Both research and CAT results suggest that boys perform relatively better than girls in the examinations of both the multiple-choice, short-answer and long-answer varieties (Cox, 2004, p38). This is concerning as 66% of the final grade for every final year Mathematics course comes from the examinations. The distribution of results for 2004 in examinations also suggests another problem similar to the one Elwood noted in the UK. The mark for coursework is more 'bunched' than examinations, which tend to have a greater spread. Since the VCE ranks students, the ranking is more likely to be effected by the exam mark - despite the intended weighting of the examination (Elwood, 2005, p 384).
Even within the design of an examination, the type of questions used can have an impact on relative gender performance. It has also been argued that 'the performance of girls in science examinations can be enhanced or disadvantaged according to the way the examination and its parts are constructed'. (Cox, 2004 quoting Whithouse & Sullivan). There are significant gender differences when, in multiple choice examinations, there are penalties for incorrect answers. This could be explained by risk-taking being a more 'masculine' trait. Similarly, there is a large gap between male performance and female when open questions are used as a major part of assessment compared to prove/disprove type long answer questions. (Anderson, 2002)
Girls tended to perform better in the internally assessed coursework; a task which focused on communicating the process and mathematics used in an investigation. 'The traditional assessment tasks appeared to favor males. More innovative but still demanding assessment task with a focus on the solution process as well as the answer, which required sustained and independent efforts over a longer period of time, and which had a stronger verbal component seemed to favor females.' (Leder, 2002) This finding is echoed by Cox (2004) and is consistent with comparatively better GAT score that females achieve in Written Communication.