Quasi-Experiment: Postgraduate Students’ Class Engagement in Various Online Learning Contexts when Taking Privacy Issues to Incorporate with Artificial Intelligence Applications

Abstract

<p>Artificial intelligence (AI) has enhanced students’ learning especially by personalizing content to students’ individual needs and automating the learning process. However, AI has also raised serious privacy issues which hindered the adoption of the technology. Therefore, this research was an attempt to contribute towards attaining the goal of investigating the above problem in terms of students’ class engagement in various online learning contexts. A quantitative (quasi-experimental design) study with pre-test and post-test was conducted, and data were collected from 99 postgraduate students from the same local university in Hong Kong. The 99 students were divided into three groups randomly with each group consisting of 33 students. One control group learned with traditional online learning without any AI applications, two experimental groups learned with AI applications in the class, and one with the privacy issues taken into consideration. The pre-test and post-test were conducted based on The Online Student Engagement Scale (OSE) across 4 main factors (skills, emotion, participation, and performance) to assess the students’ behavioral changes. Descriptive analysis and ANCOVA analysis were used to analyze the data. This study discovered that the students’ online learning with the AI applications applied for 4 weeks improved their class engagement, and the group with privacy consent taken into consideration exhibited a greater improvement in students’ class engagement than the groups without privacy consent when implementing the AI application for online learning. The second stage of this project has been exploring the reasons behind the research results via a qualitative approach (interviews). Further content analysis was recommended to investigate this issue deeper. </p>