You are on page 1of 19

AMBIGUITY TOLERANCE AND FIELD INDEPENDENCE AS PREDICTORS OF PROFICIENCY IN ENGLISH AS A SECOND LANGUAGE Carol Chapelle Iowa State University

Cheryl Roberts Columbian-American Center, Medellin


The fact that some adults are more successful at acquiring an L2 than others has led to investigations of individual characteristics as predictors of successful L2 acquisition. This paper reports the results ofan investigation of the relationship between two learner characteristics, Ambiguity Tolerance (AT) and Field Independence (FI), and adult learners' acquisition of English as a Second Language in the United States. A Multiple Regression Analysis revealed that AT and F1 accounted for a significant amount of variance on several end-of-semester language measures beyond that which could he accounted for by beginning-of-semester performance or other variables. The implications of these findings for further research are outlined.

APTITUDE FOR SECOND LANGUAGE ACQUISITION


In spite of attempts made to improve second language (L2) instruction, the fact remains that some adults are more successful at acquiring an L2 than others. In the past, researchers sought to explain this variance among

'This paper was presented at the 18th annual TESOL convention, March, 12-17, 1984 in Houston. Texas. Biographical Information: Carol Chapelle is an assistant professor of ESL at Iowa State University in Ames, Iowa. Cheryl Roberts is director of courses for the Columbian-American Center in Medeliin, Columbia. Acknowledgements: We are grateful to the following people for their assistance with this research: Lyle Bachman, Rebecca Brewer, Joan Jamieson, John Lett, Alice Omaggio, Sandra Savignon, Rand Spiro, and Jerry Walker. We are particularly indebted to H. Douglas Brownforhis wisdom, guidance and patience throughout the project.

27

28

Language Learning

Vol. 36, No. I

learners in terms of teaching methods (Chastain 1969), intelligence (Pimsleur, Mosberg and Morrison 1962), analytic language skills, referred to as foreign language aptitude (Carroll and Sapon 1959), attitude (Gardner and Lambert 1972), cognitive variables (Naiman, Frolich and Stern 1975), and social factors (Schumann 1976). Although these perspectives toward L2 aptitude have made important contributions to our overall understanding of the process of L2 acquisition, much work remains to be done before statements can be made about which combination of factors is ultimately crucial to L2 acquisition in a particular situation. To that end, we continue to quantify student characteristics, assess learning situations, and examine the skills that are required of students on particular L2 tasks. The research reported here provides further evidence that cognitive/ affective variables are components of L2 aptitude. Moreover, this evidence extends conclusions made about Anglo foreign language learners to international second language learners. Two variables, Field Independence and Ambiguity Tolerance, were found to be significant predictors of ESL proficiency for international students studying in the United States.

FIELD INDEPENDENCE
Field Independence/ Field Dependence (FI/ D), a cognitive variable, is defined as the extent to which a person perceives part of a field as discrete from the surrounding field as a whole, rather than embedded, o r . . . the extent to which a person perceives analytically (Witkin, Moore, Goodenough, and Cox 1977:7). An FI person may approach problem solving situations analytically, while a F D person may approach them in a more global way. In the area of intellectual problem solving, a highly FI person is able to detect patterns and subpatterns, while a F D person tends to get lost in the totality of the stimuli. Consequently, a FI person is at an advantage in problem solving situations in which isolating and manipulating a critical element is important (such as word problems in mathematics) (Witkin et al. 1977). A F D person, on the other hand, is more capable of perceiving the total picture in a situation. Thus, the relationship between F1/ D and success on a given task should depend on the nature of the task. FI/ D can also be described in terms of age, sex and culture. For western societies, FI/ D follows a developmental curve. Children become

Chapelle and Roberts

29

increasingly FI until about the age of 15. FI then stabilizes until approximately age 30 when it gradually begins to decrease. In addition, in western societies, consistent differences between men and women have been found: men tend to be slightly more FI than women (Witkin et al. 1977). FI/ D has also been found to be related to societal and cultural variables. For example, Hopi children in the southwestern United States are more F D than other American children (Witkin et al. 1977); Mexican-American children from traditional Mexican homes are more FD than MexicanAmericans from non-traditional homes (Ramirez 1973; Ramirez, Castaneda, and Herold 1974). From these and other comparisons (reported in Witkin et al. 1977) of FI/ D in different cultures, it has been concluded that cultures with more elaborate social structures and pressure to conform tended to have children who were more F D whereas cultures in which technology played an important role and individual freedom was stressed produced children who were more FI. Further evidence for this hypothesis has been presented recently (Hansen 1984): American high school students in Hawaii were found to be significantly more FI than their counterparts from Polynesian cultures.

