Measurement Equivalence of English and Spanish Versions of the Perceived Leader Integrity Scale

IRT;Perceived Leader Integrity Scale;DFIT;measurement equivalence;leadership

Datta, Anasuya ; Joan J. Michael, Committee Chair,Adam Meade, Committee Member,Bart Craig, Committee Co-Chair,Datta, Anasuya ; Joan J. Michael ; Committee Chair ; Adam Meade ; Committee Member ; Bart Craig ; Committee Co-Chair

University:North Carolina State University

关键词: IRT;    Perceived Leader Integrity Scale;    DFIT;    measurement equivalence;    leadership;   

Others  :  https://repository.lib.ncsu.edu/bitstream/handle/1840.16/1851/etd.pdf?sequence=1&isAllowed=y

美国|英语

来源: null

【 摘 要 】

Research over the last three decades has addressed the importance of integrity in leadership (e.g., Burns, 1978; Fairholm, 1991; Posner & Schmidt, 1984; Vitell & Davis, 1990). Research and application are stunted without adequate measures that assess the extent to which leaders demonstrate ethical or unethical behaviors. As business activities between countries increase (Stephens & Greer, 1995), having tests available in multiple languages can have various benefits (Zumbo, 2003). The availability of a measure in different languages can allow researchers and practitioners to facilitate assessment without having to build a new test, develop understandings of new cultural differences, and conduct comparative research. This study used the differential functioning of items and tests (DFIT; Raju, van der Linden, & Fleer, 1995) framework, based on item response theory (IRT), to assess the measurement equivalence between two language versions of the Perceived Leader Integrity Scale (PLIS; Craig & Gustafson, 1998) using samples collected from the United States, New Zealand, and Mexico. The U.S. and New Zealand samples formed the English speaking or US-NZ group and the Mexico sample formed the Spanish speaking group. Two indices of DFIT were used to determine item level (NCDIF) and test level (DTF) inequivalence between the comparison groups. Results showed 17.9% (5 out of 28) of the items to be differentially functioning. No significant DTF was identified at the test level. Post hoc explanations of the items with significant NCDIF in terms of possible cultural and linguistic influences provide information about the possible reasons why the items are functioning differentially (e.g. translation errors, cultural differences, or both). Practical implications of the current study are discussed.