Correcting for Response Style Effects on Service Quality Measures

Response style effects due to question items, answer formats and cultural factors have gained considerable attention in marketing research. However, the published work in this field only addresses effects based on measurements in the world of Likert-type. This study sheds light on the variation in estimated response style effects using conventional agreement scales and contrasting them with using confirmation/disconfirmation scales typically applied to service quality measurement. An empirical study within the passenger airline industry highlights substantial differences. They are partly due to different item scales used as the basis for calculating response styles, partially due to the scale types and almost independent from language (cultural) differences. In general, the results give initial directions for ´calculating response style effects, which, nevertheless, merits more experimental testing.