The Schommer’s epistemological beliefs questionnaire is regularly used in studies on epistemological beliefs. However, many researchers pose some questions about the appropriateness of this instrument. The present study has been taken with the goal of investigating the reliability and factor structure ...
Read More
The Schommer’s epistemological beliefs questionnaire is regularly used in studies on epistemological beliefs. However, many researchers pose some questions about the appropriateness of this instrument. The present study has been taken with the goal of investigating the reliability and factor structure of schommer’s epistemological beliefs questionnaire (1990). To do this, the questionnaire was translated into Persian (Farsi). Then, it was retranslated into English and compared with the original version to find and correct the gaps between the translated version and the original version. After the pilot study and resolving the possible mistakes, the final questionnaire was implemented for sample group consisting of 518 students studying in different fields in Tabriz Payame Noor University (302 students from humanities and 216 students from basic sciences).Cronbach Alpha calculation showed that none of reliability coefficients of sub-scales are acceptable. So, factor analysis was applied in this study to examine the construct validity and to reach an optimal factor structure. Before factor analysis of data, internal consistency was calculated for 63 questions. 27 questions due to under .10 and negative correlation between item-total score test were deleted from subsequent analyses. Using Principal Component's method, factor analysis was performed on 36 remaining questions. Investigatng of Scree plot of primary analysis showed two-factor solution. In order to achieve a two-factor structure, the analysis of Principal Components was done using Promax rotation method. 20 questions, due to factor load below .35 or due to meaningful and equal load on more than one factor in several rotations was deleted. Final analysis showed loading of the remained 16 questions on two factors. These factors were nominated according to content of the questions covered by simple/absolute knowledge and fast/fixed learning.Calculation of Cronbach Alpha coefficient showed that the reliability of two obtained factors is acceptable
