This study identifies and evaluates the most frequently used research methods and factors influencing academic performance, based on 95 studies published after 2012. The research focuses on academic performance and university dropout rates, using a combination of empirical studies and meta-analyses. The most commonly used methods include Educational Data Mining (EDM) algorithms such as decision tree, logistic regression, and neural networks, as well as Structural Equation Modelling (SEM). The predictive power of these methods varies, with Support Vector Machines, Multilayer Perceptron, and Naïve Bayes algorithms showing the highest accuracy. Factors influencing academic performance include grade point average (GPA), obtained credits (ECTS), and gender, which are consistent predictors. However, these factors are mediated by student factors such as intrinsic motivation, self-regulated learning strategies, and self-efficacy, as well as throughput factors like work, finances, and academic engagement. Contradictory results were found regarding age and family background. The study also highlights the importance of psychological and socio-economic factors in academic success. The research emphasizes the need for a comprehensive model that includes input, throughput, and output factors. The findings suggest that GPA, ECTS, and gender are the most significant predictors of academic performance, while factors such as intrinsic motivation, self-efficacy, and conscientiousness also play a crucial role. The study concludes that academic performance is a complex phenomenon influenced by various factors, and that a universal definition or approach is not possible due to the diversity of contexts and variables involved. The research also notes the limitations of the study, including the lack of representative samples and the challenges of comparing studies across different cultural and political contexts.This study identifies and evaluates the most frequently used research methods and factors influencing academic performance, based on 95 studies published after 2012. The research focuses on academic performance and university dropout rates, using a combination of empirical studies and meta-analyses. The most commonly used methods include Educational Data Mining (EDM) algorithms such as decision tree, logistic regression, and neural networks, as well as Structural Equation Modelling (SEM). The predictive power of these methods varies, with Support Vector Machines, Multilayer Perceptron, and Naïve Bayes algorithms showing the highest accuracy. Factors influencing academic performance include grade point average (GPA), obtained credits (ECTS), and gender, which are consistent predictors. However, these factors are mediated by student factors such as intrinsic motivation, self-regulated learning strategies, and self-efficacy, as well as throughput factors like work, finances, and academic engagement. Contradictory results were found regarding age and family background. The study also highlights the importance of psychological and socio-economic factors in academic success. The research emphasizes the need for a comprehensive model that includes input, throughput, and output factors. The findings suggest that GPA, ECTS, and gender are the most significant predictors of academic performance, while factors such as intrinsic motivation, self-efficacy, and conscientiousness also play a crucial role. The study concludes that academic performance is a complex phenomenon influenced by various factors, and that a universal definition or approach is not possible due to the diversity of contexts and variables involved. The research also notes the limitations of the study, including the lack of representative samples and the challenges of comparing studies across different cultural and political contexts.