“I can’t search on Google for answers”: validity evidence of a developed computer-based assessment application

Abstract

While learning management systems have been increasingly used as innovative ways to teach students in the post-COVID-19 pandemic, academic dishonesty is still in crisis. The open-access facility has the potential intervention in academic dishonesty. This study reported validity evidence of a developed computer-based assessment application to combat academic dishonesty: 1) measuring the development of high-order-thinking-skill (HOTS) test items; 2) examining the test item's validity integrated into the application; and 3) reporting the program analysis of the anti-cheating application for English for Specific Purposes (ESP) class assessment. This mixed-method study used focus group discussion and reading tests to collect the data. Pearson product-moment and black box analysis were used to measure test-item validity and application. The quantitative result revealed that test items were developed according to authentic journal texts and HOTS question characteristics and were statistically proven valid and reliable. Simplifying the authentic text empirically enabled EFL students to have an easier ESP reading comprehension complexity level. The black box analysis showed that the auto-lock screen feature prevented suspicious cheating access. This study contributed to extending practical and theoretical research on technology-enhanced language learning (TELL) by providing an innovative anti-cheating technology application.