Introduction
The research by Weijters, Cabooter and Schillewaert (2010) investigates the effect of the number of response categories and the labeling of response categories on misresponse, extreme response style, and the net acquiescence response style.
Market researches utilize questionnaires in obtaining consumer data.
Most researchers use Likert items in providing several design-related choices of responses for the respondents.
The choice of the format of rating the scales has raised concerns over the period in market researches.
The choice of the number of response categories and the number of response labels in the questionnaire has raised concerns.
The effect of the rating scale format, the number of response categories and the number of response labels on the respondents’ is, therefore, investigated in the study.
Different researchers use four, five, six or seven responses which are either fully labeled or extreme labeled.
There are inconclusive researches on the relationship between response styles and scale formats.
Conceptual Background
The net acquiescence response style (NARS) bias has an influence on the rating scale measures.
The extreme response style (ERS), on the other hand, affect the spread of data.
The misreponse to reversed items (MR) is used when countering the NARS, where the responses are recoded in the opposite direction and given out to respondents.
Methodology
The research was conducted using two empirical studies.
The first study utilized a 1207 randomly selected sample from Dutch-speaking men for a internet marketing research in Belgium.
The research used an online survey through manipulation of response categories labels and number of response categories.
The study used a questionnaire which consisted two parts; one part for measuring MR while the other part for measuring NARS and ERS.
NARS and ERS were measured using log odds which are illustrated with the equations below.
NASR = ln ((agreements + 1)/ (disagreements + 1)).
ERS = ln ((extreme responses + 1) / (non – extreme responses + 1)).
The analysis involved correlation between NARS and ERS together with the comparisons with the traditional options.
Reference
Weijters, B., Cabooter, E., & Schillewaert, N. (2010). The effect of rating scale format on response styles: The number of response categories and response category labels. International Journal Of Research in Marketing, 27(3), 236-247.