FI and L2 Acquisition
A relationship between FI and L2 acquisition can be proposed if the different situations in which L2 acquisition takes place are kept in mind. A FI person may be good at language skills such as those employed in a classroom in which an analytic-type method is used. In this environment, the student is asked to learn rules about the language and apply those rules to language use. The FI persons ability to approach problem solving analytically and impose structure on a situation should facilitate learning in this type of classroom environment. The analytical FI type should be good at language learning, i.e., learning rules, finding patterns, and organizing to make generalizations. The F D person, on the other hand, may be better at L2 acquisition, i.e., acquiring the language through integrative language use such as interaction with native speakers in social situations. It has been suggested that the ideal language learner would be one who was cognitively flexible so that he/she could take advantage of both his/ her FI and FD styles, and thereby benefit from the use of both L2 learning and acquisition situations (Brown 1980:93).

30

Language Laarning

Vol. 36, No. 1

Several studies have attempted to provide evidence for a relationship between FI/ D and L2 acquisition. One study using English-speaking Canadian students learning French found that FI was a significant predictor of L2 proficiency as measured by an imitation test which required the student to repeat a sentence in French (Naiman, Frohlich and Stern 1975). Two other studies in the same environment found that 1) FI, along with ethnocentrism and French class anxiety, predicted success on a general French achievement test (Tucker, Hamayan and Genesee 1976) and 2) FI showed a significant positive correlation with achievement on French reading, listening, writing and grammar tests although FI added nothing to the prediction that could not be accounted for by other variables. (Bialystok and Frolich 1978). Another investigation of Anglo students found that FI was related to better performance in a formal Spanish course (Hansen and Stansfield 1981). Three aspects of L2 competence comprised the measure of performance in this college freshman level course: linguistic, communicative and integrative. In a study of adult ESL students (Seliger 1977) a correlation was found between a sentence disambiguation test, and FI, thus supporting the hypothesis that FI is related to successful L2 learning insofar as the ability to disambiguate sentences is a measure of L2 proficiency. In another study using ESL students (Day 1984), correlations done on the data given indicated a significant relationship between FI and performance on a cloze test (r = .259, p < .05) but not between FI and performance on a test of communicative competence (. 108). Another study examining primarily ESL students found significant positive correlations between FI and several language proficiency measures (Hansen 1984). Overall, then, the research supports the hypothesis that FI is related to L2 acquisition as measured by a number of different proficiency tests; there has been no empirical evidence that indicates F D learners have an advantage in some cases.

AMBIGUITY TOLERANCE
Ambiguity Tolerance (AT) can be defined as a persons ability to function rationally and calmly in a situation in which interpretation of all stimuli is not clear. People who have little or no AT perceive ambiguous situations as sources of psychological discomfort or threat (Budner 1962). These feelings may cause them to resort to black-and-white solutions

Chapelle and Roberts

31

(Frenkel-Brunswik 1949), and refuse to consider any grey aspects of a situation. They may also strive to categorize phenomena rather than order them along a continuum (Levitt 1953); moreover, they may arrive at premature closure (Frenkel-Brunswik 1949) or jump to conclusions rather than take time to consider all of the essential elements of an unclear situation. People with little AT may also try to avoid ambiguous situations. People who have a great deal of AT, on the other hand, enjoy being in ambiguous situations and, in fact, seek them out. They are believed to excel in the performance of ambiguous tasks (MacDonald 1970). Four types of ambiguous situations have been defined: novel situations, in which there are no or insufficient familiar cues; complex situations, in which there are too many cues to take into account; insoluble situations, in which the cues suggest different structures (Budner 1962); and unstructured situations, in which the cues cannot be interpreted (Norton 1975).

