Enhancing Assessment through AI: A Comparative Study of Lecturers’ Views on Human and AI-Based Feedback in Nigerian University Settings
Keywords:
Artificial Intelligence (AI) Feedback, Lecturers’ Perception, Assessment Practices, Higher Education, Hybrid Feedback ModelAbstract
This study comparatively examined lecturers' perceptions of AI-based versus human feedback in
Nigerian university settings, focusing on attitudes, perceived effectiveness, and integration
preferences at Nnamdi Azikiwe University, Awka. The study was guided by three research questions
and two corresponding hypotheses were tested. The population comprised all 3,200 academic staff of
Nnamdi Azikiwe University, Awka, from which a sample size of 357 lecturers was selected using
Taro Yamane's formula at 95% confidence level. Adopting a descriptive survey design, data were
collected using the instrument titled "Lecturers' Perception of AI-Based Feedback Questionnaire
(LPAIFQ)" with a reliability coefficient of 0.864, and validated by three experts, two in Educational
Measurement, Evaluation and Research and one in Educational Technology. Quantitative data were
analyzed using descriptive statistics of frequencies, percentages, means and standard deviation and
inferential statistics including independent t-test and ANOVA using SPSS version 26.0. Findings
reveal nuanced perceptions: while 68% of lecturers appreciate the speed and consistency of AI
generated feedback, concerns persist regarding its contextual accuracy, emotional sensitivity, and
adaptability to individual student needs, with significant differences in perception based on years of
teaching experience (F=12.34, p<0.05) and disciplinary affiliation (t=4.67, p<0.05). The study
concluded that disciplinary differences and prior experience with AI tools emerged as significant
factors influencing attitudes toward AI in assessment, while 72% preferred hybrid models combining
human judgment with AI efficiency. It is recommended that institutions should develop balanced
integration approaches with culturally-sensitive AI feedback systems and adequate training programs
to enhance feedback practices in AI-enhanced learning environments.
