Abstract
Courses taught in Applied Statistics, such as regression or multivariate analysis, tend to have the examination component based on a final written paper, either with computer output attached for interpretation or with summary statistics given so the calculator can be used to evaluate test statistics and hence make inference, or a mixture of both. Assignments may be based on students using a statistical package to do the analysis. The authors found this approach unsatisfactory. The first author trialled at the University of Canberra in 2001 both a mid semester computer based
examination and final computer based examination with the students being allowed to use a variety of statistical packages. Student feedback was so favourable that she taught the course the following year in the same manner as well as a regression course. When she moved to Macquarie University she split the final examination for the third year regression course she was in charge of in 2004, 2005 and 2006 into a computer laboratory examination and a separate written paper. We have been unable to find more than a small number of other examples of this approach, which seems to
have considerable promise as a way of implementing authentic assessment in applied courses. As well as case studies, issues associated with setting, running and marking such examinations are discussed.
Original language | English |
---|---|
Title of host publication | Symposium proceedings |
Subtitle of host publication | assessment in science teaching and learning : September 28, 2006 : the University of Sydney |
Place of Publication | Sydney |
Publisher | UniServe Science |
Pages | 101-106 |
Number of pages | 6 |
ISBN (Print) | 1864878657 |
Publication status | Published - 2006 |
Event | Assessment in Science, Teaching and Learning - Sydney Duration: 28 Sept 2006 → 28 Sept 2006 |
Conference
Conference | Assessment in Science, Teaching and Learning |
---|---|
City | Sydney |
Period | 28/09/06 → 28/09/06 |