Measurement Invariance of Perceived Learning of Business Statistics through E-Learning Platform

The measurement of invariance is a stable relationship between items and traits that have the same psychometric properties across the groups. Measurement invariance is crucial in statistics where the responses are based on students' perceived learning of Business Statistics. In this study, this issue relates to the students' perceived learning of Business Statistics across different profiles who deal with different methods of solving statistical problems. This requires students to have multiple skills such as calculation and computational skills, thinking skills and analytical skills. One of the most important topics perceived by students to be difficult to grasp is hypothesis testing, which became more challenging when the Business Statistics learning is conducted using e-learning platform as students are required to learn statistics in a virtual environment. This study served two purposes. Firstly, to determine the invariance across different profiles of students and secondly, to investigate the students' perceived learning of Business Statistics through e-learning platform. One hundred and fourteen (114) undergraduate students enrolled in the Statistics for Business and Social Sciences (STA404) course have participated in the study. Data was collected from the Accountancy and Hotel Management students using a survey instrument which was developed for the purpose of obtaining the initial perception of students' learning of Business Statistics through e-learning platform. The data was analysed using software Winsteps 3.90 based on the Rasch measurement model. The analysis focused on structural information and differential item functioning. The results showed that male students have positive initial perceptions towards learning Business Statistics through e-learning. Rasch Measurement Model revealed measurement invariance existed across both group of students' gender and their faculties (Accountancy and Hotel Management) through differential item functioning.