Determinants of Electronic Examination Success: An Application of Robust Structural Equation Modeling

Abstract

In this paper, robust procedures for estimating path coefficients in the modeling framework of structural equations are proposed based on robust correlation coefficients. These are two estimates: robust correlation coefficients and Pearson correlation coefficients. The traditional approach performs better than the robust approach using real data of the researcher’s collected through the questionnaire. To find out factors contributing to success in electronic tests, students were sampled using the questionnaire from postgraduate applicants at the University of Mosul. Following the collection of data, 205 valid responses were obtained that contained six independent variables that represented factors which affect electronic tests. The proposed biweight midcorrelation method produced more stable path estimates and improved model fit in the presence of outliers. The results indicate that self-efficacy, quality of use, and user satisfaction are the most influential factors affecting the success of electronic examinations.