This research investigated the cognitive processes of candidates taking IELTS Academic Writing Task One (AWT1) with different graphic prompts at two different time points – before short training on how to do AWT1 tasks, and post-training. It explored the extent to which candidates’ cognitive processes are affected by the use of different graphs, their graphic skills and English writing abilities, and the short training. A grounded and multi-layered case study approach was employed to collect data on candidates’ cognitive processes. 24 intending IELTS candidates from a large Chinese university completed eight AWT1 tasks while thinking aloud their processes of doing the tasks (four before training and four after training) under examination conditions. Samples of their English writing abilities and graphicacy were also collected, as well as post-task interviews with all participants.
The think-aloud protocols were analysed to identify the common patterns of cognitive processes. A model of cognitive processes was developed, consisting of three interrelated stages – comprehending non-graphically presented task instructions, comprehending graphic information and re-producing graph comprehension in written discourse in English as a foreign language. This model guided our analyses to address the four research questions: (1) How the participants processed the graphic information and how they followed the graphic conventions to re-produce their graph comprehension in written discourse in English were affected by the types of graphs they read. Such effects of different graphic prompts on the cognitive processes were clearly evidenced in the mean scores of the writings, in the use of vocabulary, and in whether and how they would make comparisons or trend assessments, following the graphic conventions in presentation, interpretation and re-production. (2) Although graph familiarity did not seem to affect task performance in terms of the marks of the writings, the participants clearly expressed some potential psychological impact of graph familiarity on their task performance. (3) There is a strong correlation between AWT1 writing performance and writing ability as measured via topic-based argumentative essays. (4) The influence of the special training was strong, in particular, in terms of whether or not the participants tried to make interpretations, predictions and comments by linking the graphic information with their domain knowledge about the graphs.
The implications of these findings are discussed with reference to AWT1 task design, as well as other language test tasks that use graphs as prompts, particularly for listening, speaking and writing assessments.