The effect of peer response groups on writing apprehension, writing achievement, and revision decisions of adult communi...


Material Information

The effect of peer response groups on writing apprehension, writing achievement, and revision decisions of adult community college composition students
Physical Description:
ix, 146 leaves : ; 29 cm.
Sailor, Susan Hardee, 1963-
Publication Date:


Subjects / Keywords:
Instruction and Curriculum thesis, Ph. D
Dissertations, Academic -- Instruction and Curriculum -- UF
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1996.
Includes bibliographical references (leaves 129-144).
Additional Physical Form:
Also available online.
General Note:
General Note:
Statement of Responsibility:
by Susan Hardee Sailor.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
oclc - 35172099
System ID:

Table of Contents
    Title Page
        Page i
        Page ii
        Page iii
    Table of Contents
        Page iv
        Page v
        Page vi
    List of Tables
        Page vii
        Page viii
        Page ix
    Chapter 1. Introduction
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
    Chapter 2. Review of related literature
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
    Chapter 3. Research design and methodology
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
    Chapter 4. Data analysis and results of the survey
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
    Chapter 5. Summary, conclusions, and recommendations
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
        Page 115
        Page 116
    Appendix A. Writing apprehension measure
        Page 117
        Page 118
        Page 119
    Appendix B. Paul Diederich essay rating scale: Modified
        Page 120
        Page 121
        Page 122
        Page 123
        Page 124
    Appendix C. Instructor directions: Community college
        Page 125
        Page 126
    Appendix D. Taxonomy of revision changes
        Page 127
    Appendix E. Brooke's strategies for peer response groups
        Page 128
        Page 129
        Page 130
        Page 131
        Page 132
        Page 133
        Page 134
        Page 135
        Page 136
        Page 137
        Page 138
        Page 139
        Page 140
        Page 141
        Page 142
        Page 143
        Page 144
    Biographical sketch
        Page 145
        Page 146
        Page 147
        Page 148
        Page 149
Full Text








Perhaps the best part of a dissertation is thanking

everyone involved in its writing. My sincere appreciation is

expressed to all who have guided and encouraged me from the

beginning of my graduate study to the present. I am

especially grateful to Dr. Lee J. Mullally, Dr. Robert G.

Wright, Dr. Linda Crocker, Dr. Julie Dodd, Dr. Jeff Hurt, and

Dr. Craig Frisby for their faith in my abilities and for

giving me the gentle pushes and guidance necessary to

accomplish a project of this magnitude.

I particularly wish to thank the students, faculty, and

administration of Keystone Heights Jr./Sr. High School for

their concern, kindness, patience, and time. Having so many

caring people interested in this degree made it possible for

me to continue writing when times became hectic. Heartfelt

thanks go to the students and faculty of the SPEARS program

for continued inspiration and for being "number one" to me.

A special thank you is extended to the administrations,

faculties, and students of St. Johns River, Santa Fe, and

Central Florida Community Colleges, and North Florida Junior

College who participated in the study.

Finally, I'd like to thank my family: my parents, Robert

W. and Anna Hardee; my husband Joel; and my son Carson. Yes,

Carson, Mommy's finally finished on the computer and there

are two "doctors in the house."



ACKNOWLEDGEMENTS ........................................... ii

LIST OF TABLES .. ............................................ Vii

ABSTRACT .................................................. viii


1 INTRODUCTION ........................................... 1

Statement of the Problem .............................. 7
Need for the Study .................................... 8
Research Design and Methodology ...................... 13
Statement of Hypotheses .............................. 15
Delimitations ........................................ 16
Limitations .......................................... 18
Summary .............................................. 19

2 REVIEW OF RELATED LITERATURE.......................... 21

The Changing Faces of Community College Students ..... 21
Identity Negotiation Theory and the Older Student .... 23
Characteristics of Adult Learners .................... 24
Motivation of Adult Learners ......................... 27
Curriculum Design for Adult Learners ................. 28
Small Group Theory ................................... 30
Peer Response Groups ................................. 32
Adult Learners and Small Group Work .................. 41
Writing Apprehension ................................. 45
The Development of the Construct ................... 45
The Writing Apprehension Measure ................... 47
Writing Apprehension and Writing Achievement ....... 48
Changing Writing Apprehension ...................... 51
Causes of Writing Apprehension ..................... 54
Teaching Product Over Process: Student
Attitude Toward Writing .......................... 55
Revision ............................................. 59
Revision and the Composing Process ................. 59
Causes of Revision Decisions ....................... 60
Revision Classification Systems .................... 61
Peer Response Groups and Revision:
A Natural Combination ............................63

Operations of Revision ............................. 65
Summary .............................................. 69

3 RESEARCH DESIGN AND METHODOLOGY....................... 71

Design of the Study .................................. 72
The Sample ........................................... 74
Description of the Instruments ....................... 76
The Writing Apprehension Measure ................... 76
Writing Achievement ................................ 77
The Taxonomy of Revision Changes ................... 79
Data Collection Procedure ............................ 80
The Rating Procedure ................................. 81
Data Analysis ........................................ 83


The Statistical Treatment of the Data ................ 84
Introduction ....................................... 84
Writing Apprehension ............................... 85
Analysis of Apprehension Data ...................... 86
Writing Achievement .................................. 90
Comparison of Raters ............................... 91
Analysis of Composition Data ....................... 91
Revision Decisions ................................... 95
Introduction ....................................... 95
Analysis of Revision Data .......................... 96
Summary ............................................. 102


Summary ............................................. 105
Conclusions ......................................... 109
Writing Apprehension .............................. 109
Writing Achievement ............................... 109
Revision Decisions ................................ 109
Discussion .......................................... 110
Recommendations .................................... 112
Future Research Inquiries ........................... 114


A WRITING APPREHENSION MEASURE ......................... 117



D TAXONOMY OF REVISION CHANGES ......................... 127


REFERENCES .............................................. 129

BIOGRAPHICAL SKETCH ..................................... 145


Table page

3-1 Composition of Subject Sample by
Course and Age ..................................75

3-2 Composition of Subject Sample by
Course and Sex .................................. 75

3-3 Composition of Subject Sample by Group..........76

4-1 Pre- and Postapprehension Mean by
Course and Group................................ 87

4-2 Summary of ANCOVA Postapprehension..............88

4-3 Summary of Correlational Analysis for
Age and Writing Apprehension.................... 89
4-4 Pre-and Postachievement Means by
Course and Group................................ 92

4-5 Summary of ANCOVA Postachievement...............94

4-6 Summary of Correlational Analysis for
Age and Writing Achievement..................... 95

4-7 Pre- and Postrevision Means by
Course and Group................................ 98

4-8 Summary of ANCOVA Postrevision Changes......... 100

4-9 Summary of Correlational Analysis for
Age and Revision Decisions..................... 101

4-10 Summary of Findings by Course Level............103

4-11 Summary of the Relationship of Findings
to Hypotheses by Course........................ 104

Abstract of Dissertation Presented to the Graduate School
of the university of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy



Susan Hardee Sailor

May, 1996

Chairman: Dr. Lee J. Mullally
Major Department: Instruction and Curriculum

The purpose of this research was to investigate the

effect of peer response groups on students' writing

apprehension, writing achievement, and revision decisions. A

special emphasis was placed on students over the age of 25.

Writing apprehension levels were measured by the Daly-Miller

Writing Apprehension Measure. writing achievement was

measured by the holistic evaluation of compositions using the

Diederich Essay Rating Scale. Revision decisions were

evaluated by Faigley and Witte's Taxonomy of Revision

Changes. The three measures were completed by 110 community

college students in freshman level composition courses at the

beginning and ending of a sixteen-week period. The treatment


groups participated in peer response groups for 30 minutes

per week during the same time period.

After adjusting for pretest differences, three

hypotheses were tested (p. < .05) separately for ENC 1101 and

ENC 1102 classes: There will be no significant differences

among the posttest writing apprehension scores, posttest

writing achievement scores, and posttest revision decision

scores of the nontreatment and treatment groups; there will

be no significant effect of age on the adjusted posttest

scores for the three measures of nontreatment and treatment

groups; and there will be no significant interaction effect

of age and treatment on the adjusted posttest scores of the

three measures in the nontreatment and treatment groups.

The null hypotheses were rejected for writing

apprehension and writing achievement. They were retained for

revision decisions of both ENC 1101 and ENC 1102 students.

ENC 1101 students demonstrated a higher degree of writing

apprehension and writing achievement than students in ENC

1102. Age showed a direct positive correlation with both

apprehension and achievement.

More research should be conducted concerning the

reduction of writing apprehension while increasing writing

achievement. Findings at the community college were

consistent with prior research at the college level which

suggests that apprehension is associated with writing

achievement, and that freshmen students revise mainly at the

surface level.


In another decade, it is estimated, the typical

community college student will be over the age of twenty-

five. Adult learners, defined here as students age 25 and

older, are fast becoming a large segment of the community

college student population. With some reports estimating

these students comprising as much as 50 percent of the school

population, instructors should reevaluate their teaching

strategies in terms of reaching a diverse group of students,

many of whom will not fit the mold of the traditional 18-

year-old enrollee. Even the term formerly associated with

adult learners, nontraditional students, is becoming archaic

as older students enter with increasing frequency through the

open door.

Unfortunately, we don't know much about which

instructional techniques work best with the adult learner,

and much of what we do know is extrapolated from research

conducted on children and adolescents. Many studies have

sought to answer questions about why adults return to school,

with the consensus being that to acquire or keep a job new

knowledge is required, or that a major life event such as

divorce, children leaving home, or the death of a spouse

precipitates the drive to learn. Some studies report the

characteristics of adult learners by income level and amount

of previous education. But few studies have examined actual

teaching strategies and their effect on the academic

achievement and/or social adjustment of adult college

students. Even fewer studies have investigated whether gender

plays a role in effective teaching strategies for adult


Robert Brooke (1991) presents a theory of identity

negotiations occurring in the college writing classroom.

Brooke's theory provides an appropriate lens through which to

view writing instruction for adult learners. The theory

shifts the student's role as writer from regurgitator to

reflective thinker. This shift proves useful for students

beyond the writing classroom. As students negotiate and

embrace the roles a writing situation offers and make them

part of their ongoing behavior, they are learning.

The identity negotiation theory sees learning as

influenced more by the roles offered to students in school

than by any particular content or material being taught. It

is in negotiating a response to these roles that individuals

work out their future stances toward knowledge, toward

authority, and toward academic learning, stances which could

spell success or failure for the adult learner (Brooke, 1991,


Identity negotiation highlights the development of the

self within a complex arena of competing social forces. Two

assumptions guide Brooke's theory: first, that social

interaction and role expectations affect a student's choices

in negotiating his or her identity in the writing classroom,

and second, that the ever-present classroom forces of

compliance and resistance can be mediated through the use of

writing workshop format. In the first assumption, identity

negotiation focuses on what the writing class establishes as

the writer's role and what versions of these roles

participants develop as class progresses. For example, a

typical community college student may hold many different and

conflicting social roles: mother, wife, employee, and

student. In the writing classroom, participants must

negotiate a stance toward the activity of writing and what it

means in the social contexts around them. In a freshman

composition class, they are also simultaneously negotiating

their stance toward the school setting itself at the same

time. Identity negotiation theory can provide a portrait of

the important social context the individual experiences

within the writing classroom.

The second assumption begins with the notion that

college writing classes provide some inescapable roles. These

classes are required for graduation, so all sorts of

experiences and feelings about writing are represented within

the participants. Rose (1988) points out that college writing

classes act as a kind of initiation or testing ground, an

environment which all must endure to achieve the role of

college graduate. In this setting, individuals find it hard

to escape the role of a student dependent upon the teacher

for the "right" academic knowledge and must generally comply

with this role if they are to remain in school. In contrast,

Brooke argues that writer's roles are not as inescapable.

Writers are free to embrace, comply with, or reject their

words as they wish.

In many college writing classrooms the roles for

students and writers are opposed to each other. In contrast

to the passive student role, the writing classroom can

present writer's roles which value active and personal

engagement in learning, and can facilitate that negotiation

through the use of collaborative activities. Learning to

write becomes more important when it stems from writers'

roles which enhance an individual's sense of social self

(Cooper 1989). Students' ability to write, motivation to

write, and their effectiveness when writing in contexts

beyond the classroom depend upon the development of such

writer's roles.

Brooke focuses his discussion of identity negotiation on

the writing workshop, a format which allows a student free

choice of topic, ample time for drafting and revision, small

peer group critiques and teacher conferences. Writing

workshops allow students to focus energy directly on writer's

roles as opposed to traditional student roles, and help them

in identifying and exploring writing opportunities in their

lives beyond the classroom. Similarly, peer response groups

shift the learner's role from passive to active, and the

immediate verbal interaction of writer and readers helps to

solidify the newly negotiated writer's role. Writers in peer

response groups may see that they can use writing as a means

of addressing other roles they face: professional, personal,

gender, ethnic. Confident learners and writers are ready to

identify and explore various aspects of their social

cultures, a real benefit if college writing instruction is to

prepare students for thoughtful participation in their own


The community college freshman composition instructor

encounters many adult learners who bring with them a fear of

writing which can inhibit or color their identity

negotiations of the writer's role. Writing apprehension is

defined as the general anxiety that individuals might

experience about writing as determined by the Daly-Miller

Writing Apprehension Measure. Although most of us have

experienced some degree of writing apprehension toward a

composition task, adult learners are anxious for a variety of

reasons: they haven't written an essay in 20 years, they

didn't experience success with writing in their high school

curriculum, or they fear that they are not as smart as

younger students. Real or imagined, these fears interfere

with the development of writing skills, the attainment of the

writer's role, and exhibit themselves in these students'

negative attitudes toward writing. Fear may keep some

students from ever achieving the active engagement Brooke

purports is necessary to assume the role of competent writer

and thus increase a sense of social self-worth.

