Assessing the Metacognitive Growth of ESL Student Writers

Vol. 3. No. 1 — November 1997

Assessing the Metacognitive Growth of ESL Student Writers

Loretta F. Kasper
Kingsborough Community College/CUNY
<drlfk@aol.com>

Abstract

Recent research suggests a potential link between metacognitive knowledge and writing performance. This research has particularly important implications for ESL instruction because developing English language writing competence presents a major challenge to students. The present study attempted to clarify the relationship between metacognition and ESL writing performance by posing the following questions: (1) Does each of the components of the metacognitive model– personal, task, and strategy–have an equivalent effect upon the writing performance of ESL students?; (2) Do these three components impact performance equally at different levels of English language proficiency?; and (3) How do metacognitive models evolve as students advance in their English language development? This study offers both quantitative (statistical) and qualitative (protocol) data to support the hypothesis that there is a significant positive correlation between ESL students’ metacognitive growth, along and across the three components of the metacognitive knowledge base, and their actual performance on a final writing assessment. The results suggest that it is especially critical to design activities which target and develop students’ knowledge of efficient writing strategies. The results further suggest that instruction designed to strengthen students’ metacognitive models should be introduced early on as an integral part of ESL writing instruction.

Metacognitive theory deals with cognitive self-knowledge, that is, what individuals know about their own thinking (Kellogg, 1994). The emergence of metacognitive theory in the 1970’s added to our understanding of complex cognitive phenomena by providing data on learners’ awareness of and control over their cognitive activities (Devine, 1993). In 1979, John Flavell defined metacognition as knowledge which focuses on or regulates any part of cognitive activity, and he identified two general dimensions of metacognition: knowledge and experience. [-1-]

In his 1985 work, Cognitive Development, Flavell proposed that our metacognitive knowledge base consists of what we have learned, through experience, about cognitive activities, and he said that it could be subdivided into three distinct and highly interactive knowledge variables: personal variables, task variables, and strategy variables. Possessing a strong metacognitive knowledge base is critical to successful learning, and in metacognitive terms, a good learner is “one who has ample metacognitive knowledge about the self as learner, about the nature of the cognitive task at hand, and about appropriate strategies for achieving cognitive goals” (Devine, 1993, p. 109).

In 1993, Devine, Railey, and Boshoff undertook a study that attempted to provide information on the role of metacognition in second language writing by investigating cognitive models in both second language and first language basic writers and assessing the effects of these models on writing performance.While extensive research had previously been done on the role of metacognition in first and second language reading performance (see respectively, Baker and Brown (1984) and Devine (1993) for a review of first and second language research), corresponding research in writing had been sorely lacking (Devine, 1993). Thus the Devine et al. (1993) study represented an important contribution to the field.

The results of the Devine et al. (1993) study suggested a potential link between students’ metacognitive models and their actual writing performance. This potential link between metacognition and writing performance has particularly important implications for ESL writing instruction. Developing English language writing competence presents a major challenge to ESL students, and meeting this challenge is critical to the academic progress of these students. This is because falling short of prescribed levels for writing proficiency may exclude ESL students from acceptance into the mainstream curriculum and thus preclude them from earning a college degree (Kasper,1997).

In her essay, “The Role of Metacognition in Second Language Reading and Writing,” Joanne Devine (1993) points out that there are now data that suggest that “metacognitive variables play an even more important role than linguistic competence in successful second language writing” (p. 116). Devine goes on to say that second language writers often have the problem of a “limited metacognitive knowledge base,” so that they are unable to determine “if they are making progress towards the goal of the writing task” (p. 117). Additionally, students may not even have clearly defined goals for the English language writing tasks in which they are engaged. Further, as Zamel (1985) has found, ESL students often expect the instructor to assume responsibility for clarifying writing goals and monitoring progress. Yet, to progress as writers, students must learn to actively set, and then regulate and monitor their own progress towards those cognitive goals associated with writing; this process is referred to as “executive control” (Garner, 1994, p. 715). [-2-]

These findings suggest that if it is to be effective, ESL writing instruction must be based on an understanding of students’ metacognitive knowledge and must be directed toward helping students to develop their cognitive models. Exposing ESL students to instructional activities which require self-examination and self-monitoring gives them control over their writing as they learn to engage in goal-directed behaviors resulting in increased English language writing competence.

