AI-TPACK readiness and generative AI integration among mathematics lecturers in Ghana's public colleges of education: predictors of AI-TPACK readiness

Source abstract

Despite growing interest in artificial intelligence (AI) in education, empirical evidence on AI-extended Technological Pedagogical and Content Knowledge (AI-TPACK) readiness and generative AI integration among mathematics lecturers remains limited, particularly in sub-Saharan Africa. This study examined AI-TPACK readiness and generative AI integration among mathematics lecturers in Ghana's public Colleges of Education and investigated whether AI literacy, self-efficacy in AI use, and teaching experience predicted AI-TPACK readiness. A cross-sectional survey design was employed with 214 mathematics lecturers from 49 public Colleges of Education in Ghana. Data were collected using a structured questionnaire measuring AI-TPACK readiness, generative AI integration practices, AI literacy, self-efficacy in AI use, and teaching experience. Descriptive statistics and multiple linear regression were used to analyse the data. The findings showed that overall AI-TPACK readiness was moderate (M = 2.48, SD = 0.71). AI-Technological Knowledge recorded the highest mean (M = 2.62, SD = 0.69), followed by Pedagogical Knowledge (M = 2.41, SD = 0.73), while Mathematics Content Knowledge was low (M = 2.32, SD = 0.75). Generative AI integration across instructional practices was low, with relatively greater use reported for lesson preparation and worked-example generation and limited use for assessment design and student feedback. ChatGPT was the most frequently used generative AI tool among those examined. Multiple regression analysis indicated that the model was statistically significant, F(3, 210) = 54.18, p < .001, explaining 44% of the variance in AI-TPACK readiness (R 2 = .44). AI literacy ( β = .41, p < .001) and self-efficacy in AI use ( β = .33, p < .001) were significant positive predictors, whereas teaching experience was not statistically significant ( β = .08, p = .224). The findings highlight the importance of targeted professional development that strengthens lecturers’ AI literacy and self-efficacy while promoting pedagogically appropriate integration of generative AI into mathematics instruction in Colleges of Education.