Steve Cornwell

At Osaka Jogakuin Junior College (OJJC) writing is an important part of the curriculum. First year students take two semesters of composition classes (twice a week; 70 minutes per class) where they work on learning how to write paragraphs and essays using six rhetorical patterns. Then during their second year they take four content-based courses and must write a paper for each course. In these writing classes students work together on tasks such as pre-writing, group essay writing, and other group assignments. The best example of group work can be seen in the end of the unit peer editing assignment where a student's essay is read and commented on by three other students.

 This emphasis on writing at OJJC has led to an interest in the role affective variables have on students. In order to research affect, an instrument to measure students’ writing apprehension was developed. (Cornwell & McKay, 1998b) Studies have shown that students with high levels of writing apprehension perform less successfully than those with lower levels (Powell, 1984; Frankinburger, 1991). It was felt that this measurement “could identify ‘at risk’ writers, predict academic success and present bench marks to measure treatments to lower writing apprehension.” (Cornwell & McKay, 1998b, p. 86)

 This study is a follow-up to Cornwell and McKay's 1998 study measuring Writing Apprehension (WA) at Osaka Jogakuin Junior College (OJJC). It examines whether the make up of peer editing groups has any effect on writing apprehension or grades over the course of a ten-week semester.

 The literature on peer editing in second language writing is extensive and is beyond the scope of this article. For a good overview see Mitan, 1989). Peer review provides students with larger quantities of more varied feedback than if done only by the instructor. (Chaudron, 1983, James, 1981) By serving as readers, students can better understand the needs of readers. (James, 1981; Lamberg, 1980; Witbeck, 1976) Peer review increases motivation since students receive input from actual readers (Chaudron, 1984, Koch 1982, James 1981) For these and other reasons peer review has become popular in many L2 language classrooms. However, it is not without its critics. In a review of peer editing literature, Zhang (1995) found little empirical research to support the view that peer feedback had an affective advantage over other forms of feedback. In fact, students often prefer to have teachers provide feedback; it is even sometimes their first choice (Cohen, 1987; Leki, 1991).

 As mentioned earlier, at OJJC peer work is encouraged in composition classes. In addition to the benefits mentioned above, peer editing “improves students’ ability to assess their own strengths and improves class atmosphere by actively involving students....” (Berg, 1999, p. 21)

 In a recent article on issues in Second Language Acquisition (SLA), Robinson (1998) asks

How do difference in affective an personality variables contribute to differential L2 learning success? While aptitude and intelligence are to a large extent fixed cognitive attribute of the learner, motivation and anxiety can be changed and shaped through teacher intervention in learning.

He then refers to Griffiths (1990) who advocated the need to develop personality measurements appropriate to Japanese audiences. Cornwell and McKay (1998b; in press) report on their attempt to validate a reliable measurment of anxiety, not personality, in a Japanese academic writing setting. They replicated a study that created an L1 writing apprehension measurement. Such a measurement of writing apprehension, if available in L2, could identify “at risk” writers and create a benchmark against which to measure the effect of any writing apprehension treatments. This study uses the Cornwell and McKay instrument to address the following research questions:

  1. Would group makeup as determined by level of writing apprehension have any effect
  2. on an individual student’s writing apprehension at the end of a ten-week semester?
  3. Would group makeup and writing apprehension have any effect on grades?

The Study


The participants were 31 first year junior college students from an intact Academic Writing class at OJJC. Two students dropped out so the final results are based on 29 students. At OJJC students are placed into classes according to their scores on a placement test, so for the purpose of this study, these students can be considered to have the same English proficiency.


 On both the first and last day of the semester students were administered a writing apprehension questionnaire in Japanese. An earlier study ( suggested that the questionnaire was valid and reliable for that administration. A factor analysis found four factors and accounted for 44.3% of the variance. The four factors were negative perceptions of writing ability, enjoyment of writing, fear of evaluation, and showing ones work to others. For a detailed description of the questionnaire and its development see Cornwell and McKay, 1998b and Cornwell and McKay, in press. Students were asked to answer 26 questions dealing with their attitudes toward writing. The following 5 point Likert scale was used: Strongly Agree, Agree, Not Sure, Disagree, and Strongly Disagree. The split half reliability of this study's first administration was .74. The full reliability for the questionnaire using the Spearman-Brown prophecy formula was .851. (Hatch & Lazaraton, 1991) The split half reliability of the second administration was .67 with the full reliability .80. See table one for a copy of the questions in English along with a breakdown of the percentage of students that chose each answer in the pre and post questionnaire.


