The implications of extreme response style (ERS) for cross-cultural and comparative research

Abstract

Cross-cultural research in which fivepoint, Likert-type and semantic-differential scales are utilized, is a popular research practice. Extreme response style (ERS) may contaminate the validity of research results, however this possibility is often ignored in behavioural science research. In this study, the influence of biographical variables on extreme response style and the contaminating effect thereof on the validity of research results is investigated. The results of the study reveal that culture separately, and interaction with age and gender has a meaningful influence on ERS when five-point scales are utilized. The underlying causes of the phenomenon can however not exclusively be ascribed to biographical variables.

Opsomming
Kruiskulturele navorsing waarin vyfpunt, Likerttipe- en semanties-differensiale skale benut word, is 'n bekende navorsingspraktyk. Ekstreme responsiestyl is egter daartoe in staat om die geldigheid van sondanige navorsingsresultate te kontamineer, maar hierdie moontlikheid word dikwels in die gedragswetenskaplike navorsingspraktyk geignoreer. In hierdie studie word ondersoek ingestel na die invloed van biografiese faktore op ekstreme responsiestyl en die kontaminerende effek daarvan op die geldigheid van navorsingsresultate. Daar is gevind dat kultuur in interaksie met ouderdom en geslag, ERS betekenisvol beinvloed wanneer vyfpuntskale gebruik word. Die onderliggende oorsake van die verskynsel kan egter nie uitsluitlik aan biografiese veranderlikes toegeskryf word nie.