A multisample model validation of the evidence-based practice questionnaire.

Evidence-based practice may be implemented more successfully if the barriers to its implementation have been previously identified. Many of the available instruments to measure these barriers have been validated in single samples or without confirmatory analyses. The objective of the study was to contrast the goodness of fit of two measurement models (24 items and 19 items) for the Spanish version of the Evidence-Based Practice Questionnaire (EBPQ) in a sample of 1,673 full-time registered nurses in 10 hospitals and 57 primary health care centers in the Spanish Public Health Service. The 19-item model performed better in all four subsamples. A hypothesis of strict invariance, with equal factor loadings, intercepts, and error variance in all contexts in which it was evaluated, was supported. Goodness-of-fit indices provided strong evidence of good fit according to standard cut-off criteria in a multisample confirmatory factor analysis.