A Psychometric Assessment of OCB: Clarifying the Distinction Between OCB and CWB and Developing a Revised OCB Measure

Abstract

This study was performed to (1) assess the appropriateness of using negatively worded items in organizational citizenship behavior (OCB) scales, (2) psychometrically demonstrate the construct distinctness of OCB and counterproductive work behavior (CWB), and (3) report on a revised, short-form OCB scale. Leveraging classical test theory (CTT) and item response theory (IRT), we demonstrate that the negatively worded items from a popular OCB scale (Williams and Anderson 1991) do not measure OCB, but rather a unique construct (CWB). CTT analyses (factor analyses) indicate that the negatively worded items load onto a unique factor when the scale is analyzed on its own and load onto a CWB factor when the scale is analyzed with a CWB scale. Additionally, IRT analyses indicate that the negatively worded items exhibit lower discrimination parameters and higher levels of local independence than the positively worded items, and similar discrimination parameters and levels of local independence as the CWB items. In turn, IRT analyses were used to identify the best items from the OCB scale to create a revised, short-form scale. The short-form scale showed comparable or improved convergent and discriminant validity and internal consistency reliability, as well as similar patterns of psychometric information yielded from IRT analyses, compared to the original scale. In short, the revised measure better aligns with conceptual definitions of OCB, demonstrates acceptable psychometric characteristics, and, given its reduced length, is of more practical value to researchers wishing to assess this construct within different types of research designs (e.g., longitudinal, multi-source).