Thus, there appears to be a potential link between writing performance and the metacognitive knowledge base, which is, in turn, impacted by and composed of three theoretically interactive, but separate variables–personal, task, and strategy. This potential link leads to several research questions: (1) Does each of the components of the metacognitive model–personal, task, and strategy– have an equivalent effect upon the writing performance of ESL students?; (2) Do these three components impact performance equally at different levels of English language proficiency?; and (3) How do metacognitive models evolve as students advance in their English language study?

The present study sought to answer these questions by using pedagogical tools aimed at assessing the separate and combined effects of each individual component of the overall metacognitive model and comparing these effects for ESL students at the intermediate versus the advanced level of English language proficiency. Students’ metacognitive models were assessed through their pre- and post-course responses to (a) writing autobiographies and (b) cognitive style questionnaires (see Appendices A and B, respectively). These two tools have been used by previous researchers (i.e., Sandman, 1993; Sandman & Weiser, 1993; and Devine, Railey, & Boshoff, 1993, respectively) to help promote students’ awareness of the various factors involved in the writing experience.

The Writing Autobiography

Sandman and his colleagues have used pre- and post-course writing autobiographies in their basic writing classes to elucidate students’ attitudes toward writing. The writing autobiography asks students to describe and evaluate both positive and negative aspects of their English language writing experience, thereby increasing awareness. By increasing students’ awareness of their own writing experience and knowledge, the writing autobiography reflects self-knowledge and can therefore be used to help students define the personal component of the metacognitive model of writing. The pre-course writing autobiography provides instructors with a meta-cognitive “baseline” for each student. The post-course writing [-3-] autobiography helps students to monitor their own development as writers and assists students in developing sound criteria for assessing their own writing performance (Sandman, 1993).

The Cognitive Style Questionnaire

Students’ metacognitive models may be further assessed through cognitive style questionnaires (Devine et al., 1993.). The cognitive style questionnaire directs students’ attention to the goals they set and the strategies they use when writing. This instrument clarifies task and strategy knowledge by asking students to define good writing, to describe what they do when they have trouble writing, and finally to rank, in order of importance both to their writing teachers and to themselves, criteria such as clarity, fluency, grammar, originality, content, organization, and exploration.

Written Protocols

Both the writing autobiography and the cognitive style questionnaire ask students to describe their own behavior and experiences and to articulate their understanding of task and strategies. One might wonder about the reliability and usefulness of written protocols generated by students with limited English proficiency. However, in his work, Quantifying Language, Phil Scholfield (1995) reports that such protocols are being used increasingly with foreign students, who “show a greater ability than might have been expected to introspect usefully about their conscious learning strategies and communication processing activities for the language they are learning, as well as what they say in it” (p. 65).

Thus, the present study used students’ responses to writing autobiographies and cognitive style questionnaires to evaluate the three components of their metacognitive models, and then assessed the evolution of these metacognitive models by comparing pre- and post-course responses to each instrument. The study also measured the relationship between each of these metacognitive variables and students’ performance on college writing assessment examinations. The study provided between-subjects analyses of differential trends in metacognitive growth as a function of level of linguistic proficiency, i.e., intermediate versus advanced. The study further provided within-subjects analyses of the continuing evolution of individual students’ metacognitive models as these students progressed from the intermediate to the advanced level of English language study.

Method

Subjects

There were 120 students in the total subject sample. The 120 students in the subject sample represented diverse ethnic, cultural, and linguistic backgrounds, including Russian, Haitian, Hispanic, [-4-]and Asian. All of the ESL students in the subject sample were working toward, at minimum, a two-year college degree, but most would go on to pursue more advanced college degrees.

Sixty-seven students, or 56% of the total subject sample, were intermediate level (ESL 09) students. Fifty-three students, or 44% of the total subject sample, were advanced level (ESL 91) students. Students designated as intermediate had an entry level TOEFL score of approximately 350; students designated as advanced had an entry level TOEFL score of approximately 425. The total subject sample of 120 students provided the basis for the between-subjects analyses.

Fourteen students, or 12% of the total sample, continued as subjects in the study for two consecutive semesters, first as members of the ESL 09 group, and then as members of the ESL 91 group. These fourteen students provided the basis for the within-subjects analyses.

Course Parameters

Both ESL 09 and ESL 91 were integrated reading/writing courses. Both courses met for six hours per week over a twelve-week semester for a total of 72 class hours. Both the ESL 09 and the ESL 91 courses were run as discipline-oriented courses, in which the content of various academic disciplines was used as the medium of English language instruction (for a complete discussion of discipline-oriented ESL courses, the reader is referred to Kasper, 1994).

