Improving the Reliability of Kolb's Revised Learning Style Inventory

Abstract
The format of Kolb's Revised Learning Style Inventory (LSI II) was modified to eliminate a probable response bias and to investigate the inventory's subsequent reliability and stability. While internal-consistency estimates for the modified instrument dropped as expected in an initial study, test-retest reliabilities and kappa coefficients increased dramatically. A second study replicated the results of the initial research. The increased stability of the modified version argues against dismissal of the LSI as an instrument for the study of learning styles.