Table 1. Questions with percentages for pre and post administrations

SA=Strongly Agree; A=Agree; U=Uncertain; D=Disagree; SD=Strongly Disagree

           SA       A        U         D      SD

1. I avoid writing.

   pre     0.00    24.14      10.34      44.83    20.69

   post    0.00     6.90      13.79      55.17    24.14

2. I have no fear of my writing being evaluated.

  pre      0.00     34.48      6.90      34.48     24.14

  post     0.00      6.90      6.90      58.62     27.59

3. I look forward to writing down my ideas.

  pre      10.34     34.48     17.24     37.93     0.00

  post     13.79     31.03     24.14     24.14     6.90

4. I am afraid of writing essays when I know they will be evaluated.

  pre      3.45     34.48      6.90     41.38     13.79

  post     0.00     17.24     13.79     44.83     24.14

5. Taking a composition course is a very frightening experience.

  pre      0.00     6.90      6.90     48.28     37.93

  post     0.00     3.45      3.45     44.83     48.28

6. Handing in a composition makes me feel good.

  pre       6.90    31.03     27.59    34.48      0.00

  post     17.24    31.03     27.59    20.69      3.45

7. My mind seems to go blank when I start to work on a composition.

  pre      0.00     17.24     10.34    48.28     24.14

  post     0.00     10.34      6.90    62.07     20.69

8. Expressing ideas through writing seems to be a waste of time.

  pre      0.00      0.00     0.00     27.59     72.41

  post     3.45      0.00     0.00     34.48     62.07

9. I would enjoy submitting my writing to magazines for evaluation and publication.

  pre     3.45      17.24     37.9     34.48      6.90

  post     3.45      6.90     51.72    20.69     17.24

10. I like to write my ideas down.

  pre      17.24    27.59     27.59     24.14     3.45

  post     24.14    24.14     37.93     10.34     3.45

11. I feel confident in my ability to clearly express my ideas in writing.

  pre      0.00      6.90     17.24     41.38     34.48

  post     0.00     10.34     44.83     27.59     17.24

12. I like to have my friends read what I have written.

  pre      3.45     20.69     24.14     41.38     10.34

  post     0.00     24.14     41.38     31.03      3.45

13. I'm nervous about writing.

  pre      6.90     27.59     17.24     27.59     20.69

  post     6.90     20.69      6.90     48.28     17.24

14. People seem to enjoy what I write.

  pre      0.00     6.90     62.07     17.24     13.79

  post     0.00     0.00     82.76     13.79      3.45

15. I enjoy writing.

  pre      20.69     34.48     27.59   13.79      3.45

  post     27.59     31.03     27.59   10.34      3.45

16. I never seem to be able to clearly write down my ideas.

  pre      0.00     48.28     20.69     27.59     3.45

  post     0.00     27.59     31.03     34.48     6.90

17. Writing is a lot of fun.

  pre      20.69     34.48     20.69     20.69     3.45

  post     17.24     48.28     24.14      6.90     3.45

18. I expect to do poorly in composition classes even before I enter them.

  pre      6.90     31.03     37.93     24.14      0.00

  post     0.00     17.24     48.28     27.59      6.90

19. I like seeing my thoughts on paper.

  pre      6.90     41.38     31.03     13.79      6.90

  post     6.90     55.17     20.69     13.79      3.45

20. Discussing my writing with others is an enjoyable experience.

  pre      6.90     37.93     37.93      13.79     3.45

  post     3.45     51.72      1.03      10.34     3.45

21. I have a terrible time organizing my ideas in a composition course.

  pre     10.34     48.28     27.59     13.79      0.00

  post     6.90     41.38      6.90     31.03     13.79

22. When I hand in a composition I know I'm going to do poorly.

  pre      6.90     20.69     41.37     24.14      6.90

  post     3.45     10.34     48.28     31.03      6.90

23. It's easy for my to write good compositions.

  pre      3.45      0.00     0.00     55.17     41.38

  post     0.00     3.45     10.34     65.52     20.69

24. I don't think I write as well as most other people.

  pre     17.24     31.03     34.48     17.24     0.00

  post     6.90     27.59     55.17     10.34     0.00

25. I don't like my compositions to be evaluated.

  pre      0.00     24.14     13.79     44.83     17.24

  post     0.00      6.90     13.79     48.28     31.03

26. I'm no good at writing.

  pre     27.59     27.59     24.14     20.69      0.00

  post     0.00     34.48     31.03     1.03       3.45

Note: The Japanese version of the questionnaire is available by contacting the author at OJJC or

