The effect of web survey design features on user response. Master of Distance Education thesis, Athabasca University

Abstract

The purpose of this study was to determine the effect of two web survey designs on the response time of novice and of expert computer users. This thesis replicated two studies: (1) Dillman, Tortora, Conradt, and Bowker (1998), an investigation of plain versus fancy web survey design on response rates, and (2) Bowker and Dillman (2000), a study of the effect of left versus right screen alignment in web surveys on user preference. The sample (N = 40) consisted of Masters of Distance Education (MDE) students of Athabasca University during the academic year of 2001 - 2002. The design was a 2 x 2 (type by ability) factorial post-test only design. A pre-test served as a selection method to classify the "novice" and "expert" computer user. Forty subjects were assigned to two groups. While the experimental group was assigned a fancy survey with sophisticated alternative attributes, the control group was assigned a plain survey with a traditional design. A post-test comprised of subjective indicators was designed to evaluate user preference on the survey attributes. A two-way ANOVA analysis was carried out and, although there was no significant main effect of the ability factor, there was a significant effect of the type factor and a significant interaction. Possible explanations for this interaction were drawn from the results of the objective and subjective indicators. Implications and recommendations for future research were made.