In both courses, all writing assignments evolved out of the readings and required students to explore and expand their knowledge of topics presented in the texts. A number of different writing activities was used, including summaries, open-ended comprehension questions, and analytical essays in a variety of rhetorical modes.

To pass each course, students had to receive a passing grade on an end-of-semester writing assessment. For both courses, this assessment consisted of an in-class final writing examination in which students were given two hours to plan, write, and revise a persuasive essay on their choice of three assigned topics derived from a reading text. Students were given this reading text one week in advance of the writing final; however, they were not told the topic of the essay in advance of the final.

At both levels, students received either a P (Pass) or an R (Repeat) on the final writing assessment. Those who received an R had to repeat the course. At both levels, the criteria for receiving a P included appropriate organization of content, fluency and clarity of expression, and grammatical correctness. To increase objectivity in grading, the final writing assessments were not graded by the course instructor, but were cross-graded by two other ESL instructors teaching that same level course. [-5-]

Students who passed the final writing assessment were designated as the “Successful” group in this study. Students who failed the final writing assessment were designated as the “Unsuccessful” group in this study.

Data Collection

Data were collected over a period of six consecutive semesters. These data consisted of individual student responses to pre- and post-course writing autobiographies (adapted from Sandman, 1993, p. 19) and to cognitive style questionnaires (adapted from Devine et al., 1993, p. 224). Copies of the writing autobiography and the cognitive style questionnaire are provided in Appendices A and B.

The pre-course measures were completed during the first week of classes; the post-course measures were completed during the final week of classes. The time lag between completion of pre- and post-course measures was approximately twelve weeks. Moreover, students did not have access to their pre-course responses during this time. Pre-course questionnaires were collected immediately after completion. This was done to insure that students’ responses to post-course measures would be minimally affected by their responses to pre-course measures.

Data Analyses

Pre- and post-course responses to writing autobiographies and to cognitive style questionnaires were compared for each of the 120 students in the subject sample. If metacognitive variables significantly impact writing performance, we would expect the degree of change between pre- and post-course responses to be greater for students who pass (i.e., “Successful”) than for those who fail (i.e., “Unsuccessful”) the final writing assessment. To clarify the relationship between metacognition and writing performance, changes in individual student responses to each of these measures were rated according to the criteria described in Appendix C. The resulting ratings were then correlated with the student’s actual writing performance.

To control for potential experimenter bias, two independent raters, neither of whom was told the purposes of the study beyond that it was designed to measure changes in students’ attitudes toward writing, were asked to rate students’ responses. Raters were provided with a list of criteria (see Appendix C), and each student in the subject sample was assigned a rating ranging from zero to five to correspond to each of the three components of the metacognitive model: personal, task, and strategy. After all ratings had been assigned, the interrater concordance was assessed by a Spearman test and found to be .977; thus, the independent raters were in agreement almost 98% of the time. [-6-]

After ratings were generated for each individual student, both successful and unsuccessful student writers were further separated into two groups according to level of linguistic proficiency. Data were then analyzed separately for all ESL 09 and all ESL 91 students. This analysis was designed to assess any differential trends in the metacognitive growth of writers at the intermediate or advanced levels of English language proficiency.

Data for the fourteen students who comprised the within-subjects analyses were then separated and further analyzed. This analysis evaluated the extent to which individual ESL writers continued their metacognitive growth as their level of linguistic proficiency increased.

Results

If there is a significant relationship between metacognition and writing performance, then across both levels of linguistic proficiency, we would expect the ratings for passing students to be higher than the ratings for failing students. Moreover, ratings for each of the components of the metacognitive model, i.e., personal, task, and strategy, should correlate positively and significantly with scores on the writing assessment examinations. The data analyses not only confirm these research expectations, but they also provide additional insights into the specific effects of individual metacognitive components at different levels of linguistic proficiency.

Between-Subjects Analyses

The between-subjects analyses are based on the total subject sample of 120 students. Table 1 shows a comparison of the average metacognitive ratings for the ESL 09 versus the ESL 91 students.

Table 1

___________________________________________________________
Mean Metacognitive Ratings: Between-subjects analyses

                ESL 09               ESL 91
         Pass          Fail     Pass          Fail

Personal  3.4          0.9       3.6          1.8
Task      3.8          2.0       4.0          2.6
Strategy  3.2          1.2       4.1          3.8
___________________________________________________________
Note: ESL 09 n=67; ESL 91 n=53.

