Title: Creating Affective Scales: What Does the Research Say?

Abstract: Have you ever wondered about the “rules” for constructing affective items? For example, should you use vague wording to leave respondents room for interpretation, or should you be as specific as possible? Does the order in which items are presented matter, or is it arbitrary? Many researcher recommend using negatively oriented items on affective scales – is this a good idea or should you stay away from this type of item. In the presentation I first introduce and discuss theories of response processing, with a focus on how these can inform our understanding of the effects of these scale characteristics. I then review the research on the impact of vaguely worded and negatively worded items, and on how item order effects may impact responses, and present the results of research I have done recently with my students on each of these issues.

Presentation Slides

Biography: Deborah Bandalos, Ph.D., obtained the doctorate in Educational Measurement, Statistics, and Evaluation from the University of Maryland in 1991 and held academic positions at the University of Nebraska and the University of Georgia before accepting her current position as Professor and Director of the Assessment and Measurement Doctoral Program at James Madison University, where she teaches courses in exploratory factor analysis, measurement theory, and missing data methodologies. Her research areas include the effects of item wording changes in instrument development and structural equation modeling. Dr. Bandalos is the author of the forthcoming textbook Measurement Theory and Applications in the Social Sciences (Guilford) and has published articles in a variety of journals as well as several book chapters in the areas of structural equation modeling, exploratory factor analysis, and item and scale development.  She is active in professional service, currently serving as associate editor of Multivariate Behavioral Research and previously as associate editor of Structural Equation Modeling. In addition, Dr. Bandalos serves on the editorial boards of Psychological Methods and Applied Measurement in Education and holds leadership positions in Division 5 of the American Psychological Association and the Society for Multivariate Experimental Psychology.