Unravelling joint attention in foreign language writers’ collaborative revision process

By collaborating during writing, writers can pool resources and may benefit from the joint regulation of the writing process. The extent to which group members manage (individual and) joint attention may be a strong predictor of effective collaboration (Schneider & Pea, 213). Joint attention is the tendency of collaboration partners to consciously focus on and/or monitor each other’s attention to a common reference (Tomasello, 1995). However, little is known about when joint attention occurs during collaborative writing and how writers draw each other’s attention to points of interest during the writing process. To shed lights on how and when this may occur, we conducted a small-scale, mixed-method hypothesis-generating study, using a combination of keystroke logging, eye-tracking and audio- and screencapturing. In addition to generating hypotheses, the design also enabled us to explore the feasibility of different methodologies to capture joint attention in an experimental set- up. Ten foreign-language writers participated in an experimental setting, where they were assigned to five randomly composed dyads who revised a familiar and a novel persuasive text. The texts contained both lower- and higher-order textual problems. The revision partners were seated in different rooms but collaborated in the same online environment via audio and screen, a set-up similar that of working in, e.g., Zoom . We collected eyetracking- and keystroke-logging data of each dyad member as well as audio- and screen-recordings of the collaborative writing sessions. In the data-streams (i.e., written, spoken and visual), we identified moments of joint attention and overlap in joint attention. Analyses show that joint attention in gaze was higher in the familiar condition. Joint attention in speech was achieved when dyad partners were jointly coordinating the revision process and/or discussing textual problems. Mapping instances of high interaction (speech) onto instances of high joint visual attention showed that moments of co-occurrence involved focusing on and highlighting lower-order textual problems and reading words and phrases. These results also depend on task type and phase of the writing process. We tentatively explore the transfer of these results to the revision product and discuss their relevance for the study of (joint attention in) collaborative writing.