An Evaluation of the Number of Response Options for Scales in Psychology

Self-report scales are used widely in the field of psychology. These scales tend to widely differ on scale format for many reasons including consistency, time issues, and convenience. Previous studies have found that scale format has an effect on response variance, and reliability, among other psychometric properties. However, these findings have been mixed. The purpose of this study is to assess the effects of number of response options on response patterns and internal consistency. We used a 5- and 7-point scale of the Rosenberg Self-Esteem measure. Undergraduate college students were administered this scale with either 5 or 7 response options. We found that frequency and response patterns did not differ between the 2 scales, but differences in response patterns per item were present. There were also mean differences between scales, although these effects were small. The number of response options did not affect reliability. Using descriptive statistics and t-tests, differences were not detected between responses to items presented with the 5- and 7-point response scales. Further research assessing more than one measure and comparing even, and odd numbered scales is needed to better understand the effects of number of response options on response patterns.