AT and L2 Acquisition
An L2 situation can be considered ambiguous because of the characteristics it shares with each of the four kinds of ambiguous situations. An L2 situation is considered novel by learners because the grammatical, lexical, phonological and cultural cues are unfamilar and therefore insufficient for them to construct a meaningful interpretation. On the other hand, these cues may be perceived as being too numerous to interpret, resulting in a complex situation. Similarly, a learner may interpret these multiple language cues as contradicting each other, rendering the situation insoluble. Also, because language cues in many cases cannot be interpreted by the learner, the situation can be perceived as unstructured , Of course, L2 situations vary with respect to the amount of ambiguity present in each kind. Although ambiguity is present in any L2 situation, there is less in a formal language class in which individual elements of language are isolated for study than in an immersion situation in which the learner has to attend to all language cues simultaneously. A learner with little AT would be at a greater disadvantage in a situation with greater ambiguity. Two studies have attempted to relate AT to success in L2 acquisition. The first study (Pimsleur et al. 1966), using Walks A scale, found no relationship between the two factors; however, the significant problems

32

Language Learning

Vol. 36, No. 1

with this scale indicate that the results it produces might be quite distorted (Ehrlich 1965). AT was also used as one of the variables under investigation in a study that attempted to identify some of the characteristics of good language learners (Naiman et al. 1975). In this study, it was found that AT was positively related to L2 success, as measured by an imitation task and a listening task. Their subjects were students in a foreign language situation: high school students learning French in Toronto. It was reported:
...those students who have a high intolerance of ambiguity may have a great deal of difficulty dealing with the amount of ambiguity present in the second language classroom, and therefore may drop the subject as soon as possible. The possibility of such an occurrence was strengthened by the fact that tolerance of ambiguity was a significant predictor of success only in Grade 8. Grade 10 and Grade 12 students were both significantly more tolerant of ambiguity than Grade 8 students (pp. 259-260).

Thus, the small amount of research that has been done indicates that when a valid measure of AT is used and there is a large variance within the group tested, the relationship between AT and L2 proficiency is positive. This finding is consistent with the intuitive notion that a person who is more comfortable with ambiguity will function better in a n L2 environment.

LEARNERS AND LEARNING SITUATIONS


It is important to note that much of the previous research on AT and FI in L2 acquisition has been carried out in a foreign language (FL) situation using native speakers of English as subjects. Because of the impact of social, cultural and linguistic variables on L2 acquisition, caution must be taken in generalizing findings across language situations and language groups of the learners. The learners affective/ cognitive factors should be more important in an L2 situation in which they havea greater opportunity to take advantage of the L2 environment if they choose to do so. The results of the research presented here allow for generalization of previously supported hypotheses across cultures and L2 situations because the subjects came from non-Anglo cultures and were studying ESL in the United States. Another crucial element in identifying characteristics that may be important in L2 proficiency is the definition of the characteristics that are related to particular L2 tasks. For example, it would be logical to find that

Chapelle and Roberts

33

students with little AT were very poor in unstructured communicative situations, but had no more trouble than their high AT counterparts in more structured classroom activities. In studying L2 aptitude, conclusions must be drawn in view of the criterion measures (of L2 proficiency) that are used. The research reported here provides evidence that FI and AT are predictors of proficiency based on several L2 measures: the TOEFL, a multiple choice grammar test, a dictation test and an oral speaking test of communicative competence.

METHOD Subjects
The subjects were 61 adult international students who were enrolled in the Intensive English Institute at the University of Illinois during the Fall of 1982. They represented three linguistic groups: Japanese (n = 13), Spanish (n = 28) and Arabic (n = 20). All of the students in the Institute from these 3 linguistic groups were asked to participate in the study through a letter that had been translated into their native languages.

Variables and Measures


Field Independence. The Group Embedded Figures Test (Witkin et al. 197 1) was used as a measure of FI. Since its development, numerous studies have used this test as a measure of FI, including the L2 acquisition research cited above. It consists of a booklet that contains 18 pictures of complex figures within which simple figures are embedded (see Figure 1). The subject is asked to find a given simple figure in each of the complex figures on the test. The subjects ability to find the simple figures without becoming distracted by the complex figure indicates the extent to which he/ she is F1. The subject is given one point for each item that he/ she answers correctly so a subject who gets a high score is one who is FI. Ambiguity Tolerance. The MAT-50, a 62 item Likert-type scale, (Norton 1975) was used to measure AT. The scale consists of statements concerning work, philosophy, art and other topics, with which the subjects are to agree or disagree on a seven point scale. For example, item 30 is: A group meeting functions best with a definite agenda. YES! YES yes ? no NO NO!

