Little is known about cross-language measurement equivalence of the job content questionnaire (JCQ)
The purposes of this study were to assess the extent of cross-language differential item functioning (DIF) of the 27 JCQ items in six languages (French, Dutch, Belgian-French, Belgian-Dutch (Flemish), Italian, and Swedish) from six European research centers and to test whether its effects on the scale-level mean comparisons among the centers were substantial or not.
A partial gamma coefficient method was used for statistical DIF analyses where the Flemish JCQ was the reference for other language versions. Additionally, equivalence between the Flemish and Dutch translations was subjected to a judgmental review.
On average, 36% to 39% of the total tested items appeared to be cross-language DIF items in the statistical analyses. The judgmental review indicated that half of the DIF items may be associated with translation difference. The impacts of the DIF items on the mean comparisons of the JCQ scales between the centers were non-trivial: underestimated skill discretion (Milan), underestimated decision authority (Leiden), underestimated psychological demands (Milan women), and incomparable coworker support (Gothenburg 95).
Cross-language DIF of the JCQ among European countries should be considered in international comparative studies on psychosocial job hazards using JCQ scales.