An analysis of variance (ANOVA) of the effect of writing score, i.e., pass or fail, on metacognitive growth yielded significant differences for each of the components of the metacognitive model [-7-] for students at both levels of English language proficiency (all Fs (1,116) > 26.00; all ps < .0001). Thus, the average metacognitive ratings are consistently higher for passing students than for failing students across both levels of English language proficiency.

In comparing students in ESL 09 and ESL 91 in terms of the average percent who passed the writing assessment, there is no significant difference between the two levels by a chi-squared test (X2(1) = 0.28). The pass rate for ESL 09 was 84%; the pass rate for ESL 91 was 91%.

In pairwise comparisons of the effect of level of English language proficiency on metacognitive growth, the average ratings are all higher for students at the ESL 91 level. However, an ANOVA procedure showed that the effect of level was significant only for the strategy variable (Fstrategy(1, 116)=17.04; p=.001). Thus, in the between-subjects analyses, there is no statistically significant difference between intermediate and advanced level students in terms of the personal and task components of their metacognitive models (both Fs(1,116) <1.84; both ps>.05).

Within-Subjects Analysis

The previously described analyses compared different students enrolled at two different levels of ESL, and thus provided between-subjects analyses of the effects of interest in the study. However, fourteen students in the sample were subjects at the intermediate level, ESL 09, and subsequently at the advanced level, ESL 91. Separate analyses of these students’ data provide within-subjects information, thereby supplying a more comprehensive measure of the extent to which individual ESL writers continue to grow as their level of linguistic proficiency increases.

Table 2 shows students’ ratings for the personal, task, and strategy components. The data reveal that all metacognitive ratings increased as these fourteen students progressed from the ESL 09 to the ESL 91 level. This increase indicates that these students continued their metacognitive growth during the following semester.

Table 2

___________________________________________________________
Mean Metacognitive Ratings: Within-subjects analyses

                     ESL 09               ESL 91
Personal              3.9                  4.1
Task                  3.6                  3.7
Strategy              3.1                  3.9
___________________________________________________________
Note: Of the 14 students in this group, 14 passed ESL 09; 13 passed
ESL 91. [-8-]

Although all ratings increased as students progressed from the intermediate to the advanced level, only the rating for the strategy component is significantly different across the two levels by a paired t-test (t(14)=3.67; p < .003. This replicates the results found in the between-subjects analyses described previously.

Correlation Analyses

Overall Subject Sample

Because one of the hypotheses driving this research was that there was a significant positive relationship between metacognitive growth and writing performance, Spearman correlation coefficients were computed to assess the degree of relationship between a number of different variables in the study. All correlations are based on 120 observations. All three components of the metacognitive model correlate positively and significantly with performance on the final writing examination. For personal, task, and strategy ratings, the correlations range from r=+0.43 to +0.46; ps <.0005.

Within

The data for the fourteen students who were in both the ESL 09 and the ESL 91 groups provided some insight into how the relationships between metacognitive variables are affected by individual students’ increasing English language proficiency. All Spearman correlations were based on 14 observations. Analyses revealed that at both levels of English language proficiency the relationship of all three components–personal, task, and strategy–to performance on the writing final was marginally significant (rs range between +0.45 and +0.48; ps < .10).

However, strategy knowledge at the ESL 09 level correlated significantly with all variables at the ESL 91 level (rs range between +0.58 and +0.78 for ESL 91 personal, task, and strategy variables; ps <.02). Thus, the benefit of acquiring strategic knowledge at the intermediate level was related to enhanced subsequent growth in personal, task, and strategy knowledge at the advanced level.

Analysis of Students’ Written Protocols

The results of the statistical analyses can be further clarified first by examining students’ own written protocols, i.e., their responses to specific questions designed to tap into the personal, task, and strategy variables of the metacognitive model, and then by comparing the responses of successful and unsuccessful student writers, i.e. those who passed/failed the final writing assessment. [-9-]

Personal Variables

The pre- and post-course writing autobiographies (see Appendix A) presented students with questions designed to evaluate the personal component of the metacognitive model for writing. The statistical analyses indicated a significant relationship between personal knowledge and writing performance. An examination of students’ responses to the writing autobiographies confirms the results of the statistical analyses and reveals substantive differences between successful and unsuccessful student writers, particularly at the intermediate level.

