Relevance & Research Question
When developing web-based course evaluation tools (for evaluating teaching e.g. in university or secondary schools), researchers and practitioners put much effort into reaching high psychometric properties of the instruments, realizing high participation rates, suitable and useful reports of the results, and the drawing of consequences (see Thielsch & Hirschfeld, 2011). However, there are three more vital aspects not to be overlooked, especially when using technology to aid collection, processing, and presentation of data: privacy, data security, and the ethical dimension of our actions. To account for these aspects is critical not only from a judicial point of view (European Union Data Protection Directive, 95/46/EC or, for Germany, Bundesdatenschutzgesetz, BDSG); considering them will arguably also foster acceptance of course evaluations among evaluators and evaluatees. This presentation therefore aims at discussing issues of privacy, data security and its ethical dimension when conducting online course evaluations.
Methods & Data
A review of the literature on the mentioned issues of course evaluations was conducted and paralleled to existing practice. Additional theoretical and practical aspects were examined. The following are key aspects being discussed: (1) carefully selecting (and limiting) the variables to assess; (2) securing assessment and storage of data via access control and encryption; (3) anonymized or pseudonymized archiving of data; (4) ethical aspects such as using non-reactive data and reporting of negative results.
Results
Online course evaluations show potential for very high data quality and benefits over traditional paper-pencil evaluations, in part due to the facilitated way to collect and process large data sets. However, the risks accompanying these facilitations are discussed, and recommendations for researchers and practitioners are given on how to control or even avoid them altogether.
Added Value
The study raises awareness for aspects of course evaluations that are often taken for granted or neglected, but may endanger data integrity as well as acceptance when not paid attention to. The study gives recommendations for conductors of web-based course evaluations on how to handle these aspects.