34

Language Learning

Vol. 36, No. 1

Complex figure

Simple figure

Figure I . A sample item from the Group Embedded Figures Test (Witkin et al. 1971)

Chapelle and Roberts

35

A subject who answers this and other similar statements with a YES! would get a low score on the total AT test. This AT test, which has not been used in previous L2 research, was chosen because it was demonstrated to have higher reliabilities and was subjected to a more rigorous set of tests to establish evidence for validity than previously used measures. The criterion-referenced validity procedure included correlations of the MAT-50 with the measure of AT used in the Naiman et al. research; these correlations were significantly positive. English Class Anxiety. Students anxiety in their English class was measured by a portion of the Attitudes and Motivation Test Battery (AMTB) (Gardner and Smythe 1979). This scale was chosen because its authors have provided evidence for the validity and reliability of each of the subtests in the measure. Motivational Intensity. Students Motivational Intensity for learning English was also measured by a sub-scale of the AMTB. English Proficiency. Students English proficiency was measured at the beginning and the end of the semester by several English tests: TOEFL, a multiple choice grammar test, cloze and dictation tests and an oral test of communicative competence. The grammar, cloze and dictation tests were those which are routinely used for placement purposes. The oral test of communicative competence (Bachman and Palmer 1984) was developed and validated on the basis of the theoretical model of communicative competence presented by Canale and Swain (1980); therefore, it included measurement of three general competence areas: grammatical competence, pragmatic competence, and sociolinguistic competence (see Bachman and Palmer 1984).

Procedure
The subjects were tested three times during the course of the semester. At the beginning of the semester, they were given the routine placement tests, the TOEFL, and the oral test of communicative competence. In the seventh week of the semester, they were given the tests of FI, AT, English Class Anxiety (ANX) and Motivational Intensity (MOT), which had been translated into their native languages through a process of translation and back translation. At the end of the semester, all of the English tests were administered again. The data was analysed using SPSS (Nie et al., 1975) to perform three analyses. First, a Multiple Regression Analysis using language group as a

36

Language Learning

Vol. 36, No. I

predictor variable was done to determine whether it was a significant predictor of any of the cognitive/ affective measures. This analysis revealed that the sample could not be treated a s a homogeneous group; the Japanese students were significantly different on all predictors so they were omitted for subsequent analysis. Second, Pearson product-moment correlations were calculated t o find relationships among the affective/ cognitive measures and the language tests. Finally, Multiple Regression Analyses were done using end-of-semester language proficiency scores as the criterion measures and the affective/ cognitive measures as predictors.

RESULTS
Correlations
Pearson product-moment correlations were done using the affective/ cognitive variables as predictors and the language proficiency measures as criterion variables (Tables 1 and 2). There were significant correlations between FI and MOT and all of the language measures at the beginning and the end of the semester. Those students who were highly FI did better on all of the language measures; moreover, the correlation between FI and the end of semester scores were typically stronger than those between FI and the beginning of semester scores, thus indicating that the FI students were more likely to score higher on the proficiency measures after a semester of L2 study. The correlations between MOT and all of the language measures were significantly negative indicating that students who showed higher levels of Motivational Intensity as measured by the AMTB were those who did not d o well on the language proficiency measures. There are several possible explanations for this unexpected finding. First, the measure asks questions concerning the effort made by students toward their in-class work such as: When it comes to my English homework, I: a. work very carefully, making sure I understand everything. b. put some effort into it but not as much as I could. c. just skim over it. It is possible that those students who put forth a great deal of time and energy doing class assignments neglected to use the opportunities present in an actual L2 environment to use the language in real meaningful contexts. It may be that students who spend time using the language

Chapelle and Roberts

37

Table I Pearson product-moment correlaiions between the affecriveicogniiive measures and rhe whole and part TOEFL scores

Affective/Cognitive Measures Language Measures TFLl TFL2 FI .548*** .750*** .540*** .739*** .482*** .730*** .481*** .731*** AT
,106

MOT -.456*** -.465*** -.395** -.505***


-.449***

ANX
-.I73 -.303**

.237* ,094 .205* .045 .275* ,150 ,200

ST 1 ST2
LC I LC2

-. 124
-.258** -.070 -.328** -.08 1 -.283**

-.412** -.418** -.378**

RDI R D2
***p< ,001 **p < .01
p

< .05
= =

FI AT ANX MOT TFLI TFL2 STI ST2 LCI LC2 RDI ROZ

= =
= = =

=
=

= = =

Field Independence Ambiguity Tolerance English Class Anxiety Motivational Intensity Beginning-of-Semester TOEFL End-of-Semester TOEFL Beginning-of-Semester Srructure part of the TOEFL End-of-Semester Structure part of the TOEFL Beginning-of-Semester Listening part of the TOEFL End-of-Semester Listening part ofthe TOEFL Beginning-of-Semester Reading part of the TOEFL End-of-Semester Reading part of the TOEFL