The use of peer response groups, a popular collaborative

strategy in the teaching of composition, may help to reduce

adult learners' writing apprehension. A peer response group

is a small group of three to five members that meets to

fulfill three goals: improving student texts, helping the

writer to develop more strategies for writing and a better

sense of how readers respond to the text, and helping writers

see the social value in the kinds of writing they are doing

(Brooke, 1994). The groups allow students to share their

writing strengths and weaknesses as they respond to the

papers of other group members in constructive and meaningful

ways through the use of teacher-provided guiding questions.

Peer response groups may also serve to increase

students' writing achievement, defined here as a student's

competence in writing as measured by the holistic scoring of

compositions, and levels of revision, which refers to a

student's decision to change an aspect of his/her paper.

Revision can occur on seven different levels: surface, or

spelling and capitalization changes; lexical, phrase, clause,

sentence, or multi-sentence levels, which cover addition,

deletion, or substitution of words, phrases, clauses, and

sentences; and text level, which is a change in audience,

function, overall content, or total re-write of the essay

(Bridwell, 1980). Revision choices demonstrate the extent to

which the student has adopted the writer's role and dropped

the traditional student model.

Statement of the Problem

Many adult learners come to college with a high degree

of writing apprehension that interferes with their ability to

succeed in composition classes. Their lack of confidence and

low achievement in writing coupled with the fact that they

have few academic and social support networks in place on

campus sometimes prevents them from fully accepting the

invitation to academic discourse and the writer's role

offered by the composition class.

Daly, a major investigator of writing apprehension,

found that the construct inhibited students in several ways.

Apprehensive writers avoid writing tasks and instruction and

as a result do not get sufficient practice to develop as

competent writers. Highly apprehensive writers use

significantly less intense language, and use fewer words and

make fewer statements than low apprehensives. High

apprehensives reported less writing success in the past and

expected less in the future, a self-fulfilling prophecy in

many instances (Daly, 1977, 1978; Daly & Miller, 1975;

Faigley, Daly & Witte, 1981).

Writing apprehension prevents adult learners from

achieving the new identity of competent writer necessary to

successfully complete college composition classes. It also

limits their participation in writing opportunities outside

of class which may enrich their other social roles.

Need for the Study

When an educational institution's clientele changes

dramatically, it has a responsibility to adapt to meet the

needs of its new members. According to the Almanac of Higher

Education (1991) students over the age of 25 now make up 63.6

percent of the community college student population. In

Florida, the percentage is even higher--67.9 percent. From

1970 to 1985 the community college enrollment of adult

learners increased by 114 percent nationally, while students

under the age of 25 grew by only 15 percent (Watkins, 1990).

The most recent projections by the U.S. Department of

Education indicate that two-year colleges will grow by

another one million students in the upcoming decade,

suggesting that these institutions will play a major role in

higher education well into the twenty-first century

(Reynolds, 1994).

Although a 1989 Almanac of Higher Education survey of

9,996 two-year college faculty states that 85 percent of the

respondents say their students are seriously underprepared in

basic skills such as those required for written and oral

communication, the Carnegie Commission estimates that only 10

percent of colleges nationally have taken steps to assist the

adult learner in any concrete way (Clark, 1989).

Much of what we know about adult learners comes from

three relatively limited areas: (1) "my life and times in

teaching" articles, (2) "why adults decide to study"

research, and (3) extrapolation from theory derived from work

with children and/or adolescents (Zemke & Zemke, 1981). Other

research on adult learners concentrates on adult basic

literacy skills such as reading or balancing a checkbook

(Imel, 1991; 1992). How can the community college instructor

meet the mission of designing strategies for adult learners

when what we know about their preferences is sketchy at best?

Morrison (1994) notes that adult learners often bring a

set of fears and anxieties with them to the classroom, the

most common being lack of basic knowledge or skills, the fear

of making mistakes, taking tests, or writing "good" papers,

the fear of being slower than younger students, the fear of

juggling a job, family and school, and a general fear of

failure. They experience discord between the self-image of

competence outside the classroom and the student image of

incompetence and inferiority (Evans, 1989). Adults tend to

take errors personally and are more likely to let them affect

their self-esteem than younger students. Adults cite

increasing or maintaining a sense of self-esteem and pleasure

as strong secondary motivators for engaging in learning

experiences They have something real to lose in the

classroom--self-esteem and ego are on the line (Zemke &

Zemke, 1981).

Muench (1987) recognizes that adult students need help

in building their self-confidence as students, in acquiring

or refreshing study skills, and in managing their time and

other resources while in school. They also benefit from

opportunities to interact with their peers and need to be

actively involved in the educational process through sharing

their relevant work and life experiences.

Students with a high degree of writing apprehension

generally lack self-confidence and avoid writing situations

whenever possible (Donlan, 1986). When they do write, highly

apprehensive students are not likely to be receptive to

instruction (Smith, 1984).

Several researchers have suggested that prior negative

experience with writing is generally the source of the

development of negative attitude toward writing (Daly, 1979;

Harris, 1978; Rose, 1980). Smilkstein (1990) proposes that

learning is a physical activity. Doing, trying, making and

correcting mistakes is a vital part of the activity of

learning. When negative emotions are present, they physically

inhibit the learning process. Seller, Garrison, and Bookar

(1978) support this idea, noting that there is a significant

correlation between writing apprehension and course grades.

Students who have never experienced success with writing

are often discouraged by initial frustrations and discomforts

felt at the beginning of a writing assignment and allow these

attitudes to interfere with their writing. They miss the

opportunity to become better writers through practice and

they do not see the act of writing as rewarding. Students

with a low degree of writing apprehension are able to put

initial negative feelings into perspective because they are

self-confident and exhibit positive attitudes toward the act

of writing. However, the best approach to take to reduce

writing apprehension has not been clearly established by

research (Hillocks, 1986).

Although the need for collaboration and participation is

emphasized in adult education literature, there is little

empirical support for collaborative learning as the best way

to educate adults; there is also little discussion of

collaborative learning itself, that is, what it is, how it is

implemented, and its strengths and weaknesses. Research at

the primary and secondary levels suggests that students learn

better through noncompetitive, collaborative group work than

in classrooms that are highly competitive and individualized.

Whether or not that is true with adults is still largely

untested (Bruffee, 1987, Imel, 1992). Benshoff (1991) reports

that adults generally prefer more active approaches to

learning and value opportunities to integrate academic

learning with their life and work experiences, opportunities

which can be provided by collaboration in the classroom,

which also aid in the negotiations of new roles for adult


Several research studies dealing with peer groups and

the evaluation of writing indicate that improvement in theme-

writing ability and grammar usage, when small groups of

students engage in peer evaluation may equal or exceed the

improvement that occurs under evaluation procedures carried

out by the teacher alone (Ford, 1973; Lagana, 1972; Maize,

1952; Pierson, 1967; Sager, 1973; Sutton & Allen, 1964).

Additional research by O'Donnell (1985) and Hillocks (1986)

suggests that peer response or evaluation significantly

improves writing ability. However, a landmark study by Fox

(1980) showed no significant difference in college writing

classrooms using the peer response group technique versus

those using teacher-only feedback.

Exactly how response groups work to increase writing

achievement is not known. Also uncertain are which facets of

the composition process they improve. If the popular process

approach to writing is accurate, steps from prewriting to

drafting to revision to publication could be affected by the


The area of revision warrants a closer look. Sommers

(1980) indicates that many students resist changing anything

except mechanics once they have a draft of a paper. Part of

that resistance stems from the tediousness of rewriting but

another part results from students' inability to resee their

papers, to develop a reader's perspective, to assume the role

of a writer. Peer response groups offer an excellent

opportunity for students to practice the difficult role

reversal Fleckenstein (1992) describes: "Writers need to

perceive the desires or expectations their texts arouse in

their projected readers and then check to see if those

desires are satisfied" (p. 82). Again, empirical evidence on

peer response groups and their effect on the composition

process is limited.

Research Design and Methodology

In order to fulfill the purposes of the study--to

investigate the effects of peer response groups on students'

levels of writing apprehension, writing achievement, and

level of revision--eight freshman composition instructors

participated from Central Florida, St. Johns River, and Santa

Fe Community Colleges, and from North Florida Junior College.

Four instructors led nontreatment groups, conducting their

classes as similar lecture, reading, assignment courses where

drafting and revision was done outside of class, without peer

interaction, in response to teacher comment only.

Nontreatment group refers to those students who were asked to

write an expository composition and to complete the Daly-

Miller Writing Apprehension Measure at the beginning and the

ending of a sixteen-week period.

Four other instructors led treatment groups, which used

peer response groups for a sixteen-week semester in their

classrooms following a prescribed protocol. The treatment

group completed the same tasks as the nontreatment groups,

except that these students participated in a peer response

group which met weekly for 30 minutes to share and critique

their writing assignment rough drafts.

These community colleges, with the exception of Santa

Fe, had a similar type of student population--mostly

rural/suburban, with similar percentages of male/female and

traditional/adult learners. Santa Fe serves a younger student

body, with many students transferring to the University of

Florida after completing their associate's degrees. The

college was included in the study because of its well-

developed program to help displaced homemakers gain a college

education. These colleges also approximated the gender/age

distribution of other community colleges in the Florida

community college system. Teachers included in the study were

matched as closely as possible by teaching evaluations or

teacher profiles supplied by the academic dean or department


A nonequivalent control-group pretest-posttest design

was employed (Mason & Bramble, 1989). Although random

assignment of individual students to treatment and

nontreatment groups was not possible, intact classes were

randomly assigned as treatment or nontreatment.

The Daly-Miller Writing Apprehension Measure was

administered as a pretest to every student during the first

week of the study and as a posttest during week 16. In

addition, each student was given two hours during week one

and week 16 to write an expository essay which incorporated

narration in response to one of two equally difficult writing

prompts. These essays were holistically scored by four

professional raters. Holistic scoring is a guided procedure

for rating written pieces. The scoring occurs quickly,

impressionistically, by trained raters. A scoring guide which

describes desirable characteristics of writing and identifies

high, middle, and low quality levels for each feature is

usually used. In this study, a slightly modified version of

the Paul Diederich Essay Rating Scale was used as the scoring

guide. It is divided into two areas: (a) general merit, which

includes ideas, organization, wording, and flavor; and (b)

mechanics, which includes usage, sentence structure,

punctuation, capitalization, abbreviations, numbers,

spelling, handwriting, and neatness. The handwriting and

neatness components were removed to modify the scale for this

study. The raters had experience scoring the College Level

Test of Academic Skills for the State of Florida Department

of Education. They predictably had a high interrater

reliability. Finally, first draft and final copies of two

randomly selected essays per student were compared and

analyzed for revision types and frequencies using the

Taxonomy of Revision Changes devised by Faigley and Witte


An analysis of covariance was performed for each of the

dependent variables (Kennedy & Bush, 1985; Mason & Bramble,

1989). A simple bivariate correlation was also performed to

identify strengths of relationships between variables.

Statement of Hypotheses

The following research hypotheses were tested:

HI: There will be no significant differences (p < .05)
in mean adjusted posttest scores of nontreatment and
treatment groups on writing apprehension, writing
achievement, and revision measures.

H2: There will be no significant effect (p < .05) of age
on the adjusted posttest scores of nontreatment and
treatment groups on writing apprehension, writing
achievement, and revision measures.

H3: There will be no significant interaction effect
(p < .05) of age and treatment on the adjusted
posttest scores of nontreatment and treatment groups
on writing apprehension, writing achievement,and
revision measures.


Random sampling is the preferred method of doing

experimental research; however, a quasi-experimental design,

the non-equivalent control-group pretest-posttest design, was

selected for this study due to the necessity of using intact

college classes.

Eight instructors from area community colleges

participated in the study. These teachers taught at least two

sections of freshman composition. Four teachers led

nontreatment groups and four led treatment groups. Treatments

were randomly assigned to four of the groups. Pretest and

posttest data on writing apprehension, writing achievement

level, and level of revision changes were collected by the

principal investigator using specific data collection

procedures. The same procedures were followed by each

teacher, thus teacher differences were controlled.

The subject population represented a balanced sample of

community college students today. It is safe to conclude that

the findings in this study could be generalized to the same

types of populations and courses described herein.

The sixteen-week duration of the study was not a

limitation. In Hillocks' (1986) metanalysis of 60 composition

research studies, virtually no difference in effect size was

found in studies under or over 17 weeks, thus disallowing the

reporting of the link between duration and increased effect

size with any confidence.

Differences in holistic scorers or raters was minimized

through the careful selection of writing prompts which

elicited expository papers incorporating narration. Hake

(1986) found that in first-term college entry and exit essay

exams, pure narrations were more likely to be misgraded than

expositions that employed narration as a developmental

strategy. She suggests specific prompts which have been used

at Chicago State University and California State University

at Los Angeles which were found to produce a very slight

degree of grader variance. These prompts were used in this

study. Scorers were unaware whether the essays were entry or

exit, as Hake also found exit essays to be rated lower than

entry essays when the categories were known by raters.

Scorers were four teachers trained in holistic scoring

employed by the State of Florida Department of Education in

scoring the College Level Academic Skills Test (CLAST). These

scorers received extensive training and scoring opportunities

through their association with the CLAST, so interscorer

reliability was high.


Five limiting factors were identified in this study.

Limiting factors were history, maturation, teacher

differences, mortality, and pretest effects.

This study was limited to students in preestablished

groups of those teachers who teach at least two sections of

ENC 1101 or ENC 1102. Each teacher was randomly assigned a

nontreatment or treatment group in an attempt to control for

teacher differences. History and maturation were additional

threats to internal validity because of the uncertainty that

all groups would be exposed to the same events and that they

had similar maturational processes.

Enthusiasm for the use of peer response groups cannot be

controlled; some teachers may display a more positive

attitude about the technique than others. Some teachers are

also better teachers than others; this cannot be controlled

although every attempt was made to locate superior teachers

through the recommendations of an academic dean or department


Mortality was controlled by using only the scores of the

students who completed both the pretest and the posttest of

the dependent variable being measured.

There may be a pretest effect on the Daly-Miller Writing

Apprehension Measure since all students completed the

instrument at the beginning and end of the 16-week period.

Kaywell (1987), in a study employing the same measure,

minimized the effect by telling subjects that the test

measured writing enjoyment, not apprehension. Since both the

nontreatment and treatment groups underwent the same

procedure, the pretest effect was probably negligible.

Similarly, the Hawthorne effect was minimal, since the

subjects' knowledge that they were participating in an

experiment was universal.

The pretest and posttest writing samples were controlled

for time, topic, and procedure for administration. Only one

composition was obtained from each student at the beginning

and ending of the study. A slightly modified version of the

Paul Diederich Essay Rating Scale was used to determine

change in writing ability from these samples. Each paper was

rated by four professional holistic scorers and tests of

interrater reliability were performed. The scores of the four

raters were averaged to provide a single score for each of

the two compositions.


The community college student population is rapidly

changing. Adult learners, many of them women over the age of

25, are becoming familiar faces in composition classrooms,

and they sometimes bring with them a level of writing

apprehension which interferes with their ability to succeed

in writing classes and causes the uncertainty of resolving

the conflict that their many social roles create.

A positive attitude toward writing is a desirable

characteristic; therefore, a reduction of the negative

attitude of writing apprehension must occur before extremely

anxious writers can overcome their inhibitions about writing

and its evaluation.

Concerned community college instructors would likely

adopt strategies that are shown to reduce writing

apprehension and increase writing achievement with adult

learners, but few researchers have at this time investigated

specific methods for reducing writing apprehension with these

students. Although some apprehension is necessary for

students to be conscientious about their work, it has been

shown that too much anxiety has negative effects. Research is

needed that focuses on reducing levels of writing

apprehension while increasing writing achievement and the

resultant effects on level of revision. The peer response

group technique is intended to alleviate adult learners'

writing apprehension as well as increase their writing

proficiency as they adopt the role of the writer within and

outside of their composition classroom.


The purpose of this study was to investigate the effects

of peer response groups on community college students' levels

of writing apprehension, writing achievement, and revision,

with special emphasis on students over the age of 25. The

review shows that there is a need for specific quantitative

research into writing apprehension of older college students,

the effectiveness of small group theories such as peer

response groups, and the effects of small group strategies

and the reduction of writing apprehension on student revision


The Changing Faces of Community College Students

The makeup of community college student population is

changing as the population of the country changes also.

According to the Almanac of Higher Education (1991), 63.6

percent of the United State's population is older than age

25, and in Florida that figure increases to 67.9 percent. At

the present time, over four million students are enrolled in

public two-year institutions, with 43 percent of them

enrolled in classes part-time due to jobs or family

obligations. In Florida, the figure is even higher--51

percent of them are part time. Florida awards more

associate's degrees each year than bachelor's degrees.

Older students are more likely to comprise the part-time

picture than are younger students due to work or family

obligations. The Report for Florida Community Colleges (1993)

lists twice as many part-time students enrolled than full-

time students--202,837 as opposed to 105,539. Of those part-

time students, 78,957 are male and 123,880 are female.

By the end of the decade Grennan and Schneider predict

the percentage of students over 25 years of age will equal

that of students under 25 (1989). Watkins (1990) supports

that conjecture. He found that in the years 1970 to 1985 the

enrollment of students over the age of 25 increased by 114

percent, while the enrollment of students under the age of 25

grew only 15 percent. Hirschorn (1988) concurs. He reports 6

million adults studying for college credit, a figure which

comprises 45 percent of all college students. Of these, 67

percent are studying for associate's or bachelor's degrees.

Students over the age of 25 come with their own unique

set of fears and anxieties which pose instructional dilemmas

for conscientious instructors. Morrison (1994) lists these

fears as a lack of basic knowledge or skills, making

mistakes, taking tests, writing "good" papers, being slower

than younger students, juggling job, family, and school,

being the only older person in class, and a general fear of

failure. These fears, real or imagined, can affect these

students performances in the classroom.

The Almanac of Higher Education (1991) seems to support

the fact that many of these fears are real. In a 1989 survey

of two-year college faculty, 85 percent said that older

students were seriously underprepared in basic skills such as

those required for written and oral communication.

Conversely, the Carnegie Commission estimates that only 10

percent of colleges nationally have taken steps to assist the

adult learner in any concrete way (Clark, 1989).

Identity Negotiation Theory and the Older Student

One of the major problems facing the older community

college student is the need to negotiate a new role for him-

or herself--the role of a member of an academic community.

These students are faced with several competing and

conflicting roles already, such as worker, spouse, and

parent. Brooke (1994) explains a theory of identity

negotiation that must occur within these students if they are

to experience success in their college classes. Brooke notes

that one consistent need within all students is the need to

connect the life of thinking, reading, and writing (the so-

called life of the mind) with the other lives they lead. One

of the strategies instructors can use to help make this

connection is that of group learning (Needham, 1994). Needham

found that precise group formation can stimulate sparks in

the most mundane but necessary subjects.

How does this translate to composition classes? Beaven

(1977) states that when peers have regular opportunity to

share their writing and to take part in evaluation

procedures, they exercise power or control over decisions

that affected their own work. Furthermore, as the dynamics of

small group work evolve, peers develop a sense of group

inclusion, acceptance, support, trust, reality testing and

collaboration. She states that as students share and focus on

unique qualities of each paper, they begin to appreciate

differences in approach, content, organization, flavor and

wording. Then students come to expect differences, and their

expectation frees students to say or write their "own things"

in their own ways. Further, Brooke (1994) believes that peer

response helps writers develop the feelings of social

approval necessary to continue writing, an understanding of

audience reactions, and their own writing processes, and the

ability to revise particular pieces effectively.

Jacobson (1991) notes that although the institutional

and instructional environments of most colleges are

constructed with the intention of providing the adult learner

access to higher education, there are obstacles present

resulting from conflicting social forces. The adult writers

in her study showed limited growth resulting from their

inability to understand the course-related writing as a

particular style of discourse characterized by academic

community-governed expectations and rules.

Characteristics of Adult Learners

There are three main reasons that adults choose to
attend college: those highly motivated from an economic

factor, those needing mental stimulation, and those seeking a

"last chance" to gain education (Needham, 1994). Although

these motivations may not seem to deviate much from those of

younger students, adults in the classroom are, as Needham

puts it, a "new breed" of college students.

This new population brings with it its own set of

problems, worries, and fears. The number one perceived

problem is age difference, which brings with it isolation and

fear. A number of students in Morrison's (1994) study

reported feeling uncomfortable being outnumbered by the

younger students in their composition classes. Many said if

they were to do it again, they would take a night class where

a higher percentage of students tend to be adults. Younger

students report a certain irritation with the willingness of

adult students to participate in class discussions. Older

students are more outspoken and willing to express their

opinions. This made traditional students feel inferior, and

these students called for instructors to work to create a

more cooperative atmosphere between the two groups. Older

students cited the younger group's lack of preparedness and

irreverence for and lack of respect for the instructor as

major problems of being in a mixed-age classroom.

This competition based on age expands itself to grades.

Younger students think older students have only one or two

classes to study for, but 85 percent of adult students work

full time. They are assumed by their younger classmates to

have more time to study. Outside of class, a segregation

exists which is typical: there is actually more intermingling

of the ages in class than in the student lounge. Tobin (1991)

asserts that the primary thing writing teachers need to do in

order to succeed is to establish productive relationships

with and between their students.

Courage (1990) studied 24 adult learners in a first-

semester basic writing course. He describes the experiences

of the students who enter the college classroom as if it were

a foreign land with its own peculiar customs. The

difficulties encountered by the students as they try to adapt

to the academic community's patterns of language use are

examined. He states that the course was a meeting ground of

cultural and language differences, with tension, conflict,

and misunderstanding. He describes the subjects as being

constrained by limited resources, relying heavily on peer

interpretation of academic language patterns, and generating

a de facto curriculum for their life concerns. He points out

that there exists an overlay of new language patterns and

purposes associated with academic discourse onto old patterns

and purposes associated with the language of other

communities to which they belong, confirming the struggle of

students as they negotiate new identities for themselves.

The study of adult education, andragogy, is a subject of

concern for many instructors in higher education. Zemke and

zemke's (1981) article titled "Thirty Things We Know For Sure

About Adult Learners" is Training magazine's leading reprint

request. In that article, the authors describe adult learners

in the classroom, their unique motivations to learn, and the

implications for curriculum development raised by these


In the classroom, Zemke and Zemke point out that adults

have something real to lose. Self-esteem and ego are on the

line. Adults tend to take errors personally and are more

likely to let them affect self-esteem. Evans (1989) agrees

with this fact. She states that adult learners forget that

they have performed admirably at work and at home for many

years; there is discord between their self-image of

competence outside the classroom and the student image of

incompetence and inferiority. This life experience can be an

invaluable asset that Zemke and Zemke assert needs to be

acknowledged and tapped and used.

Motivation of Adult Learners

Tough (1968), Aslanian & Brickell (1980), Rubenson

(1977), and Miller (1967) have all identified "teachable

moments" in the lives of adults, at which points) they are

most likely to turn to education. Adults seek out learning

experiences in order to cope with specific, life-changing

events. Marriage, divorce, new jobs, being fired, retiring,

being promoted, losing a loved one, or moving to a new city

are all examples of a life-changing event that may lead an

adult back to school. The more life-changing events an adult

encounters, the more likely he or she is to seek out learning

opportunities. Learning is a coping response for significant

change. Once convinced that the change is a certainty, adults

will engage in any learning that promises to help them cope

with the transition. Secondary motivations are increasing or

maintaining their sense of self-esteem and pleasure.

Curriculum Design for Adult Learners

Research on lifelong intellectual development conducted

by R.B. Catell (1963) suggests that there are two distinct

kinds of intelligence that show distinct patterns of age-

related development: fluid and crystallized intelligence.

Fluid intelligence is innate, stored strings of numbers and

facts in short-term memory, spatial relations and abstract

reasoning. Crystallized intelligence is the product of

knowledge acquisition and experience. Vocabulary, general

information, conceptual knowledge, judgment and concrete

reasoning are examples of this type of intelligence.

Youth is often thought of as the life stage

characterized by an insatiable desire to acquire information,

and age as the life stage in which we use information wisely.

Catell maintains that wisdom is, in fact, a separate

intellectual function that develops as we grow older. Other

research into adult developmental stages conducted by Sheehy

(1976), and Levinson (1974) reiterate the fact that adulthood

is by no means a static period of being "grown-up."

Adults prefer single-concept, single-theory courses that

focus heavily on the application of the concept to relevant

problems. They need to be able to integrate new ideas with

what they already know if they are to keep--and use--the new

information. Information that conflicts sharply with what is

already in place is integrated slowly. Also, information that

has little "conceptual overlap" with what is already known is

acquired slowly. Programs need to be designed to accept

viewpoints from people in different life stages and with

different value sets. A concept needs to be anchored or

explained from more than one value set and appeal to more

than one developmental life stage.

Adult learners seem to prefer self-directed, self-

designed learning projects seven to one over traditional

classrooms led by a professional. They tend to select more

than one medium for learning--reading and talking to

qualified peers are most frequently cited as good sources for

acquiring information, zemke and Zemke (1981) note that self-

direction does not mean isolation. Studies show that self-

directed projects involve an average of ten other people as

resources, guides, or encouragers. A final implication for

instructors of adult learners is that adults should have a

hand in shaping the curriculum whenever possible.

Instructional strategies that work with adult learners

are based on flexibility and independence. Zemke and Zemke

state that adults can learn well (and much) from dialogue

with respected peers. They respond to facilitative control

from the instructor, such as that employed when small groups

are used. Drawing upon previous knowledge on which to base

and integrate new knowledge helps adult learners perform

well. The learning environment must be physically and

psychologically comfortable. Long lectures, interminable

sitting, and the absence of practical opportunities are

irritating to adult learners.

Small Group Theory

Although small group techniques and cooperative learning

have had a place in education for a number of years,

quantitative research is necessary to determine their

effectiveness versus "traditional" educational methods. In a

1986 metanalysis of over 2000 studies of writing, Hillocks

classified traditional instruction methods into three

processes: individualized, presentational, or natural.

Instruction which uses peer groups was classified as

environmental. The individualized mode of instruction, in

which the teacher-student conference is the primary mode of

instruction, had a mean effect size of .17. The

presentational mode, characterized by lecture and teacher-led

discussion about good writing and teacher-only feedback, had

a mean effect size of .02. The natural process mode utilizes

freewriting about student interests, and feedback from both

peer groups and the teacher had a mean effect size of .19.

The environmental mode which revolves around peer group

activity and student discussion of specific strategies they

encounter in their writing had the largest mean effect size

of .44.

Hillocks' findings were supported by Atkinson (1993).

She conducted a metanalysis of over 700 studies using writing

samples as outcome measures. She reported workshop approaches

as the most effective teaching strategies, with a mean effect

size of .519. The teaching of inquiry skills had a mean

effect size of .452, and computer applications was least

effective, with a mean effect size of .318.

Why, then, do many instructors continue to use the

least-effective presentational mode coupled with an emphasis

on grammar, particularly in the two freshman composition

classes? Such rigidity and heavy-handed teacher authority may

be paralyzing student writers. Mike Rose (1981) supposes that

rigid composing rules might actually restrict fundamentally

competent writers, while T.P. Hogan (1980) notes that student

interest in writing begins to fall off rapidly in upper

elementary grades. Of course, this is when detailed teacher

criticism of writing appears. When looking at written

language acquisition by second-language learners, Krashen

(1984) found that the best approach is not grammatically

sequenced, that is, looking at one aspect of grammar at a

time. He states "when comprehensible input is supplied in

enough variety and quantity, it is hypothesized that

acquirers automatically receive far better exposure to and

practice on those structures they are ready to acquire next,

as compared to a grammatically sequenced approach" (p. 22).

Cooperative and collaborative learning are becoming

accepted modes of instruction in higher education. Hart

(1993) studied collaborative techniques with college freshmen

enrolled in a computer-assisted composition class. He found

that writing proficiency improved significantly in the

experimental group utilizing collaborative strategies, more

so than the group using computer word processor programs to

improve their papers.

Hewlett (1990) studied whether the use of cooperative

learning as an alternative instructional strategy at the

college level had a significant positive effect upon academic

achievement, self-esteem, attitude toward learning, and

liking for the subject. He found no significant main or

interaction effects for the experimental group which was

taught using cooperative learning techniques and the control

group, which did not use cooperative strategies.

Peer Response Groups

Response groups, as noted in Anne Ruggles Gere's Writing

Groups: History, Theory and Implications (1987), have existed

in various forms for over 200 years but have recently gained

new status. Gere and Robert Abbott (1985) list peer

response's effectiveness in terms of improving critical

thinking, organization, and appropriateness of writing;

improving usage; increasing the amount of revision; and

reducing writing apprehension. They also note that teachers

endorse peer response because it develops a better sense of

audience, reduces paper grading, exposes students to a

variety of writing styles, motivates them to revise, and

develops a sense of community. The experience of writing and

revising for less-threatening audiences than the teacher,

learning to discriminate between useful and non-useful

feedback, and learning to use awareness of anticipated

audience responses are additional benefits of peer response

groups cited by Berkenkotter (1984).

There are two major reasons to use peer response groups

in a composition classroom: exposure and response. Exposure,

according to Brooke (1994) is an essential element of a

writer's life. He quotes T.S. Eliot to point out the

importance of exposure in a writer's life: beginning writers

borrow, but mature writers steal. This points out the

importance of exposure in a literate life. Positive exposure

to other people's ideas, purposes and uses of writing can

make writing appealing. "Writers need exposure to other

writers and their writing in order to see what's possible, in

order to widen the range of what they themselves might try"

(p.26). Nancie Atwell, in her notable work In the Middle

(1987) uses the dining room table analogy to explain the

importance of exposure.

Growing writers need to be surrounded with literate
talk. By hearing and taking part in such talk, growing
writers come to recognize that other people are, in
fact, excited by reading and writing, finding values in
certain books and writing projects, and may come to
realize their own interest in such activities, too

Exposure is an essential component for junior high

students' successful writing according to Rief (1992) and for

senior high students according to Romano (1987). Donald

Murray (1990) emphasizes the necessity of exposure for

students in college composition classes.

Brooke notes two ways to bring exposure into the class.

First, the teacher should set aside time for individuals to

share writing and reading with the whole class and with small

groups. Student writers need to hear from each other what

sorts of things they do read and appreciate and what sorts of

things they write. Secondly, when a teacher provides time for

small group discussion of writing, they initiate an ongoing

forum for exposure. Each time a group meets to share drafts

or ideas, the members of the group are functionally exposed

to three or four other people's interests in reading and

writing, and they are likely to hear why the person wants to

write the kinds of pieces she's writing and what sort of

pieces this writing reminds her of. Group meetings provide

ongoing exposure to the literate activities that the group

members value and in which they participate (p. 27-28).

Response is as important as exposure to the beginning

writer. There are three reasons writers need response to

their writing, according to Brooke. First, getting response

to writing brings the writer into a kind of community where

writing is valued. Discussion of the ideas a writer is

wrestling with, no matter how little they seem connected to

the text itself, creates a context where the writer's ideas

have social value and aids the identity negotiation of the

student from that of novice to academic. Secondly, through

listening to the response of others to their writing, writers

learn about the reactions of other people, about the various

ways different minds make sense of the same passages and deal

with the same writing problems. This makes them better able

to predict their readers' reactions while they write and

improves their writing processes as well. Last, response to

particular drafts can often help writers see new

possibilities and problems in their pieces, often leading to

revisions that significantly improve writing.

Brooke further states that writing classes need to

provide these minimums of response: 1. For short term goals

of improving individual text, writing classes need to provide

response from other writers, response that provides direct

suggestions for how and why the students might develop their

texts further. Response should come from several different

writers. 2. For long term goals of helping the writer develop

more strategies for writing and a better sense of how readers

respond to texts, frequent opportunities should be provided

for readers to describe how they read a text--such as in

movies of the mind (Elbow, 1981) so that writers can hear

differences in how minds read. Also, writers need frequent

opportunities to reflect on the response they received, to

identify the kinds of advice they are getting, and to

speculate about what this means for their writing, and to

formulate their own plans for what to do next. 3. For long-

term goals of helping writers see the social value in the

kinds of writing they are doing, classes need to provide

frequent opportunities to discuss the ideas and purposes

underlying the writing.

Beaven (1977) echoes the importance of both exposure and

response to the writer:

Peer evaluation offers each student an opportunity to
observe how his or her writing affects others. Peer
evaluation provides a kind of motivation not available
in other approaches to evaluation. . as trust and
support grow in these small groups, students begin
writing for peers, developing a sense of audience,
becoming aware of their own voices, and using their
voices to produce certain effects in others (150).

The value of peer response groups is both social and

academic. Students involved in Brooke's classes stated that

the social value of response groups was the single most

important precursor to their writing. One student, Peg,

shared that writing became something she wanted to do--once

she found social approval for what she was attempting. Roger,

another student, felt that the groups helped him achieve the

ability to critique his own writing and the writing of

others. Even for Roger, the short-term task of "fixing"

drafts was less important than finding there was social value

to certain kinds of writing because other people also used

writing in the same ways and finding that his ability to

critique writing generally improved through repeated group

discussion. McClelland, as far back as 1953, noted that peer

evaluation strengthened interpersonal skills needed for

collaboration and cooperation as students identified strong

and weak passages and revised ineffective ones, as they set

goals for each other, and as they encouraged risk-taking

behaviors in writing. Peer response groups support the needs

identified in work on achievement and motivation, according

to McClelland. After basic needs are met, the needs for

affiliation, power, and achievement must be satisfied to

produce growth. Achievement is dependent upon fulfillment of

the needs for affiliation and power.

Donald Graves (1984), in working with student writers at

the elementary level, identified this need for affiliation

and power: children learned to write and their writing

continued to develop when they had the opportunity to write

in a supportive environment.

An additional benefit of peer response was noted by

Weiss (1980): "students can learn from assisting one another.

. Peer critiquing has the benefit of improving student

practice without excessive teacher effort--always a key

consideration" (p.143).

Richer (1992) examined the effects of two kinds of

feedback, peer directed and teacher based, on first year

college students' writing proficiency and writing

apprehension. He found that the peer feedback group showed a

significant increase in writing proficiency; there was no

correlation between writing apprehension and writing

proficiency between the two groups. He concluded that using

peer feedback provides a viable method enabling first year

college students to enhance their writing skills and that

more emphasis should be placed on this mode of instruction as

an alternative to teacher feedback and conferencing.

Quantitative research into the effectiveness of peer

response groups is limited. Many research studies dealing

with peer groups and the evaluation of writing indicate that

improvement in theme-writing ability and grammar usage when

small groups of students engage in peer evaluation may equal

or exceed the improvement that occurs under evaluation

procedures carried out by the teacher alone (Ford 1973;

Lagana 1972; Maize 1952; Pierson 1967; Sager 1973; Sutton &

Allen 1964).

Lagana (1972) found that after specific goals for

writing improvements had been established, peer evaluation

was as effective as traditional teacher evaluation over a 15-

week period. He also found that the experimental peer

response group improved more than the control group in

organization, critical thinking, and sentence revision. The

control group, however, registered more improvements in

conventions. Ford (1973) found that college freshmen in the

peer response experimental group showed significantly higher

gains in both grammar and composition ability.

Research on classroom talk and small group work

conducted by Sweigart (1991) showed that small group

discussion was very effective in improving students'

knowledge of the writing task before they began to write.

Students indicated a preference for working in situations

where talk with peers was necessary.

Peer group work and its effects on adult learners was

studied in 1990 by Stephen Fantine. He observed nine

nontraditional students in a freshman-level college writing

course. The students were taught by a variety of

instructional techniques, including teacher modeling,

teacher-student conferencing, and small-group peer review of

drafts in progress. Five students demonstrated little change

in attitude toward writing. They left the course viewing

writing as a difficult and unappealing task. Four students

made positive attitudinal changes toward writing and seemed

committed to revising any writing to be shared with others.

Hillocks (1986) in his metanalysis of composition

research found that students who have been actively involved

in the use of specific criteria and/or questions to judge

texts of their own and of others write compositions of

significantly higher quality than those who have not (p.24).

Beaven (1977) notes that the desired growth in writing seems

to occur when students work with the same group for an

extended period and when there is less structure. Beaven's

study contradicts a number of researchers who have found that

having some structured criteria to apply in peer groups when

evaluating writing results in better papers.

McCarthey (1991) studied the way specific criteria

presented by the teacher work their way into student talk

about writing in a 5th/6th grade classroom. Her subjects

demonstrated that the internalization and use of the criteria

depended a great deal upon the teacher's ability to organize

and present the information, and they used the criteria most

frequently when talking with younger students.

Applying teacher-generated criteria resulted in limited

conversation about deep writing issues in Stamm's (1993)

study. A class of 7th grade students were subjects in her

research to determine what student-student and student-

teacher writing conferences were like and what the

differences were in interactions between female and male

writers and between high-and low-achieving writers. She found

that the structure of talk in the student-student conference

was influenced by the Content Conference Form provided by the

teacher. Responders rarely identified problems for writers or

offered constructive criticism. Female writers talked more,

and helped writers improve their writing more, than male

writers did. High-achieving writers talked much more, and

made many more suggestions to help writers improve their

writing, than low-achieving writers did. However, students

interacted more freely with each other than they did in the

student-teacher conference. In the latter type of conference,

teachers talked much more than the writers, and engaged in

more talk with female and high-achieving writers than with

boys and low-achieving writers.

Vygotsky's (1978) theory of the zone of proximal

development explains some of the reasons for the success of

small groups. He notes that during play, a child behaves

beyond his or her average age and daily behavior. Play allows

the child to explore life in a relatively supportive

environment. Even though learners are ultimately responsible

for their own conceptual development, the student does not

learn in isolation. Other learners can create a supportive

learning environment that encourages students to raise their

perspectives and ask new questions.

Can students be effective judges of each others' texts?

Wallace (1991) raises doubts. In a study of students

developing, implementing, and judging intentions in writing,

he found that students were not good judges of the

effectiveness of their own and other students' texts in

meeting application criteria supplied by the teacher. In two

of his four analyses, the students showed a bias in favor of

their own texts over the texts written by other students.

They also identified problems in other students' texts that

they could not see in their own texts.

Some group work involves students writing

collaboratively. Williams (1993) examined the kinds of talk

and behaviors that occur in group writing situations. He

reported that most eighth graders in collaborative writing

groups wrote effective persuasive essays while engaging in

on-task and off-task talk and behavior. Presumably adult

learners would benefit from the group work, since they might

be more inclined to exhibit on-task behavior.

