An investigation of the efficacy of face-to-face versus synchronous chat in the generation and development of written drafts in the EAP class

Abstract

The thesis is a study of the early stages of the writing cycle in an English for Academic Purposes (EAP) class at the University of Singapore. The study focuses on a group brainstorming activity prior to the stage of writing the first draft and on the impact of this activity on the students’ first individual drafts. In addition, the study compares two different modes of discussion: face-to-face and online synchronous chat. The comparison is concerned with the interactional patterns of the discussion in the two modes, and with the transfer of content from the discussion to the first written drafts. The use of group brainstorming at the pre-writing stage is a familiar activity in the writing class but researchers have not yet paid much attention to the way in which the ideas generated in the brainstorming activity are transferred to individual written drafts. It is this gap that this dissertation seeks to fill. A question of particular interest is the extent to which knowledge construction in the composition class is accomplished by the individual or by the group.

Data were collected from four classes of first-year undergraduate students of Science taught by the researcher. The control group, with 31 members, carried out their brainstorming activity in face-to-face mode, while the experimental group, with 27 members, carried out their brainstorming using a synchronous chat facility. The primary data were the chat scripts, face-to-face transcripts and first writing drafts. Analysis of the discussion data was carried out using a genre-based content analysis model deriving from speech act theory (Sinclair and Coulthard, 1975), rhetorical structure theory (Mann & Thompson 1986, 1988), and more recent work on collaborative writing by Plowman (1993), think-aloud protocols (Smagorinsky, 1991) and collaborative computer-based communication by Garrison and Anderson (2003), among others. The model is used to count the frequency of different ideas according to their rhetorical characteristics in the two conditions (face-to-face and synchronous chat), and to determine whether the ideas were generated by individuals or through group discussion. The analysis then looks at the extent to which the ideas were subsequently reproduced in individual drafts. The frequency analyses are complemented by detailed qualitative analysis of the discussion transcripts and the essays of four students, two from each discussion mode.

The results of the analysis suggest that collaborative brainstorming is productive in helping students with the generation and development of ideas for their writing. The findings also suggest that there is a strong link between ownership and use of ideas initiated in the discussion. This tendency is stronger in the chat group than in the face-to-face group. Analysis of the discussion transcripts suggests that this difference is a result of more ideas being initiated in the chat group. In addition to these group differences, the analysis shows that discussion in both modes is characterised by a tendency to seek consensus, with very little argument and negotiation of content. The implications of these findings for the use of group discussion in the writing class are discussed.