Authentic Assessment: How Instructors Perceive Authentic in the Assessment of Higher Education Institutes in the Era of AI

Abstract

Generative artificial intelligence (GenAI) advancements have brought up chances and issues with authentic assessment in higher education. This research aimed to identify the instructors' perceptions concerning the impact of GenAI on students' examinations, assessments, and authenticity of assessment of students' learning. An exploratory descriptive quantitative methodology was employed, which involved an online questionnaire administered to 200 higher education instructors representing various academic positions, fields, and years of teaching experience. The questionnaire instrument was focused on two dimensions: perceived benefits and perceived risks of GenAI in authentic assessment. Data analysis included descriptive statistics, reliability test, independent-samples t-test, one-way ANOVA, and Pearson correlation analysis. The findings revealed quite high levels of both perceived benefits (M = 3.66, SD = 0.57) and risks (M = 4.00, SD = 0.47) related to GenAI. The most significant risks included difficulty of distinguishing the students' and GenAI generated work, plagiarism issues, possibilities for shortcutting and validity threats. On the other hand, the potential benefits of GenAI included its ability to help in matching assessment with industry standards, provide prompt feedback, decrease grading load and promote creativity. No statistically significant differences were found by gender, academic position, teaching experience, or academic discipline. Prior usage of GenAI had a significant impact on the perception of benefits, while the prior use of the AI detection software had an impact on the perception of risks.