Adult Learners and Small Group Work

Small group work such as participating in peer response

groups fits the criteria for successful curriculum design for

adult learners. Murray (1985) points out that writing means

self-exposure. No matter how objective the tone or how

detached the subject, the writer is exposed by words on the

page. It is natural for students to feel such exposure. That

fear is best relieved if the writer and fellow students look

together at a piece of writing to see what the piece of

writing is saying. . and if they listen to the piece of

writing with appropriate detachment. He further states that

sharing writing is best done orally. When students read their

papers aloud they hear the voices of their classmates without

the interference of mechanical problems, misspellings, or

poor penmanship. It is perhaps more important for the writer

to hear his or her own voice convey intensity, drive, energy,

and more. This means a piece of writing comes from what it

says and how it says it.

Judy (1980) continues Murray's ideas. He says that the

teacher should not be the only reader of student writing.

Student's comments to one another can be at least as helpful

as the teacher's, while at the same time being less

threatening. People naturally seek out an audience for their

writing. Initial shyness about making writing public should

not be confused with a desire to keep things private. When

students write for real readers--their classmates or people

outside the class--they pay more than usual attention to

matters of form, style, and correctness. Writing for an

audience allows students to see such matters as an aid to

reaching readers rather than simply as a teacher's concern or

obsession. Lauer (1980) agrees. She sees criticism from peers

as providing valuable guidance for revision. Students do

respond as an audience to their classmates' writing. In

laboring to communicate valuable insights in genuine writing

situations, they see the importance of control over


Research by Cullum (1991) seems to support Lauer's

position. The subjects in his study of collaboration in the

writing classroom showed a great improvement in sentence

length, use of passive voice, use of the verb to be, and an

overall number of writing problems.

Murray (1985) explains this group approach as a

community of writing. The community of writers wants to help

the writer help the piece of writing find its own meaning.

A case study of one older adult involved in a college-

level English class was the focus for Mary Phillip's (1991)

research. After observing the subject for 26 weeks, she

suggests that a writer can learn from another writer

regardless of context or of either writer's station of


Elaine Hobby (1989) has conducted much research on adult

women and small group work. Since twice as many adult

students at two-year colleges are women, her work merits a

closer look. Women in her study varied in age from the late

teens to mid sixties, most falling within the thirty to

forty-five age range. Most were examples of the life-changing

events theory discussed earlier. All of them felt they were

at a point of transition in their lives: they had lost or

were bored with their jobs; their children had started school

or left home; they were divorced or approaching retirement.

She found small group work remarkably successful because it

encourages students to talk together. The classroom became a

place where students talked, when for most of these women

education had meant listening to the teacher talk. In small

groups, they realized they were not alone--the room was full

of women who felt that they had either forgotten how to think

or never knew how in the first place. Elements of their lives

they had identified as personal they soon found out were

commonplace to others. They began to recognize that others in

the class could be a source of support and shared analysis,

not competition. Hobby states that because they are women

they were responsive to one another's needs and sensitive to

each other's insecurities; they encouraged each other in

speaking up in class and in writing essays. The few men who

were in her composition class tried to dominate discussion

and were less skilled at cooperation and giving support.

Hobby notes that we live in a society where no one is

considered stupid if they cannot plan in advance what clothes

three children and their father must have clean to wear in a

week--a very complex task performed effortlessly by many

women--but in a society that thinks people are stupid if they

cannot write essays.

Barbara Cambridge (1989) also points out why small group

work is beneficial for women. She says women are essentially

interactional, relational, participatory, and collaborative

as a result of the forces that have shaped their lives.

Writing and small group discussion helps women to bridge the

gap from silence to private discourse to public

pronouncements in class.

Deborah Tannen (1989) states the difference in the way

men and women verbally communicate in terms of public talk

and private talk. Men are more accustomed to being public

purveyors of information. Women, on the other hand, are

private emotion-expressers.

Writing Apprehension

The Development of the Construct

The term "writing apprehension" was coined by Daly and

Miller (1975a) to describe an individual's tendency to

respond favorably or unfavorably toward writing situations in

order to provide empirical evidence on an attitude that

affects writing behavior. Their research was based on

communication apprehension which seriously affects a large

proportion of the population. Prior studies in communication

apprehension have shown that highly apprehensive people tend

to choose occupations they perceive as requiring little

communication (Daly & McCrosky, 1975), tend to be less

inclined to achieve in general (Giffin & Gilham, 1971), and

tend to have lower self concepts than others (McCrosky &

Daly, 1974).

Some people appear unusually fearful or hesitant about

writing and avoid writing situations whenever possible. "In

classroom situations," said Daly and Miller (1975a), "they

will be the individuals who consistently fail to turn in

compositions, who do not attend class when writing is

required, and who seldom enroll voluntarily in courses where

writing is known to be demanded" (p. 244).

The reluctance or resistance highly apprehensive

students show toward writing was the subject of Nancy

Hayward's(1991) research. She states that there is no single

profile of a resistant writer, but many of the resistant

writers in her study exhibited two distinctly different

reactions--either fight or flight. When instructors suggested

that they modify some part of their papers, they reacted with

open defiance and hostility toward the instructor, or with

withdrawal, usually resulting in incomplete or hastily

completed work, lack of revising, absenteeism, or refusal to

interact in the class.

Daly and Miller's Writing Apprehension Measure (1975a)

is an empirically based, standardized self-report instrument

intended to measure an individual's level of writing

apprehension. The original measure consisted of 63 items

which dealt with respondents' perceptions of their anxiety

about writing. It included statements about their beliefs,

likes and dislikes, and attitudes about evaluations by self,

peers, teachers, and professionals. A Likert scale format was

employed using five possible choices from strongly agree to

strongly disagree. All items with factor loadings above .60

were selected to compose the initial instrument, with these

26 items accounting for 46 percent of the total variance

(Daly & Miller, 1975a). The obtained reliability of the

measure was .940, the test-retest reliability for over a week

was .923, and the mean score was 79.28 with a standard

deviation of 18.86 (Daly & Miller, (1975a). Studies lasting

over three months have produced test-retest reliability

coefficients of greater than .80 (Daly, 1985).

The Writing Apprehension Measure

After development of the Writing Apprehension Measure,

Daly attempted to provide evidence for the predictive

validity of the measure by looking at some of the effects of

writing apprehension. He looked at the interrelationship

between writing apprehension and message intensity (Daly &

Miller, 1975b); SAT scores, success expectation, willingness

to take advanced courses, and sex differences (Daly & Miller,

1975c); occupational choice (Daly & Shamo, 1976); message

encoding (Daly, 1977); writing competency (Daly, 1978);

academic decisions (Daly & Shamo, 1978); teacher role

expectancies of the apprehensive writer (Daly, 1979); and

self-esteem and personality (Daly & Wilson, 1980). Daly found

that highly apprehensive individuals wrote less intense

messages, did not expect to succeed in writing courses, and

chose occupations and/or classes that required little or no

writing. Teachers reported better expectancies for high

apprehensive males and low apprehensive females; thus it

seems that teachers expected males not to enjoy writing but

they expected that females liked to write. There was a low

correlation between the Writing Apprehension Measure and the

SAT-verbal test. Students enrolled in advanced composition

classes are usually low apprehensives who are confident

writers and enjoy the writing act.

Writing apprehension has been studied by several

researchers in regards to teachers' levels of apprehension.

Claypool (1980) found a significant negative correlation

between a teacher's level of apprehension and number of

writing assignments made. Highly apprehensive high school

teachers assigned only an average of seven writing

assignments per year as compared to 19.9 assigned by low

apprehensive teachers. One study found a positive

relationship between a teacher's apprehension level and

concern that students use standard English (Gere, Schuessler,

& Abbott, 1984). However, a larger number of researchers have

suggested that a teacher's emphasis on rule rigidity and

"perfectionism" make students fearful of writing and can

actually result in "blocked" writers (Rose, 1980; Newkirk,

1979). Little causal research has been conducted to determine

whether highly apprehensive teachers tend to transfer their

feelings onto students by doing the same kinds of

"conventions" that made them apprehensive in the first place

(Kaywell, 1987).

Writing Apprehension and Writing Achievement

Several studies have indicated that apprehension is

associated with writing performance. In a survey of

elementary and secondary teachers, poor skill development was

the most common reason cited for writing apprehension (Daly,

1979). High apprehensives write compositions with fewer

words, convey less information, use less qualification, use

lower levels of language intensity, and have less command

over usage and written conventions when compared with low

apprehensives (Book, 1976; Daly, 1977; Daly & Miller, 1975c;

Faigley, Daly & Witte, 1981; Garcia, 1977; Reed, Vandett, &

Burton, 1983). Measures written by high apprehensives were

rated significantly lower in quality than those written by

low apprehensives (Book, 1976; Daly; 1977; Daly & Miller,

1975c). Causality has not been proven in any of these

studies: writing apprehension does not necessarily cause poor

writing nor does poor writing cause writing apprehension.

Each probably affects the other.

Results from a study by Marx (1991) confound the issue

even further. In a study of over 200 freshman composition

college students, developmental and advanced students

expressed many of the same attitudes about writing, while the

middle group stated attitudes more expected from students of

lower ability.

In 1978 Daly conducted research on 3,000 undergraduates

enrolled in a basic composition course in order to detect the

actual skill or competency differences that exist between

high and low apprehensive writers. Respondents completed the

Writing Apprehension Measure and a multiple-choice test of

writing skill designed to assess knowledge about mechanics

and grammar. Daly found that high apprehensives did not

perform as well on the test of writing skills as low

apprehensives. In every case, the direction of the means

favored the low apprehensives.

One hundred ten undergraduates were the subjects in

Faigley, Daly, and Witte's 1981 study of the effects of

writing apprehension on writing competency and performance.

Differences again favored the low apprehensives in all but

two measures (paragraph pattern and sentence pattern

subtests). The subjects took the Test of Standard Written

English, the English Composition Test, the SAT verbal test

and vocabulary subtest, and the language mechanics and

paragraph comprehension sections of the McGraw-Hill Reading


In the writing performance portion of the study, highly

apprehensive individuals wrote significantly shorter

narrative-descriptive compositions which were rated as less

syntactically mature or fluent than the compositions of their

low apprehensive counterparts (Faigley, Daly, & Witte, 1981).

Interestingly, there was no effect for apprehension on

argumentative essays. The authors concluded that highly

apprehensive writers have less command over usage and writing

conventions and are unable to develop ideas as well as low


Walker (1992) completed a study to determine whether

audience adaptation activities would affect writing

apprehension and writing achievement. The study was conducted

on over 400 eighth-grade students, who participated in 15

activities designed to increase their awareness of audience

as they wrote, an emphasis also often suggested as a benefit

of peer response groups. Walker found no significant

difference between the experimental and comparison groups in

the amount of change in writing apprehension from the

beginning to the end of the 13-week study. The experimental

group exhibited a significant increase in writing achievement

while the change for the comparison group was not

significant. There was no significant difference in

postachievement for differing initial levels of writing

apprehension, and approximately 15 percent of the eighth

graders were classified as highly apprehensive. The study

suggests that increasing an awareness of audience will

significantly increase writing achievement, but not decrease

writing apprehension.

Changing Writing Apprehension

Based on the logic that a positive attitude about

writing is a desirable characteristic and highly apprehensive

individuals should be helped to lose some of their unhealthy

anxiety about writing, some researchers have focused on the

modification of writing apprehension. Daly (1985) suggests

that modifications usually take one of two forms: examining

the effects of educational programs, such as a particular

writing course, on writing apprehension and identifying and

testing various strategies aimed at alleviating writing


In a landmark study in 1980, Fox investigated the

effects of two methods of writing instruction on writing

apprehension and writing quality. One method of instruction

was set up as a workshop format where there was free writing,

structured peer group activities in response to writing,

language problem-solving exercises, and instructor-student

conferences. The other method of instruction was more

traditional. Students received instructor lectures about

writing, participated in teacher-led question/answer and

discussion periods, did structured writing activities, and

received evaluation by the instructor exclusively. Both

groups showed a significant decrease in writing apprehension.

However, Fox found that the student-centered workshop

approach reduced apprehension significantly more than the

conventional teacher-centered approach. There were no

significant differences in terms of writing quality for

either group.

Pfeifer (1981) studied the effects of peer evaluation

and personality on the writing anxiety and writing ability of

college freshmen. She found no significant effect in regards

to peer evaluation on either writing anxiety or performance,

and she noted that students with identical apprehension

levels did not necessarily produce the same quality of

writing. She attributed this difference in quality to

personality differences. Pfeifer concluded that reducing

writing anxiety did not necessarily improve writing ability.

Thompson (1979) studied freshmen college writers to

determine if her language study approach decreased writing

apprehension and improved writing ability. The approach,

which included discussions of procrastination, standard

English and dialects, the history and formation of language,

and the connection between reading and writing, resulted in

decreased apprehension and increased writing ability.

Thompson suggested that if students discover their own

"personal writing rhythm" then they will be less apprehensive

about writing.

Weiss and Walters (1980) designed formative writing

tasks to answer two questions: "how well am I learning

something, or how well can I express something being

learned?" (pp. 4-5). They sought to find out whether an

increase in the number of traditional writing tasks or an

increase in the number of nontraditional writing tasks in

content courses would also increase apprehension. Formative

writing tasks were assigned in 15 classes, five were used as

control classes. Apprehension levels in 11 of the 15

experimental classes decreased, but not significantly.

