A Construct Validity Study of Clinical Competence: A Multitrait Multimethod Matrix Approach

作者：Baig Lubna^*; Violato Claudio; Crutcher Rodney

来源：Journal of Continuing Education in the Health Professions, 2010, 30(1): 19-25.

DOI：10.1002/chp.20052

• construct validity
• clinical competence
• multitrait multimethod matrix
• international medical graduates (IMGs)
• practice readiness

Introduction: The purpose of the study was to adduce evidence for estimating the construct validity of clinical competence measured through assessment instruments used for high-stakes examinations.
Methods: Thirty-nine international physicians (mean age = 41 + 6.5 y) participated in high-stakes examination and 3-month supervised clinical practice to determine the practice readiness of physicians. Three traits-doctor-patient relationship, clinical competence, and communication skills-were assessed with objective structured clinical examinations, in-training evaluation reports, and clinical assessments. These traits were intercorrelated in a multitrait multimethod matrix (MTMM).
Results: The reliability of assessments ranged from moderate to high (Cronbach's alpha : 0.58-0.98; Ep(2) = 0.79). There is evidence for both convergent and divergent validity for clinical competence, followed by doctor-patient relationships, and communications (validity coefficients = 0.12-0.85). The correlations between the same methods but different traits indicate that there is substantial method specificity in the assessment accounting for nearly one-quarter of the variance (23.7%).
Discussion: There is evidence for the construct validity of all 3 traits across 3 methods. The MTMM approach, currently underutilized, could be used to estimate the degree of evidence for validating complex constructs, such as clinical competence.

• 出版日期2010