Overall, in terms of the personal component of the metacognitive model, successful students at both levels seem to have a greater awareness of themselves as writers. An important indication of this growth of self-awareness was their emphasis on the progress they had made, and at the same time, their acknowledgement of the progress they had yet to make. A representative student response is, “This semester I have learned a lot about my ability in English because I didn’t know that I could write an essay of four pages. Although I did them with some mistakes, I had the boldness to do them.” In contrast, unsuccessful students showed little awareness of themselves as writers, overemphasized their weaknesses, and did not note any progress they had made. This is evidenced in the following student response, “I think I have a lot of mistakes in grammar. I have a lot of ideas in my head, but it is difficult for me to explain all my ideas. I have a big problem with spelling too. I will never be a good writer, no matter how hard I try.”

Task and Strategy Variables

While the writing autobiography assessed primarily the personal component of the students’ metacognitive models, the cognitive style questionnaire was designed to evaluate the task and strategy components. One of the questions on the cognitive style questionnaire asked students to “Define good writing.” The purpose of this question was to assess students’ understanding of the writing task and its requirements. The responses of successful students at the ESL 09 level, as well as most students at the ESL 91 level, emphasized the importance of developing fluency and clarity and indicated an understanding that good writing requires a clearly communicated message and makes the needs of the reader a priority. An example of such a student response is, “Good writing makes the reader understand and get excited when he/she reads it.”

In contrast, the responses of unsuccessful student writers, particularly those at the ESL 09 level, did not really take the needs of the reader into account and did not make clear that the purpose of writing is to communicate a message to another person. These unsuccessful students defined the goal of writing as producing a perfect paper believing that a reader would automatically [-10-] understand them if they used correct grammar. This viewpoint is indicated in the following student response, “Good writing is when you don’t have a problem with grammar, when it’s not too long, and you don’t have any errors. Then the reader will understand you.”

Although most students at the ESL 91 level identified fluent and clear communication as the priority in good writing, an interesting phenomenon occurs in their post-course responses, as even successful students began to rate grammatical correctness as increasingly important. This is not surprising in view of the fact that they will be faced with the CUNY WAT (City University of New York Writing Assessment Test) during the following semester. The CUNY WAT is an holistically evaluated test which requires students to write a coherent and persuasive essay on one of two possible topics. It is a test that places a great deal of importance on grammatical correctness. In her research Adele MacGowan-Gilhooly (1991) also found that when ESL students are faced with a writing assessment that places great importance on grammatical correctness, they readjust their priorities concerning the requirements of the writing task. Unfortunately, MacGowan-Gilhooly reports that this shift in priorities toward linguistic accuracy and away from communicative competence often causes ESL students to regress from former performance levels. These findings speak to the importance of designing institutional writing assessments which evaluate students’ competence in written communication rather than in linguistic accuracy.

The results of the statistical tests suggest that while all three metacognitive variables impact performance during a single level of instruction, only the strategy component appears to be related to performance in subsequent levels of instruction. Thus responses to the question, “What do you do when you have trouble writing?” designed to elucidate students’ strategy knowledge, was of particular interest to the purposes of this study.

Successful student writers at both levels, as well as unsuccessful advanced level students, answered this question by stating that ideas need time to incubate and that taking a break from writing often helps to facilitate performance. Moreover, student writers at both levels point to the importance of taking the time needed to find their own solutions to writing problems. This attitude is expressed in the following response, “When I have trouble writing, I leave it for awhile, and come back later. I never give up. I look over and analyze my problems, and try again.” Thus, the responses of successful ESL 09 students and all ESL 91 students demonstrate growth in strategy knowledge.

In contrast, the responses of students who are unsuccessful at the intermediate level do not demonstrate such growth. These [-11-] unsuccessful ESL 09 students express a belief that there is nothing they can do to improve their own writing. Moreover, they do not express persistent attempts to find solutions to their writing difficulties, stating that when they have trouble writing, they “give up and go to look for friends and relatives who know more English to get help.”

Discussion

The study reported in this paper represents a preliminary attempt to assess the relationship between metacognitive personal, task, and strategy knowledge and the writing performance of ESL students. The study offers both quantitative (statistical) and qualitative (protocol) data to support the hypothesis that there is a significant positive correlation between ESL students’ metacognitive growth, along and across the three components of the metacognitive knowledge base, and their actual performance on a final writing assessment. Thus students who were successful on the final writing assessment obtained higher ratings on all three metacognitive variables. It must be noted that while these results indicate a positive correlation between metacognitive knowledge and writing performance, in no way do they imply a causative relationship between these variables. Further controlled studies are needed to determine any potential causative relationship.

