Abstract: The educational institutions can use educational data mining to analysis student’s performance which can help the institution in identifying the student fail or dropouts. Classification is popular data mining technique that have been widely implemented to predict student performance. There are many attributes and methods that can be used to predict student’s performance. This preliminary study presents study review related with this current topic to reveal the most widely used classification technique and the attributes that commonly used to predict student performance. Qualitative content analysis is used as the research method. It can be concluded that the most widely used classification methods based on the literature reviews are decision tree and Bayesian network and method that outperform the other based on the comparison analysis from several studies is decision tree with CART algorithm while the most widely used attributes can be categorized into four categories, student personal information, family information, pre-university characteristics and university features.