This paper reports on a study into task difficulty in the IELTS Academic writing Task 1. The study examined firstly, the extent to which the difficulty of the task is affected by the amount of information provided to the candidate and secondly, the extent to which the difficulty of the task is affected by the presentation of the information to the candidate.
In the Academic Writing Task 1 candidates are required to examine a diagram or table, and to present the information in their own words (IELTS 2000). Four tasks, which differed in terms of the amount of information the candidates were required to process to complete the task, were developed for the study. Two of the tasks included less information on which candidates could base their responses and the other two included more information. Within each of these two types of tasks, one was designated as the control, and the other was designated as the experimental task. Five different versions of each of the two experimental tasks were developed. These versions differed in the way the stimulus material was presented to candidates. The control tasks were designed as benchmark tasks and administered to all candidates. The experimental tasks were administered to selected subgroups of the cohort.
Two hundred and ten students, who were enrolled in English for Academic Purposes (EAP) courses in Melbourne or Sydney, completed four of the writing tasks (the two control tasks and two other experimental tasks). All scripts were double rated by trained and qualified IELTS raters. Analyses of the test scores and the scripts themselves were then undertaken.
The test score analyses indicated that there were no substantial differences in difficulty between the tasks, either in terms of the amount of information presented or in terms of the differences in presentation of the tasks. Analyses of the written texts produced by the students focused on whether there were any systematic differences in their written performances across different proficiency levels (high, middle and low). Responses from all three proficiency groups to the task with less information showed greater complexity overall than the task with more information .. The trend was less clear overall in relation to accuracy. However, the high proficiency group showed a strong tendency to display greater accuracy in response to the task with more information. It appears, therefore, that tasks providing less information actually elicit more complex language. Since the goal of these tasks is to produce as high a performance from the candidate as possible it can be concluded that this is best achieved through using simpler tasks.