Development of an Instrument to Measure Quality of Nurses' Worklife

Abstract

This study reports the development process and factor structure of a questionnaire to measure quality of nursing work life for registered nurses in hospital settings. Included is information on item generation, content validation, initial pilot testing, and results from the full study.

The instrument was administered to registered nurses in Illinois. Confirmatory factor analysis using maximum likelihood estimation was used to examine the fit of the data with the hypothesized measurement model. Two rival models were also tested. Chi-square (X2 (819) = 3,220.70, p < .0001), goodness-of-fit indices (CFI = .48, GFI = .63, AGFI = .60), and parsimony ratios (GFI = .60, NFI = .39) were analyzed for the hypothesized four factor model. The chi-square analysis, fit indices, and parsimony ratios indicated another model would be a better fit. A revised four factor model was estimated based on Lagrange multiplier tests and theoretical considerations.

Chi-square (X2 (819) = 3,050.23, p < .0001), goodness-of-fit indices (CFI = .52, GFI = .67, AGFI = .64), and parsimony ratios improved slightly (GFI = .64, NFI = .43) for the revised model and parameter estimate t-tests for all but three items were significant at p ≤ .05. Internal consistency reliability for the final instrument sub-scales included an alpha coefficient of .56 for the 4 item Work life/Home life Scale, (n = 265), .58 for the 11 item Work Design Scale, (n = 265), .88 for the 21 item Work Context Scale, (n = 265), and .49 for the 6 item Work World Scale (n = 265). An exploratory factor analysis revealed a two factor solution, but the two factors made little conceptual sense and were not amenable to being named. Evaluation and recommendations regarding the questionnaire are included.