Abstract:
This paper reports the first part of a broader study examining potential cultural and disciplinary differences in self-directed learning among tertiary students. Specifically, the research compares student self- efficacy (Greenglass, Schwarzer & Taubert., 1999), motivation (Chan, 1994; Midgley et al, 1998), and reported self-regulatory and volitional behaviour (Cantwell & Moore, 1996; MCCann & Garcia, 1999) from a cultural perspective (Thailand and Australia) and from a disciplinary perspective (nursing and psychology). As the research is based on instrumentation developed in western cultures, the first stage of the project, and the subject of this report, is the validation of the instruments for use among Thai tertiary students. Methodology employed for validation was based upon the cross-cultural translation technique developed by Vallrand (1989), later confirmed by Banville, Desrosiersad refined by a bilingual committee.
Content validity was assessed using a bilingual group of Thai students who responded to both English and Thai versions. Consistency of item responses was assessed using pair t-test and Pearson correlations. Among the 134 items, seven items were found to be marginally less reliable in translation. Construct validity was then assessed using a group of 150 Thai university students who responded to Thai version only. These responses were then analysed using single factor principal component analysis. Data indicated only five items with low factor loadings on the prescribed scale. Overall, the validation process indicated an acceptable level of fit between the English and Thai versions of the questionnaires.
Content validity was assessed using a bilingual group of Thai students who responded to both English and Thai versions. Consistency of item responses was assessed using pair t-test and Pearson correlations. Among the 134 items, seven items were found to be marginally less reliable in translation. Construct validity was then assessed using a group of 150 Thai university students who responded to Thai version only. These responses were then analysed using single factor principal component analysis. Data indicated only five items with low factor loadings on the prescribed scale. Overall, the validation process indicated an acceptable level of fit between the English and Thai versions of the questionnaires.