Carnegie Mellon University

Reading for Revision: A Comparison of Teachers' and Students' Judgments

Author: Ann Chenoweth
Degree: Ph.D. in Rhetoric, Carnegie Mellon University, 1997

This dissertation investigates differences in the ways that experts and novices read to evaluate the same texts to determine what problems, if any, need to be fixed and to determine the effect of making a particular change on the text. Three basic questions are addressed in this study: When students and teachers read the same text... (1) do they agree that any change to it is necessary? (2)do they agree on what kinds of changes would improve it the most? (3) do they agree on what effect making a particular change has?

The primary data come from the responses of twenty-seven writing teachers and 1232 first-year university students. Respondents were asked to complete two tasks: one task asked respondents to locate problems in four short texts and the other asked them to rate the effectiveness of various changes to a text. The data were first analyzed to assess significant differences between the response patterns of the teachers and all of the students, and then the responses of the students were further analyzed to determine if there were significant differences in the response patterns of different ethnolinguistic groups. Statistically significant differences emerged at all levels of analysis. Teachers' responses reflected an awareness of discourse-level features of the text; students' responses tended to reflect local-level concerns. Non- native speakers of English were better at locating text problems than the native speakers of English, but were not as good at judging the effectiveness of changes. These results suggest that teachers and students differ in how they value various text features and how sensitive they are to the rhetorical purposes served by those text features. In many cases it is these differences in values and sensitivity that make it difficult for students to produce texts that their teachers like. Additionally these results underscore the need for more studies that investigate inter- relationships between writing and reading in both L$sb1$ and L$sb2.$