Factorial structure of Individual Work Performance Questionnaire (Version 1.0) revisited: Evaluation of acquiescence bias.

Abstract

The objective of this study is to evaluate the factorial structure of the Spanish version of the Individual Work Performance Questionnaire (Version 1.0) and to estimate the impact that acquiescence has on it as response bias. Exactly 500 workers from organizations from different industries, primarily located in Ecuador, participated in this study. The comparison of different models using Confirmatory Factor Analysis proved that when acquiescence is not controlled, evidence leads to the rejection of a one-dimensional-or essentially one-dimensional-structure (bifactor model), thus questioning the initial conceptualization of the construct. But when this response bias is controlled, both the one-dimensional model (χ2 = 429.608 [223], p < .001; CFI = .974; TLI = .982; RMSEA = .043; SRMR = .063) and the bifactor model (χ2 = 270.730 [205], p = .001; CFI = .992; TLI = .994; RMSEA = .026; SRMR = .047) show relevant improvement in terms of goodness of fit over the three-correlated-factors model (χ2 = 537.038 [132], p < .001; CFI = .950; TLI = .942; RMSEA = .079; SRMR = .070). However, the low reliability of the substantive factors of the bifactor model makes the one-dimensional model preferable in applied studies. Finally, the results show how mistakes could be made when concluding on the possible relationships between work performance and other relevant variables, in case acquiescence is not controlled.