Effects of the Number of Response Categories and Verbal Anchoring on Rating Scales

• In Event: Roundtable Session 60
  In Roundtable Session: Collection of Behavioral Information

Abstract

Educational and psychological scales and instruments are commonly used to measure opinions, personality traits and other phenomena in education, psychology, health, and marketing. One of the primary considerations in scale development is the choice of response options or categories. Given concerns about using ordinal data for statistical analyses that typically require interval-level assumptions, it is worth investigating how closely multiple-response scales can measure interval-level data. The research issue for this exploratory study was to use item response theory (IRT) to compare the effects of the number of response categories and verbal anchoring on Likert-scale items as determined by effective category use and distance between categories as well as distributions associated with each type of scale.

Authors

• Julia Shaftel, The University of Kansas

• Brooke Nash, The University of Kansas