Predictive Validation on Utilization of AI-Awareness Scale in Educational Assessment (AI-ASiEA) by lecturers: Using a Classical Modelling Approach

Abstract

This study investigated the predictive validation on utilization of AI-Awareness Scale in Educational Assessment (AI-ASiEA) among lecturers in South-Eastern Nigeria using classical modelling. Three research questions and two null hypotheses guided the study’s investigation of the internal consistency of the scale, differences in the perceived AI-supported learning and perceived learning outcomes, and the predictive relationship between AI awareness and these outcomes. A correlational research design was adopted, with a population of lecturers across five states and a sample of 300 selected through convenience sampling. Data were analyzed using Cronbach’s alpha, mean, standard deviation, Pearson correlation, and multiple regression. Findings revealed that the AI-ASiEA demonstrated acceptable internal consistency and that lecturers showed moderate awareness of AI assessment tools such as Gradescope and Turnitin. The results further revealed a strong positive predictive relationship between AI awareness in educational assessment and perceived AI-supported learning, and a moderate significant relationship with perceived learning outcomes. This led to the rejection of both null hypotheses. The study concludes that AI awareness is a significant predictor of lecturers’ perceptions of AI-driven learning effectiveness. Among all recommendations, the universities should implement structured professional development programs to deepen lecturers’ AI literacy in assessment practices.