Reliability Analysis of Categorical Survey Items for Social and Emotional Learning Measures

• In Event: Analyzing Survey Data in Education Research

Abstract

The current study examines the configuration of survey items used to measure social and emotional learning. Of the six scales composing this instrument, the Commitment to Learning (CtL) scale had a relatively low level of reliability. A recommendation is made to modify the CtL scale by considering the removal of two items from this scale to improve the reliability. Students taking this survey exhibited similar response patterns among items with the same stem even if the content of the items represented a different construct according to the SEL measures. A suggestion is made, then, for researchers to be conscientious about nesting items under the same stem as this is a potential source of bias in survey design.

Authors

• Amaniel Mrutu, University of Minnesota - Twin Cities

• Ernest C. Davenport, Jr., Retired