ABSTRACT
Summative and benchmarking surveys to measure the postgraduate student research experience are well reported in the literature. While useful, we argue that local instruments that provide formative resources with an academic development focus are also required. If higher education institutions are to move beyond the identification of issues and benchmarking practices, the scope of survey results and their reporting need to enable and foster appropriate changes in disciplinary practices. Robust, locally developed instruments can provide detailed, programme-specific information and foster timely changes in practice with direct benefits for postgraduate respondents. Unlike benchmarked surveys, local tools can adapt to explore and examine specific concerns of students, supervisors and academic developers. Coupling high-response rates and follow-on engagement with participant feedback, well-designed local instruments provide clear and irrefutable indicators to programme and university administrators of specific disciplinary strengths and weaknesses in postgraduate pathways. In this paper, we discuss the development of a research student survey specifically designed to support academic development purposes in strengthening and enhancing the postgraduate experience.
Notes
1. The postgraduate research student community comprises those undertaking independent doctoral-level research and masters students undertaking at least a thesis contributing 50% to the final grade.
2. Future data analysis will consider differences in responses as a function of thesis life cycle.
3. For a full list of UCPEQ items, contact the authors.
4. Questions related to thesis examination were moved to a companion exit survey in 2012.
5. The University of Canterbury is organised into five Colleges – Arts, Business & Law, Education, Engineering and Science – and within each are a number of constituent schools/departments.
6. An analysis of statistical differences between the proportions of agreement across iterations of the survey revealed few statistically significant differences (when accounting for multiple comparisons), despite the large effect sizes.