Writing apprehension levels decreased more for the

experimental classes than the control classes, but not


As Smith (1984) and Hillocks (1986) point out, little

quantitative research has been conducted on the effectiveness

of specific instruction in reducing writing apprehension.

Teachers should seek to help those students who are highly

apprehensive about writing, although some apprehension is

probably necessary to write an acceptable paper. In a study

of professional adults, Aldrich (1982) found that 49 out of

89 people reported negative feelings about writing. She

concluded that the number of negative responses to her

questionnaire" seem to indicate that dread and apprehension

are probably preventing otherwise competent people from

approaching writing tasks confidently" (p. 300). The goal of

writing teachers should be to decrease writing apprehension

and increase writing ability at the same time.

Free writing, writing whatever comes to mind for five

minutes, was the focus of Sorensen's (1993) study. Working

under the supposition that an increase in ungraded writing

opportunities would bring about a decrease in writing

apprehension and an increase in writing fluency, Sorensen's

subjects wrote freely for five minutes five times per week.

She noted that writing apprehension scores decreased,

composition quality for highly apprehensive subjects

increased, and results were mixed for fluency, supporting two

of her hypotheses.

Causes of Writing Apprehension

Procrastination, inability to organize materials, an

over-adherence to perfectionism, and impatience at the

editing and proofreading stages are several causes of writing

apprehension suggested by Cope (1978). Cope believed that

effective treatment of writing apprehension cannot begin

until the cause has been established. Once the cause of a

student's writing apprehension has been found, teachers

should spend time with their students on writing as process,

time-management, and organizational skills.

In a study netting somewhat unusual results, Powers,

Cook, and Meyer (1979) found that compulsory writing caused

apprehension. Their study included students enrolled in a

basic college level composition course, half under the

impression that the course was compensatory and the other

half believing it was a regular course. Both groups were

given five to six compositions to write. The papers received

typical teacher-only feedback. The students in the

compensatory section had a significant increase in

apprehension; the other group also had an increase but not at

a significant level. These findings were unusual because they

were not consistent with other findings. Fox (1980) conducted

a study in which students were also forced to write, but they

experienced a decrease in apprehension. Smith (1984) noted

that "it seems far more likely that the method of evaluation,

not compulsory writing, was responsible for the increase in

writing apprehension" since other studies with compulsory

writing show decreased apprehension (p. 4).

Teaching Product over Process: Student Attitude Toward

Beginning with Janet Emig's 1971 classic study of the

composing process of twelfth graders, the teaching of writing

began to shift from a product-oriented to a process-oriented

approach. Emig observed the behavior, including attitude, of

four students while they were writing; she also examined how

they wrote rather than what they wrote. She found that

students refrained from showing their real feelings because

of their fear of the teacher's criticism.

Student attitude toward writing became more of a concern

to researchers in the mid 1970s. Metzger (1976) studied a

seventh grader, a tenth grader, and a college student while

writing. These students expressed writing as a joyless chore

and they viewed their teachers as only proofreaders or

editors. Metzger pointed out that classroom experiences with

writing do not encourage students to consider writing as a

pleasant, profitable or purposeful activity. He added that

teachers should provide students with many opportunities for

pleasurable writing experiences. Hogan (1980) also suggested

that teachers should make special efforts to capture

students' interest in writing in grades six through nine,

which may help preserve students' enjoyment of writing.

Shaughnessy (1977) studied the writing problems of

college freshmen. She found that after several false starts

and fear of error, they produced sentences which were

"hopelessly entangled" (p. 7). Shaughnessy advocated focusing

on the process approach to writing instruction. She felt the

product-oriented approach centered too much on surface


Ann Berthoff (1981) proposed that teaching the

peripherals of writing such as grammar, spelling, and

punctuation is not teaching writing. She said that it is a

misconception to think that writing can be factored into

subskills. In addition, she stated that by dwelling on the

correctness of student papers, teachers may unconsciously be

encouraging students to produce trivial papers or to stop

writing altogether.

Holbrook (cited in Emig, 1983) described the effect of

the over-emphasis on correctness:

Children become so terrified of putting down a word
misspelled, particularly an unfamiliar word, that they
don't put down any words. I have seen it happen to a
child of 8, who wrote long marvellous stories. After a
year with a teacher who wrote 'Please be more tidy,'
'Your spelling is awful,' 'Sloppy'--and never a good
word, she stopped altogether. She wrote little lies, a
sentence at a time, in a 'diary.' 'Coming to school
today I saw an elephant.' It wasn't true. But that was
all she was damn-well going to write--neat, complete,
grammatical, well-spelt, short, and essentially
illiterate lies. For her the word had been divorced from
experience. The deeper effect is to make the learning
process one separated from sympathy and creative
collaborative interest in exploring the wonder of being.
(p. 95)

Teachers should notice the value of the expression, of

the message in writing. Failure to notice because of too many

mistakes may have disastrous effects. "If their instructors

continue to emphasize an avoidance of error, then students

will finally learn to avoid as many errors as possible by not

writing at all," Maimon (1979) wrote about college freshmen,

a statement especially true if those freshmen are adult

learners with fragile self-concepts on the line.

Berthoff (1981) reasoned that no student will care very

much about learning usage and mechanics unless he or she

cares about the composition being edited. Emig (1983) added

that there is very little evidence to support that the

persistent pointing out of specific errors in student papers

leads to an elimination of those errors. In addition, no

student will be able to revise well until they can write

comfortably and fluently (Kirby & Liner, 1980), although a

study by Walberg and Ethington (1991) indicates that those

students with low apprehension and high motivation to write

receive more requests and opportunities to write for

teachers. They were also more likely to show their writing to


Student attitude toward writing plays a definite role in

writing growth. No matter how skilled or capable students are

in writing, if they believe they will do poorly or if they

choose to avoid writing whenever possible, their skill

matters little. Daly (1985) claimed that writing desire is

just as important as writing skill:

An individual's attitude about writing is just as basic
to successful writing as are his or her writing skills.
For no matter how skillful the individual may be as a
writer, without a willingness to engage in writing one
can expect little more than the atrophying of composing
skills. A positive attitude about writing is associated
with, and may even be a critical precursor of, the
successful development and maintenance of writing
skills. (p. 44)

In a study of 150 12th-grade students, Lester (1991)

looked at attitude toward usage in writing and concluded that

the motivational level of writers at all learning levels was

dependent upon writer anxiety, confidence in revision

techniques, and interaction with the teacher as a mentor.

Learning to write is a sensitive process, and teachers

should offer lots of praise and encouragement during the

process. Instructors should teach the conventions of writing

after self confidence has been developed. Then there will be

plenty of time to concentrate on the "peripherals" of writing

(Kaywell, 1987).


Although revision seems to be one of the easier aspects

of the composing process to study because it leaves a record,

the truth is that revision is a complex activity. Researchers

have sought to understand this complexity in two ways: by

examining the effects of revision and by speculating on the

causes of revision.

Revision and the Composing Process

One of the benefits of the writing-as-process movement

has been the inclusion of revision as an important part of

the composing process rather than as a proofreading or

editing stage done after the paper is essentially complete, a

view initially suggested by Rohman and Wlecke (1964).

Revision, by literal breakdown of the word, is re-visioning,

or reseeing the composition. Hayes and Flower (1980) point

out that in their model of composing, which consists of

planning, translating, and reviewing, reviewing (or revision)

is the key to improving the quality of the text. Revision is

not a simple process. As Hayes and Flower assert, it can be

recursive. A writer may discover a gap that may lead to a

reader's failure to comprehend a particular point. The writer

then generates ideas and composes whatever is necessary to

fill the gap. This can interrupt the entire "process" of the

process approach to writing, but ultimately leads to a

reader's understanding of a composition closer to the

author's original intentions and meanings, hence, a better


Nancy Somer's (1980) research on the effects of

revision strongly suggests that the linear model of composing

is overly simplistic. Fennick (1991) studied prose fiction

writers and concluded that all her subjects revised as they

drafted and all relied on collaborators at some point in the

process for feedback. For experienced writers as well as for

beginners, revision can be fun or it can be excruciating, but

it is always there, a part of writing, early, late, and


Causes of Revision Decisions

Two of the most important studies on revision in the

past decade have been those done by Flower and Hayes and

Sommers. Linda Flower and John R. Hayes (1980) have attempted

to study the causes of revision by soliciting verbal

protocols that provide a running account of a writer's

conscious activities during composing. Their evidence

suggests that writers move back and forth among the various

activities of composing and that expert writers frequently

review what they have written and make changes in the text

while still writing.

Sommers (1980) demonstrated that writers of different

abilities make different kinds of revisions. She drew

distinctions between the revisions of skilled and unskilled

writers according to the length of their changes and the

types of operation. Sommers used the same categories of

operations--deletion, addition, substitution, and

rearrangement--that Chomsky used to group transformations.

Adolescent writers' revising strategies were the focus

of Andrews-Beck's (1989) research. She found that writers

value overall length, sentence length and complexity, word

choice, and elaboration of content. For those categories,

writers' perceived strategies were usually consistent with

strategies actually used. Perceived strategies were not

consistent with use in two categories: writers said they

reduced error and improved appearance as writer-perceived

quality rose, but errors actually increased and appearance

changed little, suggesting that solitary revising is not

always a reliable way to improve compositions, at least with

younger students.

Revision Classification Systems

Other studies examining the effect of revision changes

on the meaning of a text were problematic because of the

artificial systems of classification employed. The 1977

National Assessment of Educational Progress used broad

categories of meaning such as organizational, stylistic,

continuational, and holistic to classify revisions. Literary

scholars have tried to classify revisions as well. In an

extensive study of well-known authors, Hildick (1965)

described the revision changes of Lawrence, Hardy, and James.

He developed six categories of revision decisions: tidying-up

changes in grammar or punctuation, roughening-up changes such

as loosening speech patterns, changes for accuracy in

expression or argument, sweeping structural alterations,

ideological determined changes, and miscellaneous changes.

Neither the National Assessment system nor the Hildick

classifications are of help in studying revision because most

of the categories overlap (Faigley & Witte, 1981).

One of the most useful classification systems for

studying revision was developed in 1981 by Faigley and witte.

Their Taxonomy of Revision Changes is based on whether new

information is brought to the text or whether old information

is removed in such a way that it cannot be recovered through

drawing inferences. Changes that do not bring new information

to a text or remove old information are called surface

changes. These include formal changes such as spelling and

punctuation and meaning-preserving changes like addition,

deletion, and substitution. Meaning changes are divided into

microstructure changes, changes in which all concepts in a

text are included, even those that can be inferred, and

macrostructure changes, which are major revision changes in

the "gist" of the text.

After applying their taxonomy to groups of expert adult

writers, advanced college students, and inexperienced

students, Faigley and Witte determined that advanced

students, not the expert adults, were the most frequent

revisers. The inexperienced writers' changes were

overwhelmingly surface changes. The advanced students and the

expert adults' changes were more evenly distributed. About 24

percent of the advanced students' changes and 34 percent of

the expert adults' changes were meaning changes. In some

respect, the advanced students revised more like the

inexperienced writers, making surface changes about twice as

often as the expert adults. But in meaning changes, the

advanced students revised more like the expert adults,

especially in macrostructure changes, where the frequencies

of changes by advanced students and expert adults were

similar. The authors conclude that

Successful revision results not from the number of
changes a writer makes but from the degree to which
revision changes bring a text closer to fitting the
demands of the situation. . Revision cannot be
separated from other aspect of composing, especially
during that period in which writers come to grips with
the demands of the particular writing situation. Success
in revision is intimately tied to a writer's planning
and reviewing skills. . Inadequate reviewing of
extant working drafts often results from poor conception
of the audience's needs, which prevents writers from
revising their texts to suit the audience's needs and
wishes. Somehow we must teach our students to distance
themselves from what they have written, to get them to
see it again, then revise. (p. 411)

Peer Response Groups and Revision: A Natural Combination

Willis (1993), in her book Deep Revision, states that

since revision fundamentally consists of trying to look at a

paper with new eyes, it follows that one excellent approach

is to bring in literally new eyes: other readers. She

continues "there is nothing more important than the response

of a good reader in making the plunge into deep revision" (p.


Willis asserts that the writer's peer group is the best

vehicle for thoughtful and meaningful revision of papers, and

that only recently have we come to believe that writing is a

solitary activity.

The solitary artist is a relatively recent model. .
dating from the Romantic movement. Many sculptors, for
example, are dependent upon craftspeople to convert
their models to monumental casts of bronze, and one of
the twentieth century's most characteristic art forms,
film, is a group effort. . Most artists need to seek
out community at some point; whether consciously or not,
we all collect ideas and revise our own work based on
the responses of others. (p. 56)

Peer response groups may be ideally suited to providing

the kind of feedback adult writers need. In a peer group,

Willis states, you not only get responses to your own

writing, you also learn by seeing someone else's work

developing, by hearing about someone else's struggles,

challenges, and plans. "We are wrong to think of writing as

an activity done only in solitude. Much of it is, but always

with the echoes of previously written words in our ears, and

always with at least a vague image of a potential reader"

(p.37). Writer's groups offer support, critiquing, or both.

