Ceiling effects associated with response scales
The recent WORC team has been working on investigating ceiling effects that occur in survey data.
A ceiling effect in physical or medical sciences refers to a condition where an independent variable such as a treatment no longer produces changes on its dependent variable (i.e., the outcome) (e.g., Nagel & Spuler, 2019; Neijts et al., 2014).
A ceiling effect associated with statistics in social sciences refers to a condition where the majority of the data are close to the upper limit (Cramer & Howitt, 2004). A ceiling effect in survey data can occur due to limitations in the survey instrument’s sensitivity and accurately measuring individual respondents’ true responses while distinguishing them from others’ responses, resulting in little variance in the data (French et al., 2018; Taylor, 2010). Such skewed data may fail to provide you with actionable information.
The WORC team’s article has been published in the 2020 Volume 59: Issue 6 of Performance Improvement Journal:
- Chyung, S. Y., Hutchinson, D., & Shamsy, J. (2020). Evidence-based survey design: Ceiling effects associated with response scales. Performance Improvement Journal, 59(6), 6-13. https://doi.org/10.1002/pfi.21920
WORC researchers who participated in in this project are: