| The purpose of this study was to determine if the Computerized Placement Test (CPT) is a reasonable replacement for the Comparative Guidance and Placement Test (CGP). This study analyzes the relationships between students' placement scores on the CPT and the CGP and their subsequent academic grades in basic, pre-technology, and college-level mathematics. The study uses data from 161 freshman students enrolled at a technical community college in the Spring of 1993. In addition, an investigation was made to determine if there were relationships between students' final grades in basic, pre-technology, and college-level mathematics and demographic profiles based on age, sex, and race.;The findings from this study show that the correlation between the two types of tests is highly significant (p ;Findings from the study indicate that the CGP and CPT would have placed students differently. For an unknown reason, 13 of the students who scored below the 50th percentile on the CGP were permitted to take college-level mathematics courses. If the CPT had been used, only seven students would have been placed in college-level courses. Placing students in college-level mathematics using the CPT was not effective. However, as a testing instrument for the overall assessment of underprepared students, the CPT was a more effective tool in placing students in pre-technology courses rather than in basic courses.;For basic mathematics students, the coefficient of determination of.165 indicates that 16.5% of the variance in grades could be explained by the four predictor variables: CGP mathematics score, sex, race, and age. However, the F-ratio (F = 1.183, p =.34) was not found to be significant at the.05 level, thus indicating that no useful relationship was found for this group. For pre-technology students, the coefficient of determination of.122 indicates that 12.2% of the variance in grades could be explained by the four predictor variables. In this case, the F-ratio (F = 3.347, p =.01) is significant at the.05 level, indicating that a useful predictive relationship was found for the pre-technology students. Of the four variables used as predictors, only age is individually significant. Fur college-level mathematics, the coefficient of determination of.231 indicates that 23.1% of the variance in grades could be explained by three predictor variables: CGP mathematics score, sex, and age. (Race was omitted, as there were no Black students in this group.) However, the F-ratio of (F = 2.698, p =.066) is not significant at the.05 level. indicating that no useful predictive relationship exists for the college-level students. |
