Vortrag | Margit Pohl, Technische Universität Wien

Abstract:

Quantitative and qualitative methods of analysis and interpretation of data have different strengths and weaknesses. Both are appropriate for the investigation of the interaction of human users with computer systems, but it depends on the goal of the research which methods should be adopted. Evaluation studies have to be planned carefully. Researchers sometimes underestimate, for example, that the choice of variables which are used as indicators for the investigated phenomena or the development of tasks for the investigation can influence the results of the research heavily. It is, for example, not a trivial issue what questions are asked in interviews or questionnaires to assess the usability of a system. In this context, I want to address logfile analysis and thinking aloud specifically. In this context, categorization plays an important role. Categorization is discussed extensively in social sciences. In HCI, this issue has not been addressed at great length. Nevertheless, it is not a trivial issue to decide which granularity to use when interpreting data gained from logfiles or thinking aloud protocols or which categorization frame to adopt to interpret the data. In addition, I also want to discuss issues of mixing quantitative and qualitative methods, so that the methods complement each other.

Speaker's Bio:

Margit Pohl studied Computer Science and Psychology in Vienna. She is associate professor at the Vienna University of Technology where she works at the Institute for Design and Technology Assessment/HCI group. She is mainly interested in the cognitive aspects of the interaction of humans with computers. Her most important research interests are Cognitive Psychology and Visualization. She has conducted many projects in the area of Cognitive Aspects of  Visualization and more generally in HCI.