The questionnaire also asked students to rate the level of their writing experience in high school. Students rated their experience writing on the sentence, paragraph, and essay level. A high school experience score was calculated ranging from a low of 3 to a high of 12. Prior to coming to OJJC, seventy-five percent of the students had little or no writing experience beyond the sentence level (Cornwell & McKay 1998b). See table two for a copy of the questions in English along with descriptive statistics and frequencies.


Table 2. High school writing experience questions and statistics

In high school how much writing experience did you have with the following:

                       a lot     some     not much     almost none

Sentences               4          3          2              1

Paragraphs              4          3          2              1

Essays                  4          3          2              1

Descriptive Statistics
High school                                           M        SD        Min.        Max.
Writing Experience                             6.23     2.17         3.00       12.00

Frequencies of Response
Value             Frequency         Percentage
    3                     4                         12.9
    4                     2                           6.5
    5                     5                         16.1
    6                     9                         29.0
    7                     3                           9.7
    8                     3                           9.7
    9                     3                           9.7
    10                   1                            3.2
    11                   0                            0.0
    12                   1                           3.2
                      -----------                 ----------
Total                 31                         100.0



Based on the scores from the first administration of the questionnaire, students were placed into eight groups of four (one group had three members). Group size was determined subjectively by the teacher who felt that four was the optimal small group size for a writing class. Following Daly and Miller (1975) high apprehensive and low apprehensive were defined as being one standard deviation above or below the mean. There were 6 students who were considered to be high apprehensive and 6 students who were considered to be low apprehensive. Three treatment groups were created: one comprised of high apprehensive (HA), one comprised of low apprehensive (LA), and one comprised of mixed apprehensive (MA), i.e. low and high apprehensive. Students were randomly assigned to these three groups based on their scores. The rest of the students were assigned to groups which served as a control. See table three for descriptive statistics and pre and post writing apprehension scores.

Table 3. Basic Descriptive Statistics & Pre and Post WA scores listed by grouping

                        N       M     SD      Min.   Max.

PRE-Wa                 29     76.21  12.34   48.00   95.00

POST-WA                29     70.31  11.16   43.00   90.00

CHANGE IN WA           29     -5.90  10.58  -29.00   16.00

 Group     Pre  Post     Final Grade  change

    1*     60     57         70         -3

    1*     63     65         86          2

    1*     54     70         82         16

    1*     53     67         77         14

    2      68     71         71          3

    2      72     63         81         -9

    2      67     62         84         -5

    2      72     57         85        -15

    3      79     72         82         -7

    3      77     78         81          1

    3      77     77         80          0

    3      74     70         81         -4

    4      84     81         85         -3

    4      81     77         61         -4

    4      83     64         84        -19

    5*     64     61         81         -3

    5*     95     66         83        -29

    5*     48     43         66         -5

    5*     92     83         85         -9

    6      80     76         82         -4

    6      81     67         83        -14

    6      79     50         84        -29

    6      79     82         84          3

    7      85     63         82        -22

    7      86     89         68          3

    8*     88     90         80          2

    8*     90     81         73         -9

    8*     88     82         81         -6

    8*     91     75         85         -16

* = treatment groups (1=Low, 5=Mixed, 8=High)

Group Members did not know they were grouped according to writing apprehension. Groups worked together on peer editing assignments and any other group work such as pre-writing assignments, group essay writing, and other group assignments.


