Main Article Content
Abstract
Low engagement and poor retention of scientific knowledge persist among students in Nigerian colleges of education, particularly in contexts where large classes, lecturer shortages, insecurity, and limited individual support constrain learning. This study examined the effect of AI-based chatbot tutoring on student engagement and retention in Biology among NCE students in selected federal colleges of education in North-East Nigeria. Four research questions and four null hypotheses guided the study. A quasi-experimental, non-equivalent pretest-posttest control group design with a delayed posttest was adopted. The sample comprised 152 NCE II Biology students in four intact classes drawn from two federal colleges of education in Adamawa and Gombe States. The experimental group (n = 78) received six weeks of conventional instruction supplemented by a curriculum-aligned AI chatbot tutor, whereas the control group (n = 74) received conventional instruction. Data were collected with the Student Engagement Scale (α = .86) and the Biology Retention Test (KR-20 = .82), administered as pretest, posttest, and delayed posttest. Mean and standard deviation were used to answer the research questions, while analysis of covariance (ANCOVA) was used to test the hypotheses at the .05 level. Students exposed to chatbot tutoring had significantly higher engagement, F(1, 149) = 231.46, p < .001, partial η² = .608, and retention, F(1, 149) = 118.37, p < .001, partial η² = .443, than students in the control group. Gender had no significant effect on engagement or retention among students exposed to chatbot tutoring. The study recommends integrating curriculum aligned chatbot tutoring into Biology instruction as a supplement to lecturer-led teaching, alongside lecturer training, reliable connectivity, power supply, and safeguards for academic integrity and student data privacy.