A comparison of clinimetric and psychometric techniques for item reduction
Introduction! Clinimetric and Psychometric methods for it reduction have conflicting aims of scale heterogeneity and homogeneity, respectively. The two methods have seldom been compared and never prospectively. The purpose of this study was to determine if these two methodologies provided comparable scales in the development of an upper extremity disability measure. Item generation and preliminary item reduction led to a total of 70 items. The ultimate goal was a 30-item, self-administered questionnaire. Itathodai Psychometric analysis involved field testing the 70-item questionnaire on 407 patients. Bquidiscriminatory item total correlation (EITC} was used to select the top 30 items. Cronbach's alpha and factor analysis with orthogonal rotation were performed. Clinimetric testing involved a separate group of 76 patients who rated the importance and severity of the 70 items. The mean sums of the importance and severity scores were used to select the top 30 items. Clinimetric and psychometric analyses were performed independently and blind to the identity of the items. After unblinding, two groups of clinicians independently modified the 30item scales using clinical judgment. R*ult*i Cronbach's alpha was 0.97 for the top 30 items selected by EITC. Factor analysis revealed two factors: a physical function factor and a symptom/ social role factor. The two scales (after clinician modification) shared 16 items in common (p=0.17). The intra class correlation coefficient of the patient scores on the 30-item scales was 0.93 prior to clinician input and 0.97 after. Conclusion* Despite stated conflicting goals, Clinimetric and psychometric methods for item reduction led to the construction of similar functioning instruments.