The influence of knowledge and skills on self-efficacy in using ChatGPT for mathematics instructionTangkui, Rayner (Avtor) knowledgeskillsself-efficacyChatGPTmathematics instructionThis study examined the influence of knowledge and skills on mathematics teachers' self-efficacy in using ChatGPT for mathematics instruction. A total of seventy-three primary school mathematics teachers in Tambunan, Sabah, were selected through stratified sampling. Data were collected using a validated questionnaire measuring knowledge, skills, and self-efficacy and analysed using descriptive statistics, Pearson correlation, and multiple linear regression. Results showed high levels of knowledge but moderate levels of skills and self-efficacy. Significant positive correlations existed between knowledge and self-efficacy and between skills and self-efficacy. Regression findings showed both knowledge and skills significantly predicted self-efficacy, with skills being the stronger predictor.Univerza v Mariboru, Univerzitetna založba20262026-09-15 10:41:53Znanstveno delo100359UDK: 37.091.3:51:004.8COBISS_ID: 290783491DOI: 10.18690/rei.5419ISSN pri članku: 1855-4431sl