGE ACQUISITION

challenge active partic must speak pairs/group not alone cohesiveness culture goal-setting authentic

23 21 12 13 6 26 11 22 10

6 39 14 72* 73* 44* 11 2 3

46* 48* 67* –3 12 1 7 3 17

32 27 –8 15 –7 32 49* 35 53*

26 1 2 10 –1 8 13 3 8

note: Lines denote divisions between major categories. Values are multiplied by 100 and rounded to the nearest integer. Values greater than 0.4 have been flagged by an “*”. Variance explained by each factor: factor1 factor2 factor3 factor4 factor5 2.489048 1.532095 1.392704 1.326841 1.194866

Five factors related to learner preferences for different aspects of foreign language pedagogy were extracted. The largest number of our questionnaire items load on Factor 1, which we have labeled a Practical approach because of the focus in these items on everyday, communicatively relevant language, feedback from the teacher, and the belief that students should ask questions when they do not understand something. Factor 2 consists of three items concerning group and pair work, not working alone, and a desire for cohesiveness in the language class. We have labeled this a preference for Cooperative Learning. Factor 3 clearly represents the dimension of Challenge and has been so labeled. Factors 4 and 5 are somewhat harder to label. Factor 4 consists of only three items, a focus on culture in the language classroom, the use of authentic materials, and an endorsement of goal-setting by language learners. We have labeled this Innovative for lack of a better term. Factor 5, which we have labeled Traditional, is made up of two core items concerning a focus on grammar and reading and writing. However, in the Japanese data the item concerning vocabulary also loaded on this factor, while in the data from learners of European languages this item loaded most highly on Factor 1, while also loading relatively highly on Factor 5. This makes some sense in terms of what vocabulary learning likely means for students of these languages. For Americans learning a European language that uses the same script as English, learning new vocabulary means primarily matching new phonological forms with familiar meanings and is likely to be viewed primarily as an important factor in becoming communicatively competent. For Americans learning Japanese, however, learning vocabulary is often viewed primarily as a matter of learning new kanji (Chinese characters used in written Japanese) and would therefore be associated with reading and writing and other aspects of traditional pedagogy. In the end, we decided to retain this item within our measure of a preference for a traditional approach, mainly to increase the number of items in this category and thereby increase its internal consistency. For Part C (learning strategies), all the categories were also derived from factor analysis. The factor solutions were very similar across the three data sets. Tables 10, 11, and 12 summarize the factor solutions obtained for three different samples and the total

SCHMIDT & WATANABE ◆ MOTIVATION, STRATEGY USE,… ◆ 323

variance accounted for. Table 13 presents a factor solution for the combined data, specifying four factors. Table 10: Part C – All language groups (N=2023) number of factors=4 (total variance accounted for=31%) f1. Study skills f2. Cognitive f3. Coping f4. Social Table 11: Part C – Japanese (N=1007) number of factors=4 (total variance accounted for=31%) f1. Study skills f2. Coping f3. Cognitive f4. Social Table 12: Part C – European (N=790) number of factors=4 (total variance accounted for=33%) f1. Study skills f2. Cognitive f3. Social f4. Coping

324 ◆ MOTIVATION AND SECOND LANGUAGE ACQUISITION

Table 13: Factor solution for Part C — All language groups (N=2023)

relate vocab compare lgs guess meaning patterns evaluate progress preview clarification work with others classmates ask other’s help organize study review after test place for study time to prepare re-read materials review early not last minute periodic study repeat words find gaps see words first look words up try to keep up what’s on test?

f1 f2 Study skills Cognitive 20 50* –2 51* 3 50* 20 53* 33 35 24 32 21 35 15 8 11 10 2 0 39 20 38 22 33 18 47* 20 38 14 74* 4 72* –10 41* 17 23 27 17 20 4 12 30 19 13 –2 –4 –8

f3 Coping 20 5 –1 9 23 23 16 8 10 24 31 22 17 18 34 10 –13 5 34 47* 35 31 41* 45*

f4 Social 3 6 4 –5 4 12 14 70* 72* 61* 18 11 3 6 4 11 3 5 6 14 10 2 7 8

note: Lines denote divisions between major categories. Values are multiplied by 100 and rounded to the nearest integer. Values greater than 0.4 have been flagged by an “*”. Variance explained by each factor: factor1 factor2 factor3 factor4 2.521515 1.807837 1.538342 1.533918

As can be seen from Table 13, of the factors extracted for learning strategies, some are more easily interpretable than others. Factor 4, with high loadings from just three questionnaire items concerning working with others and seeking help from classmates, clearly represents social learning strategies. Factor 1, which we have labeled Study Skills, includes a coherent set of items concerning the methodical allocation of resources to getting the job of studying a language accomplished. We have labeled Factor 2 Cognitive, although we note that some of the items loading on this factor would be considered metacognitive strategies in a theoretically oriented scheme. Factor 3 is hardest to interpret, since some items loading on it do so weakly, but the high loading of items such as wanting to know what is on an upcoming test and trying to keep up with the course suggests that these are coping strategies.

