Evaluating Change and Stability in Learning Style Scores: a methodological concern

Abstract

Learning styles are purported to be relatively stable characteristics with some change or development expected. Some studies using Kolb's Learning Style Inventory (LSI) have reported significant positive test‐retest correlations of LSI scores or nonsignificant repeated‐measures ANOVAs and concluded that learning styles are stable. This study examined stability and change on Kolb's revised Learning Style Inventory (LSI‐1985) using 152 participants at two points in time separated by about 10 weeks. A variety of statistics were used to evaluate stability and change in LSI‐1985 scores for the four subscales and two dimensions and the four learning styles. The use of test‐retest correlations, differences between means and other methods emphasising group effects were criticised. It was recommended that researchers also analyse and report the stability and change of style categories directly, not just score changes. These comments are also applicable to other learning style measures such as the Learning Style Questionnaire.