Other people can offer companionship through the maze of

revision as well as concrete suggestions on how to do it.

Written peer criticism was the topic of Droge's (1991)

research involving college English classes. He found that

students perceived written peer review as helpful, with

benefits that included increased audience awareness and the

creation and fostering of productive revisions and strategies

for revision. Using Bridwell's (1980) coding scheme, he found

that peer reviewers made revision suggestions mostly at the

sentence/length, usually requesting the addition of more

information. The operation suggested most frequently was

substitution/alteration. Student writers listened to their

peer reviewers and revised at a rate of 81 percent at the

level length and in the exact operation suggested in the


Farrell (1977) found that for boys, in particular, peer

group evaluation and revision was more effective than

conferences with a tutor, teacher lecture, or no instruction

situations. Farrell's findings contrast with those of Hobby

and Cambridge stated earlier, where women were found to

perform better than men in peer groups.

In Hillock's (1986) metanalysis of revision research, he

states that for all peer group studies, when operationally

clear objectives were in use the groups were successful. When

the objectives were not so clear, studies incorporating a

combination of teacher and peer feedback have a small but

consistent advantage over those with teacher feedback alone.

Peer response groups also free up a teacher, providing

more time for conferences or planning. Willis (1993) states

that it seems true that much more writing and involvement

with writing takes place in a classroom where students are

editing one another rather than if the teacher is acting as

copy editor for every draft of every paper that gets written.

Copy editing is not the best use of a teacher's time.
Operations of Revision

In Bridwell's 1980 landmark study of the revision

processes of 12th grade students, she found 56 percent of the

revisions at the surface or lexical levels, 18 percent at the

phrase level, and 19.61 percent at the sentence or multi-

sentence levels. She recorded no revisions at the text level-

-indicating that the students took more opportunity to revise

at the lower levels than at the higher sentence or multi-

sentence levels.

Addition, it seems, is the most frequent operation of

revision performed by writers of all ages. Bridwell found

that the second drafts written by the high school seniors

were significantly longer than the first. Sommers (1979)

indicated that addition was the major action of revision

undertaken by adult, experienced writers. And Kamler (1980)

reported that successive drafts of compositions by elementary

school students were longer than the original. This is

presumably because the reexamination of a manuscript prompts

associations or memory searches which result in additions to

the original (Hillocks, 1986).

Sommers (1979) compared the revising strategies of eight

college freshmen and seven experienced adult writers. She

denoted four levels of change: word, phrase, sentence, and

theme, and four operations: deletion, substitution, addition,

and reordering. She found that the greatest number of

revisions by college students came at the word and phrase

level, with deletion and substitution being the most frequent

operations. Adult writers revised most often at the sentence

level and addition was the major operation. Their revisions

distributed over all levels, suggesting that experienced

writers perceive more alternatives than do younger or more

inexperienced writers.

Stiles (1977), however, found that weaker writers are

not totally devoid of concern for higher levels of revision

like changes to content and organization. She studied eight

basic college freshmen seriously deficient in basic writing

skills and reported that while her students were preoccupied

with mechanics, they had a secondary concern for organization

and arrangement.

The importance of criteria to be applied in revision

activities should be of major concern to educators. In

Hillocks (1986) review of revision research he found that

engaging writers actively in the use of criteria applied to

their own or others' writing results in not only more

effective revisions but in superior first drafts. Most

studies showed significant gains for experimental groups

using various types of criteria, suggesting that the criteria

learned act not only as guides for revision but as guides for

generating new material. Cohen and Scardamalia (n.d.) used

criteria called diagnostic statements with sixth graders and

reported significant gains in the quality of revisions made

and in the frequency of revisions, especially in "idea"

revisions. Rosen (1974), Wright (1976) and Kemp (1979) used

various sets of criteria with college students. In their

studies, the experimental groups made greater gains than the

control groups, but the differences were not significant.

Whether students revise differently for disciplines

other than English was the subject of Young's (1992)

research. She focused on the revisions students made to two

essays written for freshmen composition and one paper written

for another discipline to determine what prompted students to

make revisions and how they compared across disciplines.

Results indicated that seven concerns motivated revision in

all disciplines: professor influence, desire for clarity,

influence of perceived audience, polishing of written

product, conformity to perceived discipline standards, and

length requirement. Young concluded that students view

revising in all disciplines in much the same way. Operations

of revision included surface level changes and microstructure

level changes rather than macrostructure level changes which

would affect the meaning of the total text, affirming Faigley

and witte's (1981) findings. This was true of English and

classes in other disciplines.

Operating at the surface level of revision changes is

not just a problem of American students. Tagong (1991)

studied the strategies Thai students employed in the revision

of their English and Thai essays. They made very few changes

in essays of either language. Most of the revisions were made

at the meaning-preserving level, employing three principal

operations: additions, deletions, and substitutions.


One of the major concerns of English educators has been,

for many years, to seek a better way to teach writing to

students. Community college instructors find that the task is

further complicated by the need to reach students of varying

ages, abilities, and apprehension levels.

Research indicates that adult learners are usually

highly motivated and may have fragile self-concepts about

their role as students. Their preference of self-directed,

teacher-facilitated curriculum activities suggests that small

group strategies may be successful in increasing their

knowledge and aiding in identity negotiations as they begin

to perceive themselves as members of an academic community.

Writing apprehension is a problem with people of all

ages, even leading some highly apprehensive people to avoid

writing at all costs, even when selecting a job. Peer

response groups have been used to produce a significant

reduction in writing apprehension for students of all ages,

but especially in college freshmen.

Studies on revision have reported that revision decision
changes differ according to the writer's level of experience

and comfort with writing. Novice writers make frequent

surface level meaning-preserving cosmetic changes, such as

correcting spelling and punctuation. Expert writers revised

words, sentences, and passages at a higher meaning-changing

level. Further research is needed to determine whether


students' participation in weekly peer response groups

changes the frequency, type, and level of revision decisions

they make.


The purpose of this research was to investigate the

effect of peer response groups on community college students'

writing apprehension levels and revision strategies. Special

emphasis was given to students who are over age 25. The

sample was composed of four intact ENC 1101 classes and four

intact ENC 1102 classes, both required freshman level English

classes. All eight classes were taught at Florida community

colleges which had similar student populations. Treatment

classes used peer response groups in addition to teacher

feedback on writing assignments. Nontreatment classes

featured teacher-only feedback on writing assignments. The

following research hypotheses were tested separately for ENC

1101 and ENC 1102 classes:

HI: There will be no significant differences (p < .05)
in mean adjusted posttest scores of nontreatment and
treatment groups on writing apprehension, writing
achievement, and revision measures.

H2: There will be no significant effect (p < .05) of age
on the adjusted posttest scores of nontreatment and
treatment groups on writing apprehension, writing
achievement, and revision measures.

H3: There will be no significant interaction effect
(p < .05) of age and treatment on the adjusted
posttest scores of nontreatment and treatment groups
on writing apprehension, writing achievement,and
revision measures.

The Daly-Miller Writing Apprehension Measure was

administered at the beginning and ending of a 16-week period

and was used to measure initial writing apprehension levels,

and any changes in these levels, of the freshman composition

students. Compositions were written by students on one of two

topics at the beginning and ending of the 16-week period and

were used as pretest and posttest measures for determining

gain or loss in students' achievement in writing. The same

compositions were analyzed using Faigley and Witte's Taxonomy

of Revision Changes (1981) to determine changes in revision

decisions. An analysis of covariance was performed for each

of the dependent variables. A simple bivariate correlation

determined the strength of variable relationships

Design of the Study

Although random sampling is the preferred method of

conducting experimental research, a quasi-experimental design

was selected for this study because of the necessity of using

intact classes. The research design chosen for this study was

the nonequivalent control-group pretest-posttest design

(Buck, Cormier, & Bounds, 1974).

The research design was structured so that (a) two

college freshman composition courses were used: ENC 1101 and

ENC 1102; (b) eight groups were studied; (c) each group was

measured at the same time on three dependent variables before

the treatment was applied to the treatment groups; (d) each

group was measured at the same time after the treatment

was applied to the treatment groups; and (e) the subjects

were not randomly assigned to the eight groups. This design

was followed for ENC 1101 and ENC 1102 classes for a total of

eight intact classes.

Age of subjects and treatment/nontreatment groups were

the two independent variables studied. Three dependent

variables were examined for treatment effects: (a) the Daly-

Miller Writing Apprehension Measure was administered before

and after treatment to assess any changes in apprehension

levels between groups, (b) an expository composition

containing a final copy and previous drafts was written

before and after treatment and holistically scored to assess

any changes in writing proficiency between groups, and (c)

the same compositions were analyzed using Faigley and Witte's

Taxonomy of Revision Changes to determine any treatment

effects on the frequency and depth of revision changes.

The design was chosen to facilitate research conducted

in natural classroom settings in a standard sixteen-week

college semester. Since randomization of subjects was not

possible due to the use of prescheduled subjects and classes,

every effort was made to use groups that were as equivalent

as possible. Pretest, treatment, and posttest procedures were

carried out at the same time, thus controlling for history

and maturation.

At the community college level, eight instructors

participated in the study for a total of eight intact groups.

These teachers were full-time tenured faculty members of

their institutions and taught at least two sections of ENC

1101 or ENC 1102 per semester. By eliminating the need for an

instructor to conduct both nontreatment and treatment

classes, the threat of teacher differences was controlled.

The four community colleges included in this study

represented a balanced sample of community college students

in the state of Florida. Because four of the teachers in the

study taught ENC 1101 and four taught ENC 1102, this study

had representation from both levels of required freshman

composition classes. It is safe to conclude that the findings

in this study could be generalized to populations and courses

similar to those described herein.

The Sample

Eight instructors, with one freshman level composition

course each, were involved in this study. The sample was

composed of eight intact classes; four ENC 1101 and four ENC

1102 classes. One hundred ten students enrolled at Central

Florida Community College, North Florida Junior College, St.

Johns River Community College, and Santa Fe Community College

completed pre- and posttest measures.

The composition of students according to course, age and

sex is presented in Table 3-1 and 3-2.


Composition of Subject Sample by Course and AQe

Course Total Under Over Percent Percent
No. 25 25 Under 25 Over 25

ENC 1101 55 33 22 61 39

ENC 1102 55 39 16 70 30

Total 110 72 38 65 35


Composition of Subject Sample by Course and Sex

Course Total No. No. Percent Percent
No. Males Females Males Females

ENC 1101 55 22 33 40 60

ENC 1102 55 17 38 30 69

Total 110 39 71 35 65

The composition of students according to nontreatment

and treatment groups is presented in Table 3-3.


Composition of Subject Sample by Group

Course Nontreatment Treatment Total

ENC 1101 26 29 67

ENC 1102 23 32 61

Total 49 61 110

Description of the Instruments

The Writing Apprehension Measure

Writing apprehension was assessed by subjects'

completion of the 26-item Daly-Miller Writing Apprehension

Measure (1975a). The statements focus on respondents'

perceptions of their anxiety about writing, their likes and

dislikes about writing, and their responses to self, peer,

teacher, and professional evaluation of their writing. The

measure uses a Likert-type scale with five levels of

responses for each statement ranging from "strongly agree" to

"strongly disagree." Scores are obtained by using the formula

Writing Apprehension = 78 + Positive Scores Negative Scores

(Daly & Miller, 1975a, p. 246). In this study, a low score of

49 indicated a highly apprehensive writer while a high score

of 117 indicated a low apprehensive, or extremely confident,


The Daly-Miller Writing Apprehension Measure is used in

most research involving writing apprehension and has been

found highly reliable across diverse subject samples. The

reliability coefficient of the measure has been close to .94

in research using the instrument (Daly, 1985, p. 45). Test-

retest reliability ranges from .92 after a one-week period

(Daly & Miller, 1975b) to greater than .80 after a three

month period (Daly, 1985) (Lauer & Asher, 1988).

Writing Achievement

Achievement in writing was determined by the holistic

scoring of pretest and posttest compositions using a slightly

modified version of the Diederich Scale (Diederich, 1974).

The scale, based on judgments of the writing of college

freshmen, was developed by the researchers at the Educational

Testing Service (ETS) and is regularly used to evaluate the

quality of compositions. Diederich's factor-analysis study of

teacher's reasons for their judgments when grading

compositions is responsible for the validity of the scale

(Myers, 1980).

Holistic scoring assumes that the whole of a piece of

writing is greater than any of its parts, and no aspect of

writing skill can be judged independently. English teachers

may not agree on the quality of particular traits, but they

can rate papers in much the same way when judging a paper as

a whole. Holistic scoring, according to Myers (1980), is

still the best way to assess writing. The College Level

Academic Skills Test, the State of Florida Writing Assessment

Test, and Advanced Placement Tests use holistic scoring as

their means of assessment, as do most other formal

evaluations of writing proficiency.

A modified version of the Diederich scale is presented

below and was used in this study. On the original scale,

Ideas and Organization received twice the number of points as

the other items based on the emphases of the schools at which

the instrument was developed. In this study, all items

received equal weight due to the possibility of different

emphases in each college's composition program. Handwriting

was removed from the original scale for the purposes of this


Low Middle High

General Merit
Ideas 1 2 3 4 5
Organization 1 2 3 4 5
Diction 1 2 3 4 5
Style 1 2 3 4 5 ____

Usage 1 2 3 4 5
Punctuation 1 2 3 4 5
Spelling 1 2 3 4 5

If a student received the lowest possible score on

everything, the total would be 7; the highest possible score

would be 35.