SCHMIDT & WATANABE ◆ MOTIVATION, STRATEGY USE,… ◆ 325

INTERNAL CONSISTENCY Internal consistency estimates, by Cronbach’s coefficient alpha, of the three parts and of categories within each part for four language samples are reported in Table 14. Coefficient alpha ranged from .92 to .88 for Part A (47 items), from .84 to .81 for Part B (22 items), and from .83 to .79 for Part C (24 items). These estimates are acceptably high. However, coefficient alpha for categories and subcategories within each part varied greatly. Generally, the more items a category contains, the higher the reliability estimate was. This was true for value components (22 items, coefficient alpha=.93 to .86) and for expectancy components (13 items, coefficient alpha=.86 to .84) on one end, as well as for competitiveness (4 items, alpha=.55 to.39) and cooperativeness (3 items, alpha=.51 to .35) on the other. This is not surprising since internal consistency estimate is a function of the number of items and inter-correlations within a category. The more items a category contains and the more they are correlated with each other, the internal consistency estimate goes up. Due to the small number of items, the internal consistency estimates for some categories are only marginally acceptable. Table 14: Cronbach’s coefficient alpha reliability estimates categories Part A Value components Intrinsic motivation Instrumental orientation Integrative orientation Interest Task value Expectancy components Expectancy Anxiety Perceived aptitude Heritage language Motivational strength Competitiveness Cooperativeness Part B Traditional approach Practical proficiency orientation Challenging approaches Cooperative learning Innovative approaches

# of items Japanese

European

Filipino

Chinese

47 22 6 3

.91 .91 .84 .42

.92 .93 .86 .64

.88 .86 .74 .44

.91 .90 .80 .51

3 4 3 13 3 6 4 2 5 4 3 20 5 7

.69 .69 .59 .84 .69 .76 .67 .80 .65 .54 .50 .82 .65 .77

.73 .71 .64 .84 .66 .73 .64 .78 .74 .55 .51 .83 .64 .78

.64 .57 .50 .86 .66 .76 .57 .72 .44 .39 .36 .81 .70 .66

.68 .69 .68 .84 .73 .75 .64 .79 .64 .42 .35 .84 .73 .84

4 3 3

.63 .70 .49

.62 .65 .53

.49 .69 .38

.51 .53 .38

326 ◆ MOTIVATION AND SECOND LANGUAGE ACQUISITION

Part C Cognitive strategies Social strategies Study skills Coping strategies

24 7 3 8 6

.82 .69 .73 .75 .52

.83 .69 .73 .74 .64

.80 .63 .72 .62 .50

.79 .60 .71 .72 .63

Overall, this questionnaire, with only 91 items, attempts to tap into three different aspects of individual differences among foreign language learners: motivation, preferences for instructional activities, and strategy use. Considering the small number of items, particularly within categories in each of the three parts, this instrument seems to serve its purpose fairly well. Factor analyses and internal consistency estimates indicate that the following categories meet the divergent and convergent validity criteria: for Part A—Value, Expectancy, Heritage Language, Motivational Strength, Competitiveness, and Cooperativeness; for Part B—Traditional Approach, Practical Proficiency Orientation, Challenging Approaches, Cooperative Learning, and Innovative Approaches; and for Part C—Cognitive Strategies, Social Strategies, Study Skills Strategies, and Coping Strategies. In order to be able to use these factors in further analyses, we have computed composite scores for each them, by summing the scores of the items loading on each factor. In the case of the expectancy scale, the scores for anxiety items were reversed to indicate a lack of anxiety. In addition, we include in our following analyses the responses to the single item concerning the subjective importance of the language requirement, since this is of interest to the participants and their instructors. Appendix B lists the scales and the items that have been summed to arrive at a scores for further analysis. COMPARISONS AMONG STUDENTS STUDYING DIFFERENT LANGUAGES Although not the major focus of this study, one question of interest is whether students studying different languages in a single institution are all more or less alike or whether the students of each language present a particular motivational profile. We also wondered whether students studying different languages might use different learning strategies or have preferences for different styles of foreign language pedagogy. It is often argued by foreign language teachers that learning their particular language requires a specific pedagogical approach or even that specific strategies may be especially useful for specific languages. To cite only one example, teachers of Japanese and Chinese as foreign languages often argue that because of the importance of learning large numbers of characters (kanji) for these languages, memorization strategies are crucial. The question we are asking here is whether students of these languages subscribe to similar beliefs or modify their strategies and preferences depending on the language they are learning. Tables 15–18 display the results of each of the target language groups on each of these variables. In each case, an ANOVA procedure was used to determine whether there are statistically significant differences among the groups, followed by a post-hoc comparison of means (using the Scheffé test) to identify precisely where the differences

SCHMIDT & WATANABE ◆ MOTIVATION, STRATEGY USE,… ◆ 327

lie. Because of the large number of statistical tests reported in this and the following sections, alpha was set at .001. Table 15 shows the results of the analysis of variance on the scores on parts A (motivation), B (pedagogical preferences), and C (learning strategies) by target language group. As can be seen in the table, there are group differences on each of these measures, but Scheffé’s test indicates that only a few of the pair-wise differences in means are significant. Learners of Spanish in our sample are, on the whole, less motivated than the learners of Chinese, Japanese, or Filipino. The students of Filipino have, overall, a higher level of appreciation for all aspects of language pedagogy than students of either French or Spanish, and Japanese learners also have a higher level of appreciation of pedagogical techniques (in general) than the learners of Spanish. Learners of Filipino report the highest use of strategies (all types combined) and their means on this part of our instrument are significantly higher than those of the learners of both Chinese and Japanese. Table 15: Analysis of variance of scores on Parts A, B, and C by target language group French mean

Part A Part B Part C

3.300 3.773 3.526

SD

Spanish mean

.530 3.181 .479 3.762 .442 3.562

SD

Japanese mean

.503 3.421 .401 3.884 .419 3.532

SD

Chinese mean

.469 3.494 .395 3.884 .430 3.386

SD

Filipino mean

.456 3.538 .437 3.989 .400 3.716

SD

F-value p-value

.425 32.783