Original Article
Can response variance effectively
identify careless respondents to
multi-item, unidimensional scales?
Received (in revised form): 21st June 2015
Danny Weathers
is an Associate Professor in the Department of Marketing at Clemson University. His research focuses on methodological issues, pricing, and consumer behavior, and has appeared in top journals including the Journal of Marketing Research, Journal of Retailing, Journal of Business Research and Marketing Letters.
Ahmet Bardakci
is Associate Professor and Chair of Production Management and Marketing, Department of Business, Pamukkale University. Correspondence: Danny Weathers, Department of Marketing, College of Business and Behavioral Sciences, Clemson University, Clemson SC 29634, USA
E-mail: pweath2@clemson.edu
ABSTRACT Because careless responding to questionnaire items can be quite common,
and even low rates of careless responding can substantially impact some analyzes, researchers have been advised to remove careless respondents before data analysis. However, identifying these respondents is a non-trivial task. In the context of multi-item, unidimensional scales, it has been suggested that response variance may hold information about careless responding. This notion was tested in the current research with two studies. Data collected from respondents indicates that people responding carelessly display more
variance in their responses than people responding with more effort. Given thisfinding, a
procedure which uses measures of response variance and cluster analysis to identify careless respondents was developed. The effectiveness of the procedure with different
specifications was tested with simulated data and validated with data from actual
respon-dents. On the basis of the results, we advocate using the procedure with specifications that conservatively identify careless respondents. Such an approach will identify the most extreme careless respondents, while maximizing the retention of careful, honest respon-dents. We discuss the advantages the developed procedure has over existing procedures for identifying careless respondents.
Journal of Marketing Analytics (2015)3, 96–107. doi:10.1057/jma.2015.6; published online 20 July 2015 Keywords: careless responding; random responding; response styles; multi-item
scale; cluster analysis; response variance
INTRODUCTION
Scaled response items are ubiquitous in survey
research. Ideally, people should select
responses based only on the substantive
meaning of the items. However, responses
can be influenced by response styles, or
tendencies to respond in ways that re
flect
information other than what items were
intended to measure (Paulhus, 1991).
Respondents are capable of a number of
© 2015 Macmillan Publishers Ltd. 2050-3318 Journal of Marketing Analytics Vol. 3, 2, 96–107 www.palgrave-journals.com/jma/