It is not surprising, however, that metacognitive ratings were higher for students who passed the final writing assessment than for those who failed it. Research on tasks such as reading and mathematical problem solving has also demonstrated a positive relationship between metacognition and performance (Collins, 1994; Schoenfeld, 1987). In both reading and mathematics, performance improves as students develop the ability to describe what they know and what they don’t know, as well as to plan and regulate how they go about learning. In short, tasks in both reading and mathematics become easier as students develop more adequate and more accurate metacognitive models.

In terms of the three individual components of the metacognitive model for writing, the present study showed, at least for the 120 students in this subject sample, that personal and task knowledge did not significantly change as students progressed from the intermediate to the advanced level. In contrast, however, their knowledge of effective writing strategies did increase significantly as students became more proficient in the English language. The within-subjects analyses provided insight into the continuing evolution of the metacognitive model for individual students as they gained increased proficiency in English. The correlational analyses revealed that at both levels, students who were successful on the final writing assessment obtained somewhat higher ratings on all three metacognitive variables. In particular, strategy knowledge at the intermediate level was related to all three types of metacognitive knowledge at the advanced level. [-12-]

That strategy knowledge increased as students advanced in linguistic proficiency is not surprising. This finding replicates some earlier studies. In 1980, Brown found that “metacognitive deficiencies are a function of inexperience” (p. 475). Therefore, as students gain experience with a given task, we would expect them to learn to use strategies that lead to more efficient task completion. Yet, as Scruggs, Mastropieri, Monson, and Jorgenson (1985) have found, the independent use of such metacognitive strategies develops very gradually. This suggests that it would be useful to incorporate instructional activities that teach efficient strategy use.

In fact, studies show that increases in learning do follow direct instruction in metacognitive strategies (Blakey & Spence, 1990) and that students can be taught to develop self-awareness and control of learning (Collins, 1994). Therefore, if metacognitive growth is a goal of writing instruction, we must provide guidance in recognizing and practice in applying strategies during the early levels of English language instruction. We need to focus students’ attention on how tasks are accomplished most efficiently. Process goals, in addition to content goals, must be established and evaluated with students (Blakey & Spence, 1990). Students must be taught to analyze their own learning and progress. They must be trained to develop, monitor, and evaluate a plan of action that will help them successfully complete writing tasks. Introducing these activities early on will help students, from the beginning of their English language learning experience, to develop more accurate views of themselves as writers and of the writing task and its requirements, and to acquire strategies for how to deal with problems they may have when writing.

The analyses of students’ written responses to pre- and post-course writing autobiographies and cognitive style questionnaires confirmed the statistical results and provided valuable qualitative data that offer important insights into students’ own perceptions of the experience of writing. Several factors stand out as important in the protocol analysis. A focus on fluency and clarity of expression and away from linguistic accuracy was related to success on the final writing assessment. Successful student writers were aware that the goal of writing is to communicate a message, and they tended to choose effective strategies to accomplish that goal. These strategies included planning, monitoring, and evaluating during the task of composing. As Dirkes (1985) has also found, using these strategies enables students to take charge of their learning behaviors and leads to successful task completion. In the present study students who prioritized the communicative aspects of writing performed better on the final assessment than did those who prioritized the grammatical aspects. This finding supports results reported by Kasper (1998) in a forthcoming paper. Research shows that student writers acquire their priorities from the response styles of their instructors (Zak, 1990), so that efficient learning behaviors can and should be modeled and taught. For [-13-] instruction to be most effective, ESL writing faculty should adopt a nonjudgmental response style that encourages students to explore their ideas freely and one that redirects responsibility for their writing development from the instructor to the students themselves. When students take charge of the writing experience, they assume what Garner calls “executive control” (1994, p. 715), learning first to set and clearly define writing goals and then to regulate and monitor their own progress toward these goals.

Finally, in its attempt to assess students’ metacognitive growth, the present study replicated the earlier work of Sandman (1993) and of Devine et al. (1993) by confirming the value of both the writing autobiography and the cognitive style questionnaire as rich sources of metacognitive information for both the student writer and the instructor. Both of these learning tools help to bring the writer’s complex, often subconscious, metacognitive knowledge to the level of consciousness, thereby promoting awareness as the writer engages in the cognitive task of composing. As Sandman (1993) explains, when students describe the strengths and weaknesses of their writing in a writing autobiography, they become aware of the factors that have shaped their writing behaviors and led to their attitudes toward the writing task. Likewise as Devine et al. (1993) maintain, when students complete the cognitive style questionnaire they gain insight into both the adequacy and the accuracy of their metacognitive models.

