Correspondence: Bryan Lilly, University of Wisconsin Oshkosh, 800 Algoma Boulevard, Oshkosh WI 54901, USA. E-mail: lilly@uwosh.edu

¹is Professor of Marketing and Associate Dean at the College of Business at University of Wisconsin Oshkosh. His research interests include consumer decision making, research methods and the marketing of new products. He consults with companies in areas of product development and marketing strategy.

Received 17 July 2009; Revised 17 July 2009; Published online 7 September 2009.

Abstract

For consumer experiments, if stimuli are balanced across subjects, what order of stimuli within subjects is optimal? In this study, four stimulus orders were tested within subjects, all balanced across subjects. Stimulus orders were tested in a similarity judgment task and were evaluated based on six criteria. The best overall performance was achieved from a non-balanced order, wherein subjects saw all stimuli first and judged similarities as desired. A commonly used within-subject balanced order performed relatively worse, and was no better than a stimulus order proposed over 100 years ago.