The Programming Performance Prophecies: Predicting Student Achievement in a First-Year Introductory Programming Course

Ensuring student success in first-year introductory programming courses presents a unique challenge when considering the diversity of student educational backgrounds. Some students enter college having programmed for several years, while others have had little to no exposure to basic programming concepts. In addition, students frequently possess a wide range of skills in areas of study that are related to programming. When students with widely differing skill sets are enrolled in the same introductory programming course, a competency imbalance is immediately established that negatively impacts the course and leads to high attrition rates among students with less experience. In order to create a more equitable experience and to ensure that students are placed in a course whose difficulty is commensurate with their abilities, we present a methodology to predict student performance in first-year introductory programming courses. Our technique requires that all students take an online exam at the beginning of the term consisting of a variety of algorithmic, logic, and math-based questions that are not tied to a specific programming language. From a sample of over 800 students, we show a significant correlation between the performance on our exam and overall course performance, thereby giving us the ability to properly accommodate students based upon our expectation of their performance in the course and to minimize any potential imbalances that may occur.