Further, responses to the writing autobiography and the cognitive style questionnaire provide the instructor with data concerning each student’s unique metacognitive model. This information enables the instructor to develop individualized strategies for supporting each student in the effort to develop English language writing competence.

In conclusion, many ESL students fail at writing because their metacognitive models have been inadequately developed, so that these students are literally “unaware” of the components of the overall experience of writing. By helping students to become more aware of their attitudes toward writing, why they have these attitudes, and what they do when they are in the process of writing, we can encourage them to develop more accurate metacognitive models for writing. This in turn can help ESL students to attain the English language writing competence so necessary to their academic progress.

References

Baker, L., & Brown, A.L. (1984). Metacognitive skills and reading. In P. D. Pearson (Ed.). Handbook of reading research (pp. 354-394). New York: Longman. [-14-]

Blakey, E., & Spence, S. (1990). Developing metacognition. ERIC Digest (ED327218). Available: http://ericps.ed.uiuc.edu/npin/respar/texts/home/metacog.html

Brown, A.L. (1980). Metacognitive development and reading. In R.J. Spiro, B.C. Bruce, & W.F. Brewer (Eds.). Theoretical issues in reading comprehension (pp. 453-481). Hillsdale, NJ: Erlbaum.

Collins, N.D. (1994). Metacognition and reading to learn. ERIC Digest (ED376427). Available: http://www.ed.gov/databases/ERIC_Digests/ed376427.html

Devine, J. (1993). The role of metacognition in second language reading and writing. In J. G. Carson & I. Leki (Eds.). Reading in the composition classroom: Second language perspectives (pp. 105-127). Boston: Heinle and Heinle.

Devine, J., Railey, K., & Boshoff, P. (1993). The implications of cognitive models in L1 and L2 writing. Journal of Second Language Writing 2: 203-225.

Dirkes, M.A. (1985). Self-directed thinking in the curriculum. Roeper Review, 11,: 92-94.

Flavell, J. H. (1979). Metacognition and cognitive monitoring: A new area of cognitive-developmental inquiry. American Psychologist 34: 906-911.

Flavell, J. H. (1985). Cognitive development (2nd ed.). Englewood Cliffs, NJ: Prentice-Hall.

Garner, R. (1994). Metacognition and executive control. In R. B. Ruddell, M. R. Ruddell, & H. Singer (Eds.). Models and processes of reading (fourth edition) (pp. 715-732). Newark, DE: International Reading Association.

Kasper, L. F. (1994). Discipline-oriented ESL reading instruction. Teaching English in the Two-Year College 22: 45-53.

Kasper, L. F. (1997). Meeting ESL students’ academic needs through discipline-based instructional programs. In T. Smoke (Ed.). Adult ESL: Politics, pedagogy, and participation in classroom and community programs (pp. 147-157). Hillsdale, N.J.: Erlbaum.

Kasper, L. F. (1998). ESL writing and the principle of nonjudgmental awareness: Rationale and implementation. Teaching English in the Two-Year College, 25, 60-68.

Kellogg, R. T. (1994). The psychology of writing. New York: Oxford. [-15-]

MacGowan-Gilhooly, A. (1991). Fluency before correctness: A whole-language experiment in college ESL. College ESL 1: 37-47.

Sandman, J. (1993). Self-evaluation exit essays in freshman composition: “Now I have new weaknesses.” Teaching English in the Two-Year College 20: 275-278.

Sandman, J., & Weiser, M. (1993). The writing autobiography: How to begin a two-year college writing course. Teaching English in the Two-Year College, 20: 18-22.

Schoenfeld, A.H. (1987). What’s all the fuss about metacognition? In A.H. Schoenfeld (Ed.). Cognitive science and mathematics education (pp. 189-215. Hillsdale, NJ: Erlbaum.

Scholfield, P. (1995). Quantifying language. Clevedon: Multilingual Matters.

Scruggs, T., Mastropieri, M.A., Monson, J., & Jorgenson, C. (1985). Maximizing what gifted students can learn: Recent findings of learning strategy research. Gifted Child Quarterly, 29: 181-185.

Zak, F. (1990). Exclusively positive responses to student writing. Journal of Basic Writing, 9: 40-53.

