Analysis of linear congruent methods and multiplicative random number generator in computer-based test

Abstract

This research focuses on the implementation of computer-based test exams in high schools which face the problem of not having differences in exam questions which results in weak security and validity of exam results. Therefore, a randomization method is needed to overcome this problem. The method used in randomization is linear congruent methods and multiplicative random number generator. There are 101 random questions, but only 40 questions are displayed for each student with a reference value of Xn and C, 2 will be added for each package of exam questions and to avoid question code=0, the calculation results will be added 1. The linear congruent methods (LCM) results achieve 100% accuracy, while the method The multiplicative random number generator (MRNG) only achieved 62.5% accuracy in randomizing the exam questions. This accuracy comparison highlights the difference in the ability of the two methods to generate random permutations of test item packages. LCM randomization accuracy ensures that each student will receive a different set of test questions in a consistent manner. However, the low accuracy of randomization using MRNG indicates a weakness in generating permutations of exam question packages. The results of this study show that the LCM method is better than the MRNG in conducting exams.