Search

The transition from paper-based to computer-based assessment raises a number of important issues about how mode might affect children's performance and question answering strategies. In this project 104 eleven-year-olds were given two sets of matched mathematics questions, one set on-line and the other on paper. Facility values were analyzed to explore the impact of the mode on performance. Errors were coded and this allowed further investigation of the differences between questions in the different modes. The study also investigated children's affective responses to working on computer, attempting to gain an insight into the effect of motivational factors. Tis was made possible by observing and interviewing a sub-sample of children. Findings suggested that although there were no statistically significant differences between overall performances on paper and computer, there were enough differences at the individual question-level to warrant further investigation. Close analysis of the data suggests that it is possible that the question type, the way it is asked, and the numbers involved, might interact with mode to affect students' willingness to show working methods. The findings also suggest that certain types of questions in certain domains might have different impacts according to mode. The study concludes that there is scope for more research to probe further any links that may exist between children's thinking, behavior and assessment mode in order to satisfy concerns about the relative reliability and validity of computer-based and paper-based testing. (Contains 2 endnotes, 4 tables, and 4 figures.)

Tests, especially high-stakes ones, occupy a prominent role in education and impact on students' personal and professional trajectory. Therefore, it is crucial that they are well understood. Enhancing understanding of the workings of tests requires a multidisciplinary approach, one that treats tests not only as assessment instruments, but also as sociolinguistic and sociopolitical products. Using data collected by means of non-participant observation, stimulated recall and one-to-one interviews with writers of GCSE and IGCSE tests, this paper attempts to provide a macro-analysis of the process of test writing. It does so by identifying the various actors and discourses that directly or indirectly take part in test writing. It uses Bakhtin's framework of voice as a tool to facilitate this. The paper concludes by proposing a macro model of test writing, one that explains the macro process of test writing by capturing the main voices involved in it and their interrelationships.

The focus of this research is on the practice of education examination question writing. Educational examinations are tests that are taken by candidates in schools or colleges but that are externally developed, administered and marked by an assessment organisation. Whilst the practice of writing external examination questions is ubiquitous, previous research in this area is limited and has tended to adopt a cognitivist approach. This current project sought to extend this area of enquiry by looking at educational examination question writing as a socio-cognitive process. The project involved a group of seven examination question writers (known as "setters") from a range of subject areas. To capture evidence of question writing practices, each setter was observed remotely using video technology whilst they carried out a question writing task. The video material was then used to facilitate a stimulated recall session where individual setters could explain the processes that motivated their observed behaviours. Analyses suggested that the setters shared a common model of the writing process, which comprised three basic but interconnected phases: thinking about writing; writing and reflective thinking; and reviewing. In addition, question writing practice was influenced by the social system of examining through the way in which the setters considered a variety of perspectives during the writing task. This insight gives an indication of how setter practice develops; implicating a sociocultural perspective which suggests that the broader social context of examination question writing is an inevitable influence on setter practice.