Testing and Assessment in an International Context: Cross- and Multi-cultural Issues

Abstract

Globalisation, increase of migration flows, and the concurrent worldwide competitiveness impose rethinking of testing and assessment procedures and practices in an international and multicultural context. This chapter reviews the methodological and practical implications for psychological assessment in the field of career guidance. The methodological implications are numerous and several aspects have to be considered, such as cross-cultural equivalence or construct, method, and item bias. Moreover, the construct of culture by itself is difficult to define and difficult to measure. In order to provide non-discriminatory assessment, counsellors should develop their clinical cross-cultural competencies, develop more specific intervention strategies, and respect cultural differences. Several suggestions are given concerning translation and adaptation of psychological instruments, developing culture specific measures, and the use of these instruments. More research in this field should use mixed methods, multi-centric designs, and consider emic and etic psychological variables. A multidisciplinary approach might also allow identifying culture specific and ecological meaningful constructs. Non-discriminatory assessment implies considering the influence and interaction of personal characteristics and environmental factors.Testing and assessment in an international context 3 Testing and Assessment in an International Context: Cross-and Multi-Cultural Issues Globalisation, increase of migration flows, and the concurrent worldwide competitiveness impose a shift on intercultural research, towards an integrative background to both common and regional competencies to achieve added value and usefulness of testing and assessment techniques. Whenever a test is translated and adapted for use in another language or culture, it is mandatory to know that the process begins always with evaluation: the first issue concerns the conceptual definition and the context of its operationalisation, which means the identification of the relevant contents of cultural knowledge. In this case, the ingredients of knowledge mean the understanding of how culture is expressed through beliefs and values, behavioural expressions, symbols and habits, but also mean a balance between cognitive knowledge and attributes of good judgement to deal with the culture or sub-culture variables. The close connection between these two aspects can lead to a deep awareness of construct validation research within each population for which translation or adaptation occurs.As Hambleton (2005) pointed out, a distinction has to be made between test adaptation and test translation. In his view, the term "adaptation" is broader and more reflective of what should happen in practice when preparing a test that is constructed in one language and culture for use in a second language and culture. Test adaptation includes all the steps involved in developing a measure to assess a specific construct in a different language and culture, taking into account the specificities of that l...

Validation of a French Version of the Career Decision-Making Difficulties Questionnaire: Relationships With Self-Esteem and Self-Efficacy

“…Ensuring that "the instrument measures the same construct(s) in exactly the same way across all group" (Byrne & van de Vijver, 2010, p. 107) is a crucial prerequisite to valuable cross-group comparisons (Rossier & Duarte, 2019). In order to verify whether the same norms can be used in Switzerland and France, and among different groups, and to ensure that these groups can be meaningfully compared, measurement invariance was tested.…”

“…Following the procedure suggested by Van de Vijver and Leung (1997; see also He & van de Vijver, 2012), the configural, metric, and scalar invariance were tested by constraining all loadings across groups for the scalar invariance and by constraining all loading and intercepts for the metric invariance. Metric invariance thus also implies configural invariance, and scalar invariance implies both metric and configural invariance (Rossier & Duarte, 2019). From one level of invariance to the other, the change in the CFI should be less than .01 (Byrne & van de Vijver, 2010) or less than .002 according to Meade and colleagues (2008), and the change in RMSEA should be less than .05 (Cheung & Rensvold, 2002).…”

“…To be able to offer a validated instrument for both career counseling and research, the primary purpose of this study is to validate the French version of the CDDQ and to test whether the structure proposed by Gati et al (1996) can be replicated in two French-speaking countries, with students whose age corresponds to periods of major career changes. Measurement invariance across gender, age groups, countries, and general versus career counseling samples is investigated in order to identify whether unique or specific norms should be developed (Rossier & Duarte, 2019). The secondary purpose of this study is to examine the sensitivity of the CDDQ to discriminate a career counseling from a general student sample.…”

Exploring the stability of HEXACO‐60 structure and the association of gender, age, and social position with personality traits across 18 countries

“…A χ 2 / df bellow 5 reflects an acceptable model, GFI and CFI values above .95 indicate a good fit, and values ranging between .90 and .95 indicate an acceptable fit, RMSEA values below .08 indicate an acceptable fit, and values below .05 indicate a good fit. In order to analyze configural, metric, and scalar invariance, changes in model fit statistics were considered (Rossier & Duarte, 2019). To provide evidence of invariance, change in CFI values should be lower than .01, and change in RMSEA lower than .05 (Byrne & Van de Vijver, 2010; Rossier et al., 2016).…”

Dark Triad Traits, Social Position, and Personality: A Cross-Cultural Study

“…Considering parcels did allow to reach configural and somehow metric invariance, but clearly not scalar invariance, as did the ZKA-PQ/SF and HEXACO-60 in the same samples (Aluja et al, 2020; García et al, 2021). Several methodologists would claim that in this case means cannot be compared across cultures (Rossier & Duarte, 2019). However, this type of mismatch between the psychometric approach (large sample of uncorrelated items) and the statistical approach (parsimonious highly correlated observed variables), and the over emphasis given to the question of the measurement invariance did lead several methodologists to suggest that the criteria to assess invariance across cultures were too strict, considering differences not directly related with the latent constructs (Millsap, 2011).…”