This study aims to measure student cognitive and metacognitive perceptions regarding computer use in university. This study addresses three key questions: (1) how does the model fit the key construct of metacognitive and self-efficacy beliefs; (2) to what degree do independent variables predict dependent factor outcomes; and (3) how can understanding student metacognitive and self-efficacy beliefs inform ICT policy in education? This representative model uses four latent factors to measure student perceptions: self-efficacy (SE), computer anxiety (CA), vicarious experiences (VE), and belief of benefits (BB). This study uses a 16-item survey model to test a sample of 602 students from a university in Japan. Quantitative data were analysed using a confirmatory factor analysis to test the validity of dependent latent factors, and a multiple regression analysis to identify independent variables that can predict dependent variable outcomes. Results identify statistically significant predictors of outcome related to two of the dependent variables (self-efficacy and computer anxiety) concerning the demographic groups: (1) gender; (2) computer use age (the age when participants begin using computers); (3) computer use based on hours per day; (4) ownership of computer devices; and (5) location of computer use. Findings conclude that the tested model gives substantial insights into the cognitive and metacognitive processes of students, which can meaningfully contribute to the development of ICT policies in schools.