achieve higher levels of proficiency than those who simply study the language. Another possibility is the fact that the Motivational Intensity measure was validated using Anglo students in a FL situation. Anglo students may respond to questions of effort differently from the way international students do. The correlations between AT and the language measures also followed a pattern. There were no significant correlations between AT and beginning of semester language scores but the correlations between AT and end of semester scores were, in almost every case, significantly positive. This indicates that students level of AT had no relationship to entry proficiency but that after a semester of L2 study, those students who had higher levels of AT tended to attain higher levels of proficiency at the end of the semester. This provides slight evidence for the hypothesis that AT is related to successful L2 study in an L2 environment; those students who are able to

38

Language Learning

Vol. 36, No. 1

Table 2 Pearson product-moment correlations between rhe affective1 cognitive measures and the structure, cloze. dictation and oral speaking rests

Affective/ Cognitive Measures Language Measures


ST I ST2

F1 .600*** .640*** .555*** .600*** .544*** .525** .485*** .533***

AT

MOT -.561*** -.487*** -.561*** -.599*** -.591*** -.581*** -.453***


-.528***

ANX
-. I78

. I30
.250* ,144 .I27
,078 .219*

-.348**

CL I CL2

-.338** -.274**
-.337** -.290**

D1 1
D12

cc 1 cc2
***p<.w1
* * p < 01 * p < 05

.088 ,054

-. I16
-. 183

FI
AT ANX

Field Independence

= Ambiguity Tolerance
= English Class Anxiety = Motivational Intensity = Beginning-of-Semester Structure Test = End-of-Semester Structure Test = Beginning-of-Semester Cloze Test = End-of-Semester Cloze Test = Beginning-of-Semester Dictation Test

MOT
STI ST2

CLI CL2 DII


DI2

CCI cc2

= End-of-Semester Dictation Test = Beginning-of-Semester Oral Test of Communicative Competence = End-of-Semester Oral Test of Communicative Competence

tolerate the ambiguities present in the L2 environment apparently gain more from their L2 study. Similarly, English class anxiety was not significantly related to beginning of semester proficiency, but in all but one case was negatively related to end of semester proficiency. Although the correlations are not very strong, their negative direction indicates that students with low levels of anxiety tended to be those who, after a semester of L2 experience, performed better on the language measures.

Multiple Regression Analysis


As the correlations indicate, FI and AT were related to performance on the end-of-semester language proficiency measures; however, in order to assess the importance of these two variables in view of other factors such as the students' proficiency a t the beginning of the semester and the other

Chapelle and Roberts

39

affective measures, a Multiple Regression Analysis was done. Because the subjects began the intensive English program with a variety of levels of proficiency, the analysis was designed to account for this beginning variance (Linn and Slide 1977). Each of the end of semester scores was used as a criterion variable; the corresponding beginning of semester score was entered into the equation as the first predictor to account for the subjects entry level proficiency (Tables 3 and 4). The next two variables to be entered as predictors were ANX and MOT; finally, ATand FI were entered to determine whether a n y variance in the end of semester scores could be attributed t o these two variables beyond that which could be accounted for by the other three variables. For all but one of the language measures, FI was a significant predictor; for the TOEFL and two of its subparts, AT was a significant predictor as well. There were four variables that were significant predictors of performance on the TOEFL at the end of the semester. First, students performance on the TOEFL at the beginning of the semester was a significant predictor; yet, the beginning of semester scores did not account for all of the end of semester variance (88%). That is, each student did not progress at the same rate during the course of the semester; instead, there were other factors that came into play in end of semester proficiency scores. Those other factors were ANX, FI and AT. ANX accounted for a significant additional amount of the variance and FI and AT accounted for significant variance even beyond that which could be accounted for by the other variables. The parts of the TOEFL followed a pattern similar to that of the total TOEFL; differences were found in the Structure sub-part for which ANX and MOT were not significant predictors; Reading for which neither AT nor MOT were significant predictors; and Listening for which MOT was not a significant predictor. Overall, the results indicate that Ft and AT are important in L2 acquisition, as measured by the total TOEFL. In particular, A T is important in structure and listening, while FI is a significant predictor of all the skills measured. The patterns of predictors on the other language proficiency measures were less regular than for those for the TOEFL. On the structure and dictation tests, FI was the only significant predictor beyond beginning-ofsemester performance. There were no significant predictors of end-ofsemester performance on the cloze test beyond the beginning-of-semester score. It should be noted that this analysis does not address the question that has been raised (Stansfield and Hansen 1983) as to whether there is a

