Papers from the International Association for Cross-Cultural Psychology Conferences

Publication Date

2011

Abstract

Bias in terms of culture has been posing threat to cross-cultural research since the very beginning of crosscultural endeavor. Metric and statistical methods have been discussed in literature in order to deal with this type of bias; however, some of these methods show side-effects on the level of scale validity and some others with so stringent effects on the available information and allow for very limited variance to be interpreted. The present twofold study describes yet another method, this time based on country clusters, following the idea introduced by Georgas & Berry (1995) of employing country sets based on their eco-cultural or psychological variables rather than single countries. In our study, the country clusters were derived from a different construct than the target one; the clusters of countries were formed using information from the European Value Survey Work Values, but the target construct in respect to bias reduction was the Person-Job Fit. Starting with 33 European countries and through trigonometrically transformed Multidimensional Scaling solutions, we arrived at a system of homogeneous clusters of countries in respect to their factor structure similarity. This similarity is not based on actual distribution resemblance levels, but on factor structure similarity as computed and utilized through the “hit” matrix. Testing for factor structure equivalence in the Person-Job Fit construct(s) for four European countries through covariance structure analysis, we contrasted two research methods, namely, the traditional across-countries approach and the method of aggregating some of the countries involved into clusters with a homogeneous factor structure. The findings showed that the aggregation technique reached acceptable levels of statistical support for the emerging factor structures, whereas the traditional approach did not statistically support the structures reached. Possible statistical artifacts were also tested through a third research condition, under a “homogeneity” rationale.

Included in

Psychology Commons

Share

COinS