Short-Form Measures of Well-Being Centered Leadership for Matrixed Healthcare Systems

This study extracted the empirically strongest short-form abbreviations of the adapted Mayo Leadership Impact Index (MLII) and evaluated their comparative validity and reliability versus the full-length reference scale.

Data were sourced from physician-respondents to the 2023 (n=158) and 2024 (n=112) annual "well-being" surveys in a tri-hospital Cardiovascular healthcare system in Texas. Priority ranking of the full-length scale's items based on factor loadings from single-factor confirmatory factor analysis and information function plots from unidimensional graded response model analysis yielded single-, 3-, and 4-item short-form scales. Convergent validity of short-form versus full-length scales was compared using Spearman correlations with professional fulfillment, self-valuation, autonomy support, peer connectedness, and peer respect measures. Divergent validity was tested via correlations with occupational burnout. Internal consistency was based on Cronbach's and ordinal reliability coefficients. Inter-rater reliability was based on areas under the ROC curve, accuracy, precision, sensitivity, specificity, and Cohen's kappa coefficient.

Single-item, 3-item, and 4-item short-form scales, from the four highest ranked items, had similar convergent and divergent validity indexes plus inter-rater reliability to the 9-item adapted MLII. Internal consistency reliabilities for the 9-item, 4-item, and 3-item versions of the adapted MLII were >0.900. Sensitivity/specificity ratios and kappa coefficients for MLII-4, MLII-3, and MLII-1 extracts versus the adapted MLII-9 were all >0.700.

Findings verify the reliability and validity of short-form abbreviations of the adapted MLII as lower-burden options for assessing well-being centered leadership within matrixed healthcare systems.