210 likes | 307 Views
Scenario-Based Scales Measuring Cultural Orientations of Business Owners Christine Koenig, Holger Steinmetz, Michael Frese, Andreas Rauch University of Giessen, Germany Zhong-Ming Wang University of Zhejiang, China. Introduction.
E N D
Scenario-Based Scales Measuring Cultural Orientations of Business Owners Christine Koenig, Holger Steinmetz, Michael Frese, Andreas Rauch University of Giessen, Germany Zhong-Ming Wang University of Zhejiang, China
Introduction • The scales differ from other scales commonly used in cross-cultural research: • Instead of measuring culture at the aggregate level, the scales measure cultural orientations at the individual level. • Instead of being based on Likert items, the scales are based on scenarios.
Individual-Level Measurement • Cross-cultural studies on individual business owners should measure cultural orientations rather than culture. • Cultural orientations are manifested in practices and values of individuals. • Among the manifestation of cultural orientations, we consider practices to be more relevant for studying business owners than values.
Individual-Level Measurement • We are not aware of any cultural orientation scales measuring practices that are suitable for business owners. • Therefore, we developed cultural orientation scales that measure the practices owners apply in their businesses.
Individual-Level Measurement • The scales measure seven cultural orientations: • uncertainty avoidance • power distance • collectivism • assertiveness • future orientation • humane orientation • performance orientation
Scenario-Based Measurement • Cultural orientations should be measured using scales based on scenarios rather than using scales based on Likert items. • Likert items and scenarios differ in the measurement of cultural orientations.
Scenario-Based Measurement • Scales based on Likert items tend to hold lower cross-cultural validity than scales based on scenarios: • Likert items are more likely to be interpreted differently by people from different cultures than scenarios. • Likert items are more affected by the reference group effect than scenarios.
Development • We developed the cultural orientation scales in a team of Chinese and German researchers. • We created scenarios that consist of social situations and behavioral options. • The social situations describe problems owners may encounter in their businesses. • Each of the social situations represents one of the cultural orientations.
Development • The behavioral options describe behaviors owners may show to solve the problems. • The first option represents of low score on the cultural orientation, whereas the second option represents a high score. • Between the two behavioral options, there are two mirror-inverted three-point scales.
Participants • Among the 461 owners who completed the scales, were 260 Chinese and 201 German owners. • They were not only owners but also managers of their businesses and had at least one employee. • Their businesses belonged to information technology, hotel and catering, automobile, or construction industry.
Cross-Cultural Validation • To ascertain whether the cultural orientation scales hold cross-cultural validity, we tested five forms of invariance: • configural invariance • metric invariance • scalar invariance • factor variance invariance • error invariance
Configural Invariance • The model of configural invariance comprised 23 scenarios that measured the seven cultural orientations.
Configural Invariance • The model of configural invariance comprised 23 scenarios that measured the seven cultural orientations. • The model of configural invariance provided adequate fit (Chi2(418) = 603.45; RMSEA = .044; CFI = .94).
Metric Invariance • The factor loadings were constrained to be equal across the Chinese and the German samples. • The model of full metric invariance achieved adequate fit (Chi2(434) = 628.26; RMSEA = .044; CFI = .93).
Scalar Invariance • The item intercepts were restricted to be equal across the Chinese and the German samples. • The model of partial scalar invariance achieved adequate fit (Chi2(446) = 641.92; RMSEA = .044; CFI = .93). • Given metric and scalar invariance, the means of the seven cultural orientations can be meaningfully compared across both samples (Steenkamp & Baumgartner,1998).
Factor Variance Invariance • The factor variances were constrained to be equal across the Chinese and the German samples. • The model of partial factor variance invariance provided adequate fit (Chi2(451) = 646.71; RMSEA = .044; CFI = .93). • Given metric and partial factor variance invariance, different covariances can be interpreted as different correlations for the remaining five cultural orientations (Marsh & Hocevar, 1985).
Error Variance Invariance • The error variances were restricted to be equal across the Chinese and the German samples. • The model of partial error variance invariance achieved adequate fit (Chi2(463) = 664.85; RMSEA = .044; CFI = .93). • Given metric and factor variance invariance, this result implies that the Chinese sample showed a lower reliability in half of the scenarios.
Conclusion • All five forms of invariance were at least partially supported. • Given metric and scalar invariance, the means of the seven cultural orientations can be meaningfully compared across the Chinese and the German samples. • Given metric and factor variance invariance, different covariances can be interpreted as different correlations – except for uncertainty avoidance and assertiveness. • Given metric invariance, regression coefficients can be compared across the Chinese and the German samples.