40

Language Learning

VoI. 36, No. I

Table 3 Multiple regression analysis using beginning-of-semester TOEFL scores and affectivelcognitive scores as predicrors of the end-of-semester TOEFL scores

End-of-Semester TOEFL Scores Total TOEFL Score

Step

Variable Entered TFI ANX MOT F1 AT

R ,936 -.336 -.467 ,782 ,212

R2
,876 .888 ,890 ,911 .921

F 225.8 3.3 .7 6.7 3.7

I 2 3 4 5

***
*

** *
***

TOEFL Parts Scores Structure


1

2 3 4 5 Listening

STI ANX MOT AT FI

,855

* **
*** **

-.272 -.490 ,233 .737

,731 ,744 ,756 ,775 ,808 ,713 .746 ,747 ,794 ,855 ,624 ,655 .662 .666 ,752

103.4 1.9 1.7 3.0 5.7 94.3 4.8

I 2 3 4 5
1

LCI ANX MOT AT FI

** *** *** * **

,844 -.296 -.396 .274 ,726


,793 -.282 -.352 ,224 ,724

.I 8. I 14.3
64.3 2.8 .7 .4 11.8

Reading

RD1
ANX MOT AT FI

2 3 4 5
***p<.001 *'p<.01 'p < .05 FI AT ANX MOT TFLl TFL2 STI ST2 LCI LC2 RDI R D2

= Field Independence = Ambiguity Tolerance = English Class Anxiety = Motivational Intensity = Beginning-of-Semester TOEFL = End-of-Semester TOEFL = Beginning-of-Semester Structure part of the TOEFL = End-of-Semester Structure part of the TOEFL = Beginning-of-Semester Listening part of the TOEFL = End-of-Semester Listening part of the TOEFL = Beginning-of-Semester Reading part of the TOEFL = Endaf-Semester Reading part o f t h e TOEFL

Chapelle and Roberts


Table 4 Multiple regression analysis using beginning-of-semester language scores and affecriveJ cognitive measures as predictors of end-of-semester language measures

41

End-of-Semester Language Measure Structure Test

Step
1 2 3 4 5

Variable Entered
STI ANX MOT FI AT

R .757 -.293 -.465 ,748 .I98


,855 -.593 -.395 .672 ,072

R2 ,573 ,577 ,579 .703 ,710 .731 ,746 ,767 ,803


205

F 42.9 .3 .I 12.2 .7 86.8 1.9 2.7 5.3 .2 46.1 1 .o .8 2.4 3.1 72.4
5.8

*** ** ***

Cloze Test

I 2 3 4
5

CLI MOT ANX F1 AT DII ANX MOT AT FI CCI ANX MOT FI AT

Dictation Test

I 2 3 4 5
1 2 3 4 5

***

* *** ** * *

,768 -.334 -.563 ,184 ,619 333 -.388 -.533 ,653 ,021

390 ,603 .613 .643 ,678 ,693 .742 ,764 ,790 .796

Oral Communication Test

2.9 3.5 .8

~~

- * p < ,001 **p< .01 * p < .05

FI
AT ANX MOT STI ST2 CLI CL2 DII DIZ CCI cc2

= Field Independence = Ambiguity Tolerance = English Class Anxiety = Motivational Intensity = Beginning-of-Semester Structure Test = End-of-Semester Structure Test = Beginning-of-Semester Cloze Test = End-of-Semester Cloze Test = Beginning-of-Semester Dictation Test = End-of-Semester Dictation Test = Beginning-of-Semester Oral Test of Communicative Competence = End