The Taxonomy of Revision Changes

One of the most useful classification systems for

studying revision was developed in 1981 by Faigley and Witte.

The Taxonomy of Revision Changes they devised is based on

whether new information is brought to the text or whether old

information is removed in such a way that it cannot be

recovered through drawing inferences. Changes that do not

bring new information to a text or remove old information are

called surface changes. These include formal changes such as

spelling and punctuation and meaning-preserving changes like

addition, deletion, and substitution. Meaning changes are

divided into microstructure changes, changes in which all

concepts in a text are included, even those that can be

inferred, and macrostructure changes, which are major

revision changes in the "gist" of the text.

Faigley and Witte tested their system until they were

satisfied that it could be applied reliably, with two

researchers reaching over 90 percent agreement on types of

revisions. They then assigned a numerical code to each

category to facilitate analysis of revisions. When a revision

change spanned more than one sentence, each sentence was

analyzed separately. The resulting frequencies of categorical

revision changes per 1000 words in final drafts and total

revision changes per 1000 words in final drafts was reported.

Data Collection Procedure

Permission for this study was obtained from the Human

Subjects Committee of the University of Florida's

Institutional Review Board. Eight composition instructors

participated in the study: one from Central Florida Community

College, one from North Florida Junior College, five from St.

Johns River Community College, and one from Santa Fe

Community College. All teachers participated in a brief

training session at which time the "teacher instructions"

were explained (see Appendix C).

Teachers administered the writing apprehension protests

during the second week of the semester. During the pretest

period, students were instructed to write a composition on

one of two prompts: Advantage/Disadvantage of Belonging to a

Group in High School, or the Pressures of Being an

Only/Oldest/Middle/Youngest Child. Students were given one

week to complete the essays, and turned in their rough drafts

and final copies to their instructor. The posttest activity

followed the same procedure during the next-to-the-last week

of the semester with only a change in composition topic.

Groups which initially wrote on topic one wrote on topic two

and vice versa.

Classes designated as treatment participated in 30-

minute peer response groups during class once a week. The

study lasted for 16 weeks. Groups read their papers aloud and

then discussed their writing assignments following the

guidelines Brooke (1994) proposed for small group discussions

about writing. An effort was made to mix high and low

apprehensive writers in the groups, as well as mixing

students under and over the age of 25.

The Rating Procedure

Compositions were coded by the subjects based on the

last four digits of a phone number. The investigator compiled

an identification list which identified each student by

pretest, posttest, and composition codes. Those students who

took the pretest but no posttest and those who took the

posttest but no pretest were eliminated from this study.

Compositions were read and scored by four professional

holistic scorers. Papers were divided into four folders and

randomly shuffled and read so that the readers were unaware

of whether the papers are entry or exit level, as Hake (1986)

described the phenomenon of exit papers receiving lower

scores when raters knew the category.

Each scorer read 220 papers in two Saturday sessions,

one week apart. The four scorers were trained together

through the university of Florida, had worked together for

several years, had participated on state-funded holistic

scoring events such as the State of Florida Writing

Assessment Test (WAT) and the College Level Academic Skills

Test (CLAST), had been teachers, and had their master's


The four scorers were provided with instructions

defining each point on the modified version of the Diederich

Rating Scale used in the study. Raters were trained together

and read together in the same room under common direction,

which, according to Myers (1980), is the most reliable method

of scoring writing samples. The investigator chose high,

middle, and low model papers for each category. These papers

were duplicated and served as practice papers for the

readers. After each group of practice papers had been rated,

they were compared. A head scorer was appointed to monitor

and resolve scores on a paper that varied widely in the

categories. Training was conducted until the average

difference between ratings was less than two points.

Each scorer read each composition and scored it on the

scales; thus, each paper was read at least four times.

Scorers marked the scale only; no marks were made on the

papers. After the paper was finished, the scorer put each

sheet on the bottom of the stack until all papers had been

rated. Once a stack of papers had been read and scored,

readers passed the folder to the reader on the right. After

the complete scoring of a folder, the investigator summed

totals and notified the head scorer of any discrepancies in

scoring. The head scorer made the final decision regarding

extreme scores. Scorers took breaks and were supplied meals

and snacks in order to control for rater fatigue.

Data Analysis

The independent variables for this study were course and

age. The three dependent variables were (a) level of writing

apprehension, (b) writing achievement, and (c) depth and

frequency of revision.

Two nontreatment and two treatment groups were used for

both ENC 1101 and ENC 1102 classes. An analysis of covariance

was performed on the adjusted posttest means of the two

treatment and two nontreatment groups based upon pretest

scores. A simple bivariate correlational analysis by course

and group was conducted to determine the strength of

relationships among the variables. The .05 level of

significance was selected for all analyses of data.


This study was designed to test the effect of peer

response groups on community college students' levels of

writing apprehension, writing achievement, and revision

decisions, with special emphasis on students who were over

age 25. This chapter contains the statistical treatment of

the data and findings relative to (a) the effect of peer

response groups on writing apprehension, (b) the effect of

peer response groups on writing achievement, and (c) the

effect of peer response groups on students' revision

decisions at the community college level.

The Statistical Treatment of the Data


The study was limited to eight freshman level community

college composition classes: four ENC 1101 classes and four

ENC 1102 classes. For each course, there were two

nontreatment groups and two treatment groups. Students in the

treatment groups participated in peer response groups for

thirty minutes per week. Pre- and posttests measuring writing

apprehension, writing achievement, and revision decisions

were collected from the students in all eight groups.

Table results are reported from reduced models because

no instructor interactions occurred upon first analysis. The

test for pretest by group was also nonsignificant, thus the

slopes of the regression lines for pre- on posttest were

assumed homogenous for the treatment groups, allowing

interpretation of the F statistic for treatment effect.

Writing Apprehension

One of the major purposes of this study was to determine

the effect of peer response groups on students' levels of

writing apprehension. To determine if an effect existed,

students completed the Daly-Miller Writing Apprehension

Measure at the beginning and ending of a sixteen-week period.

These tests were used as pre- and posttest measures for

determining gain or loss in students' apprehension in


The analysis of data involved two major steps. First,

apprehension scores were calculated for each student. Second,

an analysis of covariance (ANCOVA) was performed to determine

the following: (a) if a significant difference existed among

the adjusted mean apprehension scores of subjects in the two

groups, (b) if a significant effect of age existed on the

adjusted mean posttest apprehension scores of subjects, and

(c) if a significant interaction effect of age and treatment

existed on adjusted posttest apprehension scores of subjects.

Hypotheses for the study were stated in null form, and

tested separately for ENC 1101 and ENC 1102 courses. A

summary of the findings follows each hypothesis.

Analysis of Apprehension Data

A computation of total apprehension scores and the

difference for each student was performed. In ENC 1101, the

mean apprehension pretest score is 77.73 and the standard

deviation is 18.81. The mean apprehension posttest score is

82.31 and the standard deviation is 23.57. For ENC 1102, the

mean pretest apprehension score is 85.65 and the standard

deviation is 19.52. The mean apprehension posttest score for

ENC 1102 is 83.01 and the standard deviation is 22.31. A

comparison of pre- and postapprehension means by course and

group is shown in Table 4-1.

The trend of scores reveals that in ENC 1101, students'

apprehension can be slightly reduced using the method

investigated. In ENC 1102, however, apprehension increased

slightly after administration of the treatment.

To determine the significance of gain or loss in

students' apprehension in writing, various statistical

procedures from the Statistical Analysis System (SAS) were

performed to test the null hypotheses.

Hypothesis 1. There will be no significant differences

(p < .05) among the posttest writing apprehension scores of

the nontreatment and treatment students.

Hypothesis 2. There will be no significant effect

(p < .05) of age on the adjusted posttest writing

apprehension scores of the nontreatment and treatment


Pre- and Postapprehension Means by Course and Group

Course N Mean SD

ENC 1101
Nontreatment Preapprehension 26 78.692 22.444
Nontreatment Postapprehension 26 78.769 28.919
Treatment Preapprehension 29 76.068 15.151
Treatment Postapprehension 29 85.862 17.326

ENC 1102
Nontreatment Preapprehension 23 85.956 19.065
Nontreatment Postapprehension 23 82.565 20.659
Treatment Preapprehension 32 85.343 20.147
Treatment Postapprehension 32 83.468 23.750

Hypothesis 3. There will be no significant interaction

effect (p < .05) of age and treatment on the adjusted

posttest writing apprehension scores of the nontreatment and

treatment students.

The ANCOVA design was used to control statistically any

initial differences present in the students which might

confound differences among the four groups per course. The

covariate was the score on the writing apprehension pretest.

The dependent measure, the score on the Daly-Miller Writing

Apprehension Measure posttest, was adjusted on the basis of

the covariate. After checking all possible two-way

interactions, the model was reduced to include significant

variables. The results of this analysis are given in Table 4-


Summary of ANCOVA Postapprehension

Source SS df MS F PR>F

ENC 1101
Pretest 3939.078 1 3939.078 9.02 0.069*
Treatment 1707.543 1 1707.543 8.97 0.006*

ENC 1102

Pretest 6533.489 1 6533.489 48.88 0.001*
Treatment 8888.158 1 8888.158 34.01 0.001*

R < .05

In Table 4-2, the calculated F ratio for nontreatment

groups in the ENC 1101 classes is 9.02, which is significant

at the .05 level of probability. The model accounts for 23%

of the variance. The calculated F ratio for treatment

groups in ENC 1101 classes is 8.97, which is significant at

the .05 level of probability. The model accounts for 26% of

the variance. For ENC 1102 nontreatment classes, the

calculated F ratio is 48.88, which is significant at the

.05 level of probability. The model accounts for 20% of the

variance. The calculated F ratio for ENC 1102 treatment

classes is 34.01, which is significant at the .05 level of

probability. The model accounts for 29% of the variance. The

posttest writing apprehension scores reject the hypothesis

that there is no significant difference in the mean

performance of subjects among the groups.

A simple bivariate correlational analysis was performed

to detect the strength of the age/writing apprehension

relationship. For ENC 1101 treatment groups, age correlated

with writing apprehension at a moderate 0.456. Thus, for this

group, the older the student, the higher the degree of

writing apprehension. Results of the correlational analysis

are presented in Table 4-3.

Summary of Correlational Analysis for Age and Writing

Group Preapprehension Postapprehension

ENC 1101

Nontreatment -0.009 -0.123
Treatment -0.272 0.456*

ENC 1102

Nontreatment 0.070 0.139
Treatment 0.144 0.122

* P < .05

In short, although the posttest writing apprehension

scores reject the null hypothesis that there is no

significant difference in the mean performance of subjects

among the groups, age appears to correlate directly only with

ENC 1101 treatment subjects' scores.

Writing Achievement

In addition to investigating the effect of peer response

groups on students' writing apprehension, the purpose of this

study was to investigate the effect of peer response groups

on students' writing achievement. Two papers, one at the

beginning and one at the end of a sixteen-week period, were

written by students on assigned topics to determine if an

effect existed. Students' papers were used as pre- and

posttest measures for determining gain or loss in students'

proficiency in writing. Papers were read by four readers

using a modified version of the Diederich Essay Rating Scale

(see Appendix B) whereby paragraphs were evaluated on seven

items in two main areas: general merit and mechanics. Pre-

and posttest papers were collected from the students in all

eight classes.

The analysis of the data involved three major steps.

First, a comparison of the ratings of Rater A, Rater B, Rater

C, and Rater D was made to determine the degree of

relationship among the four ratings. Cautions were taken to

prevent raters from knowing which papers were protests and

which were posttests. Second, a computation of total scores

and the difference for each student was performed to identify

growth scores of each student. Third, an ANCOVA was performed

to determine the following: (a) if a significant difference

in writing achievement existed among the adjusted mean

posttest scores of treatment and nontreatment students, (b)

if a significant effect of age existed on the adjusted

posttest scores of subjects in writing achievement, and (c)

if a significant interaction effect of age and treatment

existed on adjusted posttest writing achievement scores of


Hypotheses for the study are stated in null form.

Following the results of the comparison of raters, a summary

of the findings follows each hypothesis.

Comparison of Raters

All 220 papers, 110 pretest papers and 110 posttest

papers were used in comparing the ratings of Rater A, Rater

B, Rater C, and Rater D. A correlation of total scores on

pre- and posttests and on each section of the Diederich Essay

Rating Scale--general merit and mechanics--using the Pearson

Product r was performed to determine rater reliability. Total

paper pretest scores ranged from 17 to 35 for Rater A, from

17 to 33 for Rater B, from 13 to 34 for Rater C, and from 12

to 34 for Rater D. Total paper posttest score ranged from 18

to 35 for Rater A, from 16 to 34 for Rater B, from 13 to 35

for Rater C, and from 12 to 34 for Rater D. Scoring

reliabilities were quite high (r = +.68 to +.82) for summed


Analysis of Composition Data

A computation of total composition scores and the

difference for each student was performed. Appendix J shows

for each 1101 and 1102 student the total composition pre- and

posttest scores and growth scores. A comparison of pre- and