Bias in an Adapted Version of the 15FQ+in South Africa

Abstract

This study addressed the cross-cultural suitability of an adapted version of the Fifteen Factor Questionnaire (15FQ+) in South Africa. Criteria for changes to the items of the original 15FQ+ included comprehension and cultural appropriateness. The instrument was administered in English to a pool of 16 339 participants from all parts of South Africa who had applied for entry-level police jobs in the South African Police Services (SAPS). Bias was studied at construct and item level. A higher level of overall structural equivalence was reported for the adapted version when compared with results found in research with the original version. However, some scales remain problematic. A slight decrease in the number of biased items was also found for the adapted version and it was concluded that item bias is not a major problem in this version. Despite these results pointing to the apparent adequacy of the adaptations, only marginal increases were found in terms of the internal consistencies when compared to the original version and for the black groups, in particular, consistency levels remained low. These low levels of consistency continue to limit the usefulness of the questionnaire.

Validating a Measure of Productive Organisational Energy in the South African Context

“…Multiple studies have provided the empirical evidence to substantiate the notion that cultural differences may have an impact on participants' responses to particular psychometric instruments (Van Eeden & Mantsha, 2007;Meiring, Van de Vijver, & Rothmann, 2006). This results in the instrument being less valid for specific cultural groups.…”

“…This results in the instrument being less valid for specific cultural groups. Although research has confirmed that some psychometric instruments developed in other countries are valid when applied in the South African context (Storm & Rothmann, 2003), there have also been multiple instances where instruments which were not adapted for use in the South African context resulted in difficulties with reliability and validity (Gray & Durrheim, 2006;Meiring et al, 2006). This can be attributed to the fact that an individual's subjective culture provides the basis through which they perceive their social environment, which in turn discourages certain behaviors and cultivates others (Anastasi, cited in Schaap, Buys, & Olckers, 2003).…”

“…The aim of this research is to validate the Productive Energy Measure (Cole, Bruch & Vogel, 2005;2012) in the South African context, and to determine whether construct equivalence was present for White and African ethnic groups. South Africa is an inherently heterogeneous culture, and as a result researchers should exercise caution when applying measurement instruments developed overseas in the South African context (Meiring, et al, 2006). It is important that instruments developed and validated overseas are re-validated for use in the South African cultural context to ensure that construct and measurement equivalence and ultimately comparability are established.…”

Analysis of Corporate Entrepreneurship in Public R&D Institutions

“…Indeed, the application of such instruments across different cultures without possible adaptations may severely affect the validity of the respective instruments [24][25][26]. Due to the way in which individuals perceive their social and cultural milieu, they show certain forms of behavior which influence in their turn activities of interpreting and responding to the questionnaires.…”

Towards a Comprehensive Procedure for Developing Measurement Scales for Cross-Cultural Management Research

“…Another approach is to evaluate the influence of the biased items on the cross-cultural differences; a simple way is to compare the difference in size of the means on the original instruments with the difference on the instruments from which all biased items are removed. Although the differences can be very large in theory, the removal of biased items may not have major implications for their size and the implications for the interpretation of the cross-cultural differences may be minimal (e.g., Meiring, Van de Vijver, & Rothmann, 2006). Finally, the size of the intercept differences and the cross-groups mean differences should be taken into account when interpreting lack of scalar invariance.…”