Identifying students who may experience difficulty in an introductory computer science course

A high percentage of the students in introductory computer science (CS1) courses either drop out or fail to perform at an acceptable level. In past years when students were flocking to computer science, student failure rales were not viewed as a problem. Now that student interest in computer science has declined significantly, departments are faced with the problem of how to increase retention rates without compromising the quality of the course. This can be partially accomplished by identifying those students who may experience difficulty and then closely monitoring and assisting these students. This study is concerned with the question: Is it possible to develop a simple and reasonably effective procedure for identifying those students who may experience difficulty in the CS1 course? Prior studies have attempted to predict success In CS1 courses with mixed results. Of these studies, the one most closely related to the present study was conducted by Fowler and Glorefeld [4]. In this study, college GPA, number of math courses, SAT math score, and age were used to develop a model which correctly classified 81 per cent of the students in CS1 courses into two categories, those who received a grade of A or B, and those who received a grade of C, D, or F. Thirty students from two sections of a CS1 course using Pascal at a comprehensive university participated in the present study during the 1988 Fall semester. The final grade was used as a measure of success in the course and was used to classify the students into two groups according to the degree of difficulty that they experienced in the course. Those who received a final grade of Cor lower were placed in a "high difficulty" group while those who received a final grade of C or higher were placed in a "low difficulty" group. The predictors consisted of classification (CLASS), reason for taking the course (UHY), and four tests covering mathematics background (MATH), logical