42

Language k a r n i n g

Vol. 36, No. 1

bias toward FI students on the cloze test. This analysis indicates that none of the variables predicts improvement on the cloze test. To show bias, correlations between FI and the cloze test must be compared with correlations between FI and other language measures given at the same time (Tables 7 and 8). In fact, the correlations between F1 and the cloze tests were no stronger than those between FI and the other language measures. Thus, these results provide little evidence for the FI/cloze relationship question. The test of communicative competence had a different set of predictors from the other language proficiency measures. Along with beginning of semester performance, ANX and FI, a lack of Motivational Intensity was a significant predictor of communicative competence. This negative correlation on the communicative competence measure alone supports the hypothesis that those students who spend all of their energy on classoriented L2 learning may be those who d o not spend time using the language to acquire communicative competence. The results of these Multiple Regression Analyses provide some evidence for the importance of FI in L2 acquisition as measured by a variety of L2 tests, The hypothesis that F D would be related to performance on the oral test of communicative competence was not supported; instead, FI was found to be a significant predictor of this measure as well. It may be difficult to find such relationships because the GEFT used for measuring FI/ D could more accurately be described as a measure of a persons ability to use a FI style. Assuming the cognitive flexibility suggested by Brown (1980), each individual is capable of using both F1 and F D styles to some degree. The style he/ she uses depends on the task at hand as well as the degree to which he/ she is able to respond to it appropriately (be flexible). When doing the GEFT, subjects are asked to use their FI style to find the simple figures in the more complex ones. Some subjects have a greater ability to use their FI style to perform this task; they get high scores on the test. However, they are never called upon to use a F D style and we do not have a measure of the extent to which they are FD. It is thus inappropriate to label a lack of FI as FD; consequently, from this and previous research, we cannot make claims about the relationship between F D and L2 acquisition. The importance of AT in L2 acquisition is evidenced by the fact that it was a significant predictor of performance on the TOEFL. However, the expected relationship was not found between AT and the oral test of communicative competence, in which a great deal of ambiguity should

Chapelle and Roberts

43

have been perceived. However, on the tests related to structure (sub-part of TOEFL and multiple choice structure) and listening (sub-part of TOEFL and dictation) there was no relationship between performance a t the beginning of the semester and AT but a t the end of the semester the relationships were significant. Apparently, then, during the course of the semester, those students with higher levels of AT had an advantage in acquisition of English structure and listening comprehension.

CONCLUSIONS
Conclusions from the study must be drawn in view of the fact that there are some limitations inherent in this type of research. In particular, this study was faced with a limit on the number of international students that could be treated as a homogeneous group for the purposes of data analysis. In addition, there was a limit on the amount of time that could be requested of the students so that additional measures that may be covariants of FI and AT could not be administered. Consequently, these results cannot be considered ultimate answers to L2 acquisition questions; instead, they point toward evidence that may help answer some of these questions. For the question of good language learner characteristics, these results indicate that AT and FI should be considered components of L2 Aptitude. Thus, a good language learner may be one who is, among other things, Field Independent and Tolerant of Ambiguity; furthermore, these results provide evidence using non-English speakers in an L2 environment thereby confirming and expanding hypotheses that have previously been supported using native English speakers in a FL environment. Because this research is unique in its use of international students in an L2 environment, replication of these findings is needed. In particular, it is necessary to d o research using larger samples of students from different language groups. In addition t o defining good language learners in terms of cognitive/ affective characteristics, it is necessary to describe the learning strategies that facilitate their success. Empirical research is needed that will provide evidence for the assumption that, for example, the learner with a great deal of AT does indeed make logical inferences concerning language cues and that those inferences facilitate his success on particular L2 tasks. When some of the strategies ofgood language learners have been identified the subsequent question can be faced: Can good language learner strategies be taught to less successful learners? An affirmative answer to this question

44

Language Learning

Vo1. 36, No. 1

has been assumed by some (Omaggio 1981), but research is needed to provide evidence for this assumption. Regardless of whether good language learner strategies can be taught, there remains a question of how learners with particular cognitive/ affective characteristics should be taught. Because L2 acquisition is a multi-faceted task and learners with varying degrees of AT and Fl may prefer and excel in different situations, it may be appropriate to match teaching approaches to students. This matching has been tried in both L2 instruction (Wesche 198 1) and instruction in other subject areas (Pask 1976) with some success. Research needs to proceed in this direction by placing students in instructional environments based on their level of AT and Fl and by further investigating approaches that can be taken in individualized forms of instruction. In short, it is not appropriate to assume that all learners will benefit from the same kind of L2 instruction; it is one of the tasks of researchers to determine how instruction ought to vary from one learner to another.