Zamel, V. (1985) Responding to student writing. TESOL Quarterly, 19: 79-101. [-16-]

Appendix A: Writing Autobiography

(adapted from Sandman, 1993, p. 19)

Return to main text

Note: The pre-course writing autobiography consisted of responses to questions 1 through 3. The post-course writing autobiography consisted of responses to questions 1 through 4.

1. Think of a particular time in your life when writing in English was a positive (good) experience. Describe that time in as much detail as you can. What was the task, the assignment, the circumstance? What are some of the factors that make writing a positive experience for you?

2. Describe one negative (bad) writing experience you have had. Think of a time when writing was difficult, frustrating, unsatisfying. Again, please be as specific as you can. What are some of the conditions that make writing difficult for you?

3. What are your strengths and weaknesses as a writer? What experiences have led you to believe that you have these strengths and weaknesses?

4. What have you learned this semester about your ability as a writer? How, specifically, do you think your writing has improved? What areas of your writing do you think still need work? [-17-]

Appendix B: Cognitive Style Questionnaire

Return to main text

(adapted from Devine et al., 1993, p. 224)

     1. Define good writing.
     2. What do you do when you have trouble writing?
     3. Rank the following in order of importance--in the first
        column rank their importance to you, personally; in the
        other column, rank their importance to the teachers who
        grade your papers.  Use #1 for most important through #7
        for least important.

     Importance to:            You            Teachers
     1. Clarity
     2. Originality
     3. Grammar
     4. Organization
     5. Exploration
     6. Fluency
     7. Content 

[-18-]

Appendix C: Description of Rating Criteria

Return to main text

Instructions to raters:

Read the pre- and post-course writing autobiography and the pre- and post-course cognitive style questionnaire for each student in the subject sample.

Based on the student’s responses to the writing autobiography, you should assign one rating to represent the personal variable.

Based on the student’s responses to question 1 of the cognitive style questionnaire, you should assign one rating to represent the task variable. Based on the student’s responses to question 2 of the cognitive style questionnaire, you should assign one rating to represent the strategy variable.

These ratings should be determined by which of three areas–grammar and correctness, personal voice, or communication–is indicated by the student as the most important concern, or the dominant focus, in effective writing. Consider the response to question 3 of the cognitive style questionnaire, together with the other responses in determining your ratings.

Grammar and correctness includes mention of punctuation, grammatical usage, and/or spelling as major concerns in writing.

Personal voice includes mention of creativity and self-expression as major concerns in writing.

Communication includes mention of sensitivity to audience and communicating thoughts clearly as major concerns in writing.

If a student identifies only grammar and correctness as the dominant focus, he/she should receive a rating of 1. If a student identifies personal voice as the dominant focus, he/she should receive a rating of 2. If a student identifies communication as the dominant concern in writing, he/she should receive a rating of 3. If a student identifies two of these areas, he/she should receive a rating of 4. A student whose response indicates that effective writing should communicate a message clearly, express the writer’s feelings, and be grammatically correct, thus identifying all three areas as important, should receive the maximum rating of 5.

The pre- and post-course measures should each be rated separately; yielding for each student 2 sets of personal ratings, 2 sets of task ratings, and 2 sets of strategy ratings to correspond to the pre- and post-course measures. [-19-]

To determine the final ratings for the personal, task, and strategy components, take the difference between the pre- and the post-course ratings for each.

To determine the overall metacognitive rating, average the ratings for the personal, task, and strategy components.

Therefore, a pre-course personal rating of 2 and a post-course personal rating of 4 yields a final personal rating of 3. A student with a personal rating of 3, a task rating of 5, and a strategy rating of 4 would have an overall metacognitive rating of 4.

Author Notes

This research was supported by a Kingsborough Community College Released Time Grant for Scholarly and Applied Research.

I thank Kingsborough Vice President and Provost Michael Zibrin and the 1995-96 Committee on Scholarly and Applied Research for giving me the opportunity to conduct this study.

I thank Morton Fuhr and Howard Kasper for sharing their ideas and insights and for their invaluable assistance in this research project.

I thank Robert Singer for his thoughtful comments on an earlier draft of this paper.

&copy Copyright rests with authors. Please cite TESL-EJ appropriately.

Editor’s Note: Dashed numbers in square brackets indicate the end of each page in the paginated ASCII version of this article, which is the definitive edition. Please use these page numbers when citing this work.

[-20-]