Disentangling Administration Errors From Scale Development Errors in Survey Research

Abstract

In a recent manuscript, Friesner, et al. (2023) used the concept of information entropy to assess the quantity of information in survey responses. They demonstrate how assessments of the quantity of information can be used to identify possible errors in a survey’s administration. A major limitation of their methodology is that it assumes the survey items used to elicit consumer preferences were created appropriately and contained a meaningful quantity of information. The current study addresses this limitation by incorporating a methodology developed by Friesner et al. (2021) into the Friesner et al. (2023) methodology. The combined methodology is applied to the same data studied in both Friesner et al. (2021) and Friesner et al. (2023), which allows for a direct comparison of the quantity of information gained/lost from survey administration versus scale development. The results indicate that the survey used in the empirical application exhibits flaws in both scale design and survey administration.