An analysis of student placement into College Algebra

The purpose of this study was to analyze placement practices into College Algebra and to determine the effect of these practices on student success rates in College Algebra. This ex post facto study was conducted at a private, urban university with an enrollment of 3700 students and included all students who took College Algebra from Fall Semester 1993 through Summer Semester 1998. The 657 students were divided into four groups: students who did not take the placement test before taking College Algebra; students who took the placement test, were recommended to take College Algebra, and did; students who took the placement test, were not recommended to take College Algebra, but did anyway; and students who took the placement test, were not recommended to take College Algebra, took Intermediate Algebra first, and then took College Algebra. The percentages of students who took the mathematics placement test and who took the recommended course; the success rates of the four groups; the correlation between the placement test scores and the College Algebra grades; the optimal cut-score on the placement test; the prediction model of success; and the variation in the prediction model from three variables, student gender, the semester of enrollment for College Algebra, and the number of semesters between taking the placement test and enrolling in College Algebra were determined. The study revealed the following: (1) Students followed placement recommendations. (2) Students who took the placement test and followed the recommendations were more successful in College Algebra than students who did not. (3) Students who took Intermediate Algebra were not as successful in College Algebra as the students who did not take Intermediate Algebra. (4) Placement test scores and College Algebra grades were significantly correlated. (5) Raising the cut-score can help to lower the number of students who are recommended for College Algebra but are unsuccessful. (6) Predicted grades and actual grades were significantly correlated. (7) Stepwise multiple regression analysis selected the placement test score and student gender but excluded the semester of enrollment and the number of semesters between the placement test and College Algebra. These results indicate that assessment and placement should be continued.