Search

A SERIES of papers, by M. Mercadier, on the dimensions of physical quantities, has recently been appearing in the Comptes Rendus. They are summarised and extended in the Journal de Physique (July, 1893, p. 289).

The self-report final-say decision measure of marital power has been declared invalid on the basis of low levels of interrelationships between this and other measures of marital power. However, there are alternative explanations for the poor performance of this measure to which little attention has been given previously: improper analytical procedures and computational errors. Data from 51 pairs of parents are used to show the results of miscalculations and inadequate procedures (similar to those encountered in previous marital power validity research) on the assessment of the relative validity of thefinal-say measure. In addition, these data are used to demonstrate that, even when analytical procedures are carefully selected, external criterion variables must be used if the meaning of interrelationships among measures is to be correctly interpreted.

OBJECTIVE:To develop and validate SF-12 osteoarthritis-specific health index (SF-12 OASHI). METHODS:Patient data on SF-12 and six osteoarthritis (OP) clinical variables (physician and patient global assessments, pain intensity, knee pain on weight bearing and motion, time to walk 50 feet) at baseline and week, 6, from two placebo-controlled clinical trials (n = 422), assessing efficacy of NSAIDs in OA patients were used. Using canonical correlation analysis, a SF-12 OASHI was developed in 75% of the sample (n = 317) by adding individual SF-12 item scores at baseline, each multiplied by their respective OA specific weights (canonical crossloadings on clinical variables). Validation (developmental sample) and cross-validation (25% holdout sample [n = 105], and another clinical trial sample [n = 170]) of the SF-12 OASHI were conducted by examining its correlation with clinical variables, and by computing the relative validity (RV) estimates of SF-12 OASHI as compared to physical (PCS12) and mental component summary scores (MCS12), using baseline and change scores at 6 weeks. Correlation between SF-12 component score and clinical variable was divided by correlation between OASHI and respective clinical variable to arrive at the RV. RESULTS: SF-12 OASHI demonstrated significant correlations with individual clinical variables ranging from −0.19 to −0.54 (p < 0.05). In general, SF-12 OASHI was more sensitive than the PCS12 and MCS12 scores as indicated by higher correlation coefficients with clinical variables, at baseline, in developmental and two cross-validation samples. At baseline, the RV coefficients for SF-12 OASHI ranged from 0.64 to 1.09 for PCS12 and 0.37 to 0.89 for MCS12. In general, SF-12 OASHI also showed more responsiveness to changes in clinical variables at 6 weeks as compared with PCS12 and MCS12 scores. CONCLUSION:The SF-12 OA-specific health index is a comprehensive and more sensitive measure of patient quality of life in OA as compared with PCS12 and MCS12.