Comparative Study of The Difficulty Index of Items Displayed with The Paper-Based Test and The Computer-Based Testing Paper

The focus of this study is to analyze whether there are differences in the difficulty index of items if the same items are displayed with the Paper Based Test (PBT) model and the Computer-Based Testing (CBT) model. The PBT and CBT models have the same paradigm of measuring estimated abilities but have differences in the context and feel aspects. These differences include the number of items in the range of eyesight, use of tools, how to work on items, basic knowledge needs about computer operations, and habit factors. These differences can affect the results of estimating the ability of test participants. This study uses development methods and quantitative methods. Development methods are used to develop package items and CBT software. Two groups of respondents were used with equal ability to work on the same question package. One group uses the PBT model, and the other group uses CBT so that the response choices are obtained in each test model. The results of the respondent’s answer choices were then analyzed by the ITEMAN software to get the item difficulty index. The item difficulty index of each model was tested statistically using SPSS software, whether there was a significant average difference between the two test models. From the results of the study and analysis in classical theory, it can be concluded that statistically there are differences in the average difficulty index of the items if the same item is displayed with the PBT and CBT models. It was found that the items displayed with the PBT model had a more difficult tendency than when displayed with the CBT model.