 As mentioned earlier, on the last day of the semester students were again given the WA questionnaire. The scores of the pre and post questionnaires were analyzed using a repeated measures analysis of variance (ANOVA). See Table four. An initial ANOVA showed that there was a significant difference between the groups. Since an ANOVA can only show if there are significant differences, but not where they are located, a Scheffe post hoc test was run. Scheffe was chosen since it is a conservative test which is unlikely to falsely reject the null hypothesis. (Hatch & Lazaraton, 1991, p. 354) It showed one significant interaction p > .04 between treatment group 1 (LA) and treatment group 8 (HA) and the first administration of the WA questionnaire. The other groups did not have any significant interactions.

Table 4. Repeated measures ANOVA of 8 groups
(HA, N=4; LA, N=4; Mixed, N=4; & Control, N=17)
STATISTICA  summary of all effects; Effect 1=Group; 2=Time


Effect df Effect MS Effect df Error MS Error F p-value
1 7 414.32 21 144.30 2.87 .029
2 1 381.01 21 59.78 6.37 .020
1 * 2 7 93.91 21 59.78 1.57 .199


Finally, correlations were run between the WA questionnaire scores and the students final score. Taken as an entire class there were no significant correlations between the WA scores and the final scores. However, looking at individual groupings there were some interesting results. The HAs pre WA and post WA scores were negatively correlated at -.84 significant at p < .05.

Results and Discussion


Before starting the discussion, it is important to note several shortcomings this study has. The sample size was very small therefore any effect would have to be very large to appear. Only one group could be formed for each treatment (high, low, mixed) and each group could only have 4 members. In addition, one of the control groups started with three members but ended with two members after a student dropped out due to attendance problems. Given the small sample size drawn from an intact class, no generalizations to a larger population can be made. Furthermore, from time to time group members were absent necessitating an occasional combining of groups.

 Additional comments about the design should examine the possibility of a teacher effect. Although the researcher and the teacher were the same, after making the groups, group membership was not referred back to until the semester was over. This should lessen the possibility of a teacher effect. Another comment related to the teacher is that the teacher assigned the students’ final grade. The final grade is made up of students’ composition grades plus their class work grades. Since it was not possible to get other raters to rate the students’ work, it is not possible to report the reliability of the students’ final grades.

 A final subjective observation is that the low apprehensive group had some low achievers, not in terms of their grade, but in terms of their attitude toward the class; in other words, it had students who did not always appear prepared for class or who were often absent. This could have affected their performance. It is difficult to say whether or not the student’s HA was the reason for their low achievement.

Changes in Apprehension

 With the above shortcomings in mind, what can we learn from this study? It is interesting to note that while most groups apprehension level dropped (declines ranged from 2.5 to 11.5), the LAs' level rose 7.0 points. The mixed group attained the largest apprehension level drop with a fall of 11.5 and the high apprehensive level dropped 7.25.

 On an individual level some students' WA levels changed drastically between the pre and post administrations. While most groups' standard deviations ranged from below 1 to a little over 5, the mixed group’s was 22.65 on the pre and 16.46 on the post. This is to be expected since it is a group comprised of high and low scores. However, one of the members dropped 29 points from a pre WA score of 95 to one of 66. One of the control groups also had a member whose score fell drastically from 85 to 63, a drop of 22 points. Finally, one of the low apprehensive scores rose from 54 to 70, an increase of 16 points. Given the size of the study, these three large, extreme changes probably affected the results. (See Figure One).

Post hoc Interviews

Why did the low apprehension group’s apprehension rise while everyone else's fell? To find the answer to this question I interviewed some students. I was able to interview six students whom had been in one of the three groups of interest. Two of the students’ apprehension had risen, two had fallen and two had remained the same. and give them the writing apprehension measurement again. The interviews took place after the students had completed the first semester of their second year, and had thus experienced writing two papers for the content-based classes mentioned in the introduction. Table five shows students responses and writing apprehension scores.


