Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)



First Advisor

Sarah Liggett


Writers unable to improve their own work benefit when an outside reader suggests revision strategies, but what constraints shape the advice students give? This research presents a descriptive/naturalistic study of how students in one composition class viewed peer evaluation. The research explores how students' expectations of the procedure shape their comments and presents a system to code comments for revision/praise content. It considers nine editors' responses to the same essays. The Flower et al. three-part self-revision categories were modified and expanded to code editors' comments: those that referenced an essay's basic features ("thesis") were coded revision level (RL) 1; comments that offered a diagnosis of the problem ("Introduction needs to be clearer") were coded revision level (RL) 2; and advice that offered specific revision strategies (" ... separate into general sentences instead of throwing all ... facts into one") were coded revision level (RL) 3. The current study also presents a three-tier taxonomy of praise which coded interchangeable, ambiguous peer feedback ("good!") as praise level (PL) 1; comments that targeted an aspect of the essay ("introduction was good,") as praise level (PL) 2; while comments that referenced a specific element of the essay (" ... good point when she asks the question 'Is worrying about grades really worth it?"') were coded praise level (PL) 3. Of 525 revision comments, 217 were coded RL 1; 177 RL 2; and 131 RL 3. Students wrote 117 PL comments: 49 PL 1; 55 PL 2; and 13 PL 3. Forty-one percent (266 of 642) of the comments were written at the lowest comment level (RL or PL 1). Editors' awareness of evaluation's social aspects controlled the length, content, and tone of their comments. End-of-semester questionnaires revealed that writers' suspicions were aroused by brief comments, even if the comments were positive; this indicates that writers needed to be convinced of the editors' sincerity/competence before the comments could be accepted. Editors tempered their comments according to their perception of the social ramifications of evaluation. The results highlight the importance of praise and the effect evaluation's social dimension exerts upon peer evaluators.