Analisis faktor, kesahan dan keboleh percayaan instrumen inventori Stereotaip Multietnik Malaysia (ISMEM)

Abstract

This study aimed to test the validity and reliability of the Inventory of Malaysian Multi Ethnic Stereotypes (ISMEM) instrument developed by the researchers. The qualitative phase was using in-depth interviews and focus group discussions. The employment of qualitative method was to explore the concept of stereotypes among the ethnic groups. Quantitative phase involves the formation of items by themed and categorised the items that were built from the qualitative phase. A total of 1,000 university students from various ethnic groups in Malaysia were involved in the quantitative phase. Exploratory factor analysis on ISMEM has formed two main components labelled as Antecedents Stereotypes and Phenomenon Stereotypes. The psychometric assessment of ISMEM involves i) testing the construct validity between the Antecedents Stereotypes and the Phenomenon Stereotypes; (ii) determining the convergent validity between the Antecedents Stereotypes and the Phenomenon Stereotypes; (iii) testing the reliability coefficient of the Antecedents Stereotypes and the Phenomenon Stereotypes, and (iv) examining the item quality of the Antecedents Stereotypes and the Phenomenon Stereotypes. An assessment of the suitability of data for factor analysis was conducted before the psychometric assessment was carried out. The correlation (r≥.30) between the scores of the Antecedents Stereotypes and the Phenomenon Stereotypes suggested a reasonable factoring. Construct validity test such as the Test Kaiser-Meyer-Olkin, the Bartlettt's Test of Sphereity, the communality, the load factor, the Eigen value and the percentage of variance factor in both components of the Antecedents Stereotypes and the Phenomenon Stereotypes are reasonable. The reliability value of Cronbach Alpha for the Antecedents Stereotypes (α = .897-.933) and the Phenomenon Stereotypes are well accepted (α = .706 -.821). Therefore, ISMEM is a reliable and valid measure. However, further studies need to be conducted in other groups of sample to further validate the instrument.