Table 5. Student Interviews

Student Change Pre Post Interview HS writing Prior attitude
1 little 63 65 52 a lot easy/fin
2 up 54 70 56 none* difficult in any lang.
3 up 53 67 59 a lot easy
4 down 95 66 62 none nervous
5 little 48 43 54 a little** not afraid of mistakes
6 down 91 75 60 none*** confused

* wrote speeches in high school; likes writing letters in English
** wrote diary
*** wrote letters on her own

The low apprehensive group was comprised of students whom had quite a bit of high school writing experience or had studied abroad or had written letters on their own. This is consistent with Cornwell and McKay’s (1998b) finding that high school writing experience and writing apprehension are significantly negatively-correlated; as one goes up the other goes down. One of the students who studied abroad stated she had learned to not be afraid of mistakes; they were good for her. These students stated that upon entering school they weren’t nervous. Instead they thought writing was easy and looked forward to it.

 This is the opposite of what students with no writing experience expressed. For them, entering writing class was a confusing and nervous experience. They had no knowledge of what was expected of them and one even had negative ideas toward learning English. How these two groups’ apprehension changed is interesting and may provide areas for future research.

 Student 1 remained the same. When asked why she thought she remained the same, she replied, “I liked writing from my high school experience. I had a good experience writing before so I thought it would be fun to write.” This is in contrast with student 2 and 3 from the same group. Student 3 stated, “I first thought it was really easy—academic writing. But we learned about comparison and contrast and classification. I couldn’t understand the differences. Unit 1 and 2 were easy but Unit 3 and 4...I was very confused what is the difference between them.” One of these students changed her opinion on what the researcher labeled “enjoyment of writing.” She went from saying she enjoyed writing to saying she wasn’t sure. The others opinion changed in the area of “showing her writing to others” and in her “own opinion of her writing.” She went from a positive attitude to a negative one.
 Confusion also plays a role in the students whose apprehension went from high to low. Student 6 said that she was confused when entering the program , but that “writing became fun little by little during the first year, so in second year I had not worry. It wasn’t hard [to] write; it wasn’t was just writing.” Student 4 also reported that although she was nervous when starting, “ the end of the semester, [she] can understand the way of writing.”
 In addition to the changes that occurred it is interesting to see what students felt were the hardest and easiest parts of writing. Many students felt that doing the outline and collecting information were the hardest thing about writing a paper. They also mentioned the paper length (5 pages) and the fact that they had to write two papers at the same time. Surprisingly, several of the students mentioned that the actual writing of the paper was the easiest.
 Finally, it is interesting to see that all students’ writing apprehension dropped between the end of the first year measurement and the interview measurement. With the exception of one student, everyone’s interview writing apprehension score is lower than their post score and all are what were considered low apprehensive scores in the initial study.


 This study was done to see if group makeup would have any effect on WA or final scores. The results of this study seem to indicate that group makeup alone does not appear to change writing apprehension or affect final scores. An interesting anomaly in this study was the tendency for low apprehensives’ apprehension scores to rise. Due to the sample size and other problems mentioned earlier, these results are not generalizable. So, what does this study mean for us as teachers? Based on the data provided by the questionnaires and the interviews, teachers might consider the following:

 1) Prior writing experience has an influence on writing apprehension. While the measurement only looked at high school writing experience, the fact that all apprehension scores were in the low apprehensive range by the middle of the students’ second year is important. The more writing we can get our students to do, the less apprehensive they may be in their next writing class.

 2) Writing apprehension is not fixed, but can be changed over time. Therefore, apprehension scores should be seen as indicators of where a student is at one point in time. Also, we should be encouraged as writing teachers. If apprehension can be changed over time, there is hope for all our students.

 3) Prior experience alone will not develop successful writers. For some students academic writing’s rigor was a shock and their apprehension shot up. Others took it in stride and their anxiety remained low. We need to inform students of the difference between academic writing and other forms of writing they may have been exposed to.