REFERENCES
Bachman, L., and A. Palmer. 1984. The construct validation of some components of communicative proficiency. TESOL Quarterlv 16:449-465. Bialystok, E., and M. Frohlich. 1978. Variables of classroom achievement in second language learning. Modern Language Journal 32:327-336. Brown, H.D. 1980. Principles of Language burning and Teaching. Englewood Cliffs, New Jersey: Prentice-Hall, Inc. Budner, S. 1962. Intolerance of ambiguity as a personality variable. Journal o Personality f 30:29-50. Canale, M. and M. Swain. 1980. Theoretical bases of communicative approaches to second language teaching and testing. Applied Linguistics 1: 1-47. Chastain, K.D. 1969. Prediction of success in audio-lingual and cognitive classes. Language Learning 19:27-39. Carroll, J.B., and S.M. Sapon. 1959. Modern Language Aptitude Test: Form A Manual. New York: Psychological Corporation. Day, R. 1984. Student participation in the ESL classroom or some imperfections in practice. Language Leurning 34(3):69- 102. Ehrlich, D. 1965. Intolerance of ambiguity, Walks A scale: Historical comment. Psvchological Reports I7:591-595. Frenkel-Brunswik, E. 1949. Intolerance of ambiguity as an emotional and perceptual personality variable. Journal of Personality 18: 108-143. Gardner, R.C., and W.E. Lambert. 1972. Attitudes and Motivation in Second Language Learning. Rowley, Mass.: Newbury House Publishers. Gardner, R.C., and P.C. Smythe. 1979. Attitude/ Motivation Test. Language Research Group. London: University of Western Ontario.

Chapelle and Roberts

45

Hansen, J., and C. Stansfield. I98 I . The relationship between field dependent-independent cognitive styles and foreign language achievement. Language Learning 3 1:349-367. Hansen, L. 1984. Field dependence-independence and language testing: evidence from six pacific island cultures. TESOL Quarier1.v 18:3 11-324. Levitt, E. 1953. Studies in intolerance of ambiguity: I The Decision Location Test with grade school children. Child Development 24:263-268. Linn, R.L., and J.A. Slide. 1977. The determination of the significance of change between preand posttesting periods. Review o Educational Research 47: 121 - 150. f MacDonald, A.P. 1970. Revised scale for ambiguity tolerance: reliability and validity. Ps,vchological Reports 25:79 1-798. Naiman, N., M. Frohlich, and H.H. Stern. 1975. The G o o d h n g u a g e Learner. Toronto: The Ontario Institute for Studies in Education. Nie, N.H., C.H. Hull, J.G. Jenkins, K. Steinbrenner, and D.H. Bent. 1975. Siatisrical Package f o r the Social Sciences. United States: McGraw-Hill, Inc. Norton, R. W. 1975. Measurement of ambiguity tolerance. Journalof Personality Assessment 39:607 -6 18. Omaggio. A. 198 I , Helping Learners Succeed: Acriviiies f o r the Foreign Language Classroom. Washington D.C.: Center for Applied Linguistics. Pask. G. 1976. Styles and strategies of learning. British Journal o Educational Psychology f 46: 128-148. Pimsleur, P., L. Mosberg, and A. Morrison. 1962. Student factors in foreign language learning. Modern Language Journal 46: 160-170. Pimsleur, P., D. Sundland, and R. Mclntyre. 1966. Underachievement in Foreign Language Learning. New York: MLA Materials Center. Ramirez, M. 1973. Cognitive styles and cultural democracy in education. Social Science Quarterly 532395-904. Ramirez, M., A. Casteneda, and P.L. Herold. 1974. The relationship of acculturation to cognitive style among Mexican Americans. Journal o Cross-Culrural Psychology f 5:424-433. Schumann, J. 1976. Social distance as a factor in second language acquisition, Language Learning 26:?. Seliger, H. 1977. Does practice make perfect?: A study of interaction patterns and L2 competence. Language Leurning 27:263-278. Stansfield, C., and J. Hansen. 1983. Field dependence-independence as a variable in second language cloze test performance. TESOL Quarterly 17:29-38. Tucker, R.G., E. Hamavan, and F.H. Genesee. 1976. Affective, cognitive and social factors in second-language acquisition. Canadian Modern Language Review 32:2 14-226. Wesche, M.B. 1981. Language aptitude measures in streaming, matching students with methods, and diagnosis of learning problems. In Individual Differences and Universals in Language Learning Aptiiude, ed. K.C. Diller. Rowley, Mass.: Newbury House Publishers, Inc. Witkin, H.A., P.K. Oltman, E. Raskin. and S.A. Karp. 1971. A manual f o r rhe Embedded Figures Test. Palo Alto: Consulting Psychologists Press. Witkin, H.A., C.A. Moore, D.R. Goodenough, and P.W. Cox. 1977. Fielddependent f and field-independent styles and their educational implication. Review o Educational Research 47: 1-64,

You might also like