Multiple Biases in Rating Scale Construction
Friedman, H. H. and Amoo, T. (1999). "Multiple Biases in Rating Scale Construction," Journal of International Marketing and Marketing Research, Vol. 24, October 1999, 115-126.
25 Pages Posted: 28 Apr 2014
Date Written: October 1, 1999
A significant portion of research relies on surveys that employ rating scales. It is thus very important to ensure that the rating scale as a measuring instrument is both reliable and valid. Unfortunately, there are many ways that a rating scale can be biased. The current study investigates the extent to which a study can be biased — wittingly or unwittingly — by choosing a particular type of rating scale. Dishonest researchers can manipulate the outcome of their research, if they wish, and purposely bias a study. An experiment was conducted in which seven questions were asked in two different ways in order to determine whether and how much a researcher could influence the outcome of the results by manipulating the rating scale. The results confirmed that the wording used in the rating scale could slant results in the direction desired by the researcher.
Keywords: rating scales, measurement bias, ethics in research
JEL Classification: C00, C10, C40, I20
Suggested Citation: Suggested Citation