IJSTIP Published Article Details

Abstract This study determined the predictors of research self-efficacy among STEM graduates in a science investigatory project class and examined the levels at which these predictors occur, their influence on STEM graduates’ research self-efficacy, to develop a model for facilitating research self-efficacy. Utilizing a descriptive-correlational research design, the study employed Partial Least Squares - Structural Equation Modeling (PLS-SEM) with 241 STEM graduates of Surigao del Norte National High School who responded to adapted and researcher-made survey questionnaires via Google Forms. The findings revealed that the highest level of research self-efficacy was in Research Ethics and Integrity (M=3.55), while the lowest was in Data Analysis (M=3.38). Regarding the extent of factors that predict research self-efficacy, Research Mentoring Experience recorded the highest mean (M=3.63), while Research Training Environment scored the lowest (M=3.31). Significant intercorrelations were observed between all factors of research self-efficacy, with correlation coefficients ranging from moderate to weak (r = 0.16 to 0.78). Crucially, although the level of research self-efficacy in science investigatory project were all positively correlated with all the factors that predict research self-efficacy, multiple linear regression analysis only identified Investigative Interest (β=0.41, p<0.001) and Research Training Environment (β=0.12, p=0.019) as the only significant predictors of research self-efficacy, while Research Anxiety, Working Alliance, and Research Mentoring Experience were non-significant predictors. Further, this was translated to the findings of the structural equation modeling and validated Investigative Interest and Research Training Environment as the only factors that significantly and positively influence research self-efficacy. The study contributes to the SIRSE (Social Cognitive Processes, Investigative Interest, Research Training Environment, Self-Efficacy) Model, providing an empirically derived research self-efficacy model mediated by Social Cognitive Theory. Thus, the study recommends the implementation of hands-on data analysis workshops and establishing a dedicated research budget to provide students with research materials. Ultimately, the adoption of the SIRSE Model is proposed to systematically recalibrate DepEd’s research curriculum for fostering confident researchers.