Measuring health sciences students' attitudes toward persons with disabilities: Is one scale better than another?

Area of Science:

• * Educational Psychology
• * Disability Studies
• * Psychometrics

Background:

• * Negative attitudes toward persons with disabilities (PWDs) contribute to stigmatization and exclusion in educational settings.
• * Effective measurement tools are crucial for assessing and addressing these attitudes among health science learners.
• * Existing scales require rigorous psychometric evaluation for reliability and validity.

Purpose of the Study:

• * To compare the psychometric properties of the Multidimensional Attitudes Scale Toward Persons with Disabilities (MAS) and the Attitudes and Perspectives Toward Persons with Disabilities Scale (APPD).
• * To evaluate internal consistency, factor stability, factor replicability, and convergent validity of both scales.
• * To assess scale performance across different measurement occasions in health science students.

Main Methods:

• * Exploratory factor analysis (EFA) with bootstrapping was used to analyze data from Medical (n=102) and Doctor of Physical Therapy (n=39) students.
• * Internal consistency (Cronbach's alpha) and factor replicability (Tucker congruence coefficients) were assessed.
• * Confirmatory factor analysis (CFA) and convergent validity analysis (correlation) were conducted.

Main Results:

• * Both MAS and APPD yielded three-factor solutions, but MAS demonstrated superior internal consistency (α ≥0.838) and factor replicability (Tucker's $\phi$ ≥0.88).
• * Neither scale showed ideal model fit in CFA when applying post-intervention data to the pre-intervention model.
• * Convergent validity analysis revealed a moderate correlation (r=0.368, p<0.001) between MAS and APPD, indicating limited overlap.

Conclusions:

• * The MAS exhibits better psychometric properties than the APPD for assessing health science students' attitudes toward PWDs.
• * Both scales present limitations in factor stability and replicability across measurement occasions.
• * Further refinement and validation of both scales are necessary for reliable and consistent application.