AbstractA study was conducted on the Felder‐Soloman Index of Learning Styles© (ILS) to assess reliability, factor structure, and construct validity as well to determine whether changing its dichotomous response scale to a five‐option response scale would improve reliability and validity. Data collected in this study had internal consistency reliability ranging from 0.55 to 0.77 across the four learning style scales of the ILS. Factor analysis revealed that multiple factors were present within three of the learning style scales, which correspond to known aspects of the scale definitions. The factor analysis and direct feedback from students on whether they felt their scores accurately represented their learning preferences provide evidence of construct validity for the ILS. Changing the response scale improved reliability, but it did not change the factor structure substantially nor did it affect the strength of the evidence for construct validity based on student feedback.
Read full abstract