 4) Understanding the writing process is one way that anxiety is lowered. This should be a priority in our writing classes. Also, confusion about rhetorical patterns’ relationship to research papers has a negative effect on students’ apprehension levels. (See Cornwell and McKay, 1998a for some suggestions on how to address this confusion).

 5) Students see the actual writing as the easiest part of the writing process. This should be encouraged while at the same time, more time needs to be spent on the so-called hard parts: outlining, coming up with ideas, introductions and conclusions, readability, etc.

 This study should be replicated with larger samples to see if this study’s lack of significant findings is an artifact of the small sample size. In addition a more in-depth study of students’ prior writing experience’s effect on their writing apprehension might be an area for future research. Also, a closer look at how students’ apprehension changes both positively and negatively would add to our knowledge of anxiety and it’s role in second language writing. Finally, a study to examine the connections between low achievement (i.e poor attendance/lack of preparation) and high apprehension is needed. Does one cause the other or are they coincidental?


  • Berg, E.J. (1999). Preparing ESL students for peer response. TESOL Journal, Summer 1999, 20-25.
  • Chaudron, C. (1983, March). Evaluating writing: Effects of feedback on revision. Paper presented at the annual meeting of the Teachers of English to Speakers of Other Languages, Toronto, Canada (ERIC Document Reproduction Service No. ED 227 706)
  • Chaudron, C. (1984). Evaluating writing: Effects of feedback on revision. RELC Journal 15, 1-14.
  • Cohen, A. (1987). Student processing of feedback on their compositions. In A. Weden & J. Rubin (Eds.), Strategies in language learning (pp. 57-69). Englewood Cliffs, N.J. : Prentice Hall.
  • Cornwell, S., & McKay, T. (1998a) Making the transition from writing short essays to long research papers. The Language Teacher, 22(4) pp. 15-20
  • Cornwell, S., & McKay, T. (1998b) Measuring writing apprehension at Osaka Jogakuin Junior College. Osaka Jogakuin Junior College Kiyo, 27 pp. 85-96.
  • Cornwell, S., & McKay, T. (in press). Establishing a valid, reliable measure of writing apprehension for Japanese students. Jalt Journal.
  • Daly, J.A., & Miller, M.D. (1975). The empirical development of an instrument to measure writing apprehension. Research in the Teaching of English 9, 242-249.
  • Frankinburger, P. (1991). An annotated bibliography of the literature dealing with the effects of student attitude and teacher behaviors on writing apprehension and composition quality. Exit Project, Indiana University at South Bend. ERIC Document Reproduction Services No. ED 333 482.
  • Hatch, E. & Lazaraton, A. (1991). The research manual: Design and statistics for applied linguistics. New York: Heinle & Heinle.
  • James, D. (1981). Peer teaching in the writing classroom. English Journal, 70 (7), 48-50.
  • Koch, R. (1982). Syllogisms and superstitions: The current state of responding to writing. Language Arts, 59, 464-471.
  • Lamberg, W. (1980). Self-provided and peer-provided feedback. College Composition and Communication, 31, 63-69.
  • Leki, I. (1991). The preference of ESL students for error correction in college-level writing classes. Foreign Language Annals, 24, 203-216.
  • Mitan, R. (1989) The peer review process: Harnessing students’ communicative power. In D. Johnson & D. Roen (Eds.), Richness in Writing: Empowering ESL Students. (pp. 207-219). New York: Longman.
  • Powell, B. J. (1984). A comparison of students attitudes and success in writing. Journal of Negro Education, 53, (2), 114-123.
  • Spielberger, C. D. (1983) Manual for the state-trait anxiety inventory (Form Y) . Palo Alto, CA: Consulting Psychologists Press.
  • Witbeck, M. (1976). Peer correction procedures for intermediate and advanced ESL composition lessons. TESOL Quarterly, 10, 321-326.
  • Zhang, S. (1995). Reexamining the Affective Advantage of Peer Feedback in the ESL Writing Class. Journal of Second Language Writing, 4 (3), 209-222.