Kesahan dan kebolehpercayaan soal selidik gaya e-pembelajaran (eLSE) versi 8.1 menggunakan model pengukuran rasch

Abstract

The e-Learning Style Questionnaire (eLSE) version 8.1 is developed to measure e-learning style based on the Asian culture and learning environment using the hybrid method involving face-to-face, online and self-learning method. The study was conducted to produce empirical data on the reliability and validity of the eLSE questionnaire version 8.1 using the Rasch Measurement Model. A survey was conducted in the Faculty of Education towards 302 third year students from the TESL, Science, Sport and Recreation, Special Education and Islamic and Al-Quran Education. Out of the 302 questionnaires given only 295 were returned for analysis. The reliability and validity of the eLSE questionnaire were tested with the Rasch Measurement Model using the Winstep version 3.68.2 program. The unidimensionality test conducted showed standardized residual variance of 29.6% with the biggest secondary dimension in the first contrast amounting to 12.1%. The Rasch analysis showed person reliability index of 0.83 and item reliability index of 0.96. In terms of item polarity, each item was able to contribute to e-learning style measurement since the PTMEA CORR exceeded 0.30, in specific between 0.31 and 0.57. Item fit analysis showed that none of the items needed to be dropped since infit mean square values are between 0.82 and 1.29, and the outfit mean square values are between 0.80 and 1.40. Items in the questionnaires are ordered in a continuum of increasing intensity for the measurement of the e-Learning style construct. This shows the validity of the constructs in eLSE. The Rasch Measurement Model shows that the student’s e-Learning style questionnaire has a high reliability and validity to be used for measuring e-Learning style