Winter 1993 - Volume 1 Number 2
Predicting the Academic Success of Re-Entry College Students From Placement Test Scores: A Multiple Regression Analysis
Many colleges and universities administer admission tests to determine applicants' chances of academic accomplishment. The American College Test, the Test of English as Foreign Language, the Scholastic Aptitudes Test, the Nelson-Denny Reading Test (NDRT), and the Canadian Achievement Test are some widely used admission criteria. They are said to enhance college predictions about their applicants' possibilities of success.
A number of studies have been conducted to determine whether placement test scores along with selected student characteristics are good predictors of academic achievement. These studies, though differing in focus, design, sampling procedures and measurement techniques, have focused mostly on traditional students. Moreover, only limited analyses have been conducted to determine whether the Nelson-Denny Test which is widely used for identifying and placing adult learners in academic programs is a good academic predictor, especially among those whose education has been interrupted for a long time.
The purpose of this study was to test the general predictive value of standardized admissions test scores and selected student characteristics on the academic success of adult learners enrolled in a first-year social work program. More importantly, it was to explore whether admission test scores on the Nelson-Denny Reading Test could be used as an early warning system for students' academic adjustment.Data and Measurement
The data analyzed in this study came from the Student Information System (SIS) and the students entrance test scores in vocabulary, comprehension, and total scores on the Nelson-Denny Test. Once obtained, the data were recorded by code number for each student to ensure anonymity. The following input variables were retrieved from the student profile: age, sex, marital status, years since last attending school before enrolling in the college, grade point average (GPA), number of high school credits, previous attendance at the college, and ethnic background. Students' entrance scores on the Nelson-Denny Reading Test were obtained from the Admissions Office. Statistical analysis was accomplished with the computer package SPSS (windows, release 6.0).
The central dependent variable in this research, academic success, was measured by cumulative GPA. Exact values ranged from 1.69 to 3.40 on a four point scale. The independent variables were allocated to non-institutional and institutional categories. The non-institutional variables were: age, gender, ethnicity, marital status, number of children, number of high school credits, and number of years since last attending school before entering college. The main institutional variable was the Nelson-Denny Reading Test. The NRDT is composed of two subtests, comprehension and vocabulary.
The sample consisted of all 60 students that enrolled in the first year of a two-year social work diploma program at the Alberta Vocational College (AVC), Lac La Biche, during 1990, 1991 and 1992. Of the group, 14 (23%) were males and 46 (77%) were females. The ages ranged from 20 to 50 years, with 20 (33%) between 20-29 years, 26 (43%) between 30-39 years, and 14 (24%) 40 years and older. With regard to marital status, 24 (40%) were single, 17 (28%) were married, and 19 (32%) were either divorced or separated. Of the 41 students with dependents, 26 (63%) had one or two dependents, while the remaining 15 (37%) had three or more dependents. The sample consisted of 21 (35%) aboriginal and 39 (65%) non-aboriginal students. On average, these students had been out of school for 16 years before enrolling at the college. A total of 45 (75%) attended the college in some capacity before enrolling in the first-year social work program. GPAs were distributed as follows: under 2.0 = 33%; 2.0-2.49 = 15%; 2.5-2.99 = 29%; 3.0 and above = 23%.Correlation Analysis
Correlations were run in order to reveal the relationships among dependent and independent variables. Table I shows measures of variables, means and standard deviations for variables in the regression model for nine independent and the dependent variable (GPA). Table II shows a negative correlation (r=.-55) between the number of years students have been out of school before attending AVC and their number of high school credits. The Table displays low positive correlation between age and vocabulary scores on the NDRT (r=.22), total scores (r=.08), and GPA (r=.24). The negative correlation (r=.-53) between age and number of high school credits is moderately high, suggesting that as age increases, reported number of high school credits decreases.
According to Table II, both Nelson-Denny Total Score and scores on the Nelson-Denny Vocabulary subsets are positively correlated with academic success. The positive correlations between Nelson-Denny Total Scores (r=.51), Nelson-Denny Vocabulary Scores (r=.54) and GPA, indicate that students who score well on the test, especially in the vocabulary subset, are more likely to succeed in the program.
The strong positive correlation between the Nelson-Denny Total Score and the Nelson-Denny Vocabulary Score is noteworthy. The high positive correlation (r=.93) between Vocabulary Scores and the Nelson-Denny Total Scores (upon which student selection is based) indicates that students who score very high on the vocabulary section of the NDRT are more likely to receive a very high overall score. What is even more noteworthy is that gender, ethnicity, marital status, previous enrollment at the college, and high school English scores had no significant effect on GPA and, hence, were not reported.Regression Analysis
In order to decide which of the independent variables are most important in predicting academic success among first-year social work students, a step-wise multiple regression analysis, using students' GPAs as a dependent variable, was carried out. In this procedure, entry of an additional independent variable is based on the value of the partial correlation coefficient. Variables with the highest partial are entered first. Additional predictor variables are entered, or are removed if they are no longer useful, until all the variables are in the equation. Only significant predictors are included at .05. As Table III shows, the results of the step-wise multiple regression analysis reveal that of all of the predictor variables, only vocabulary scores on the Nelson-Denny are significant predictors of academic success within the sample.
These findings are compatible with previous research. Age, for instance, has been found to be a poor predictor of academic success among adult students. Indeed, while examining the factors associated with academic achievement, researchers have concluded that information collected at admission was of little value in making predictions about individual students.Summary and Conclusion
This study sought to determine how accurately student academic success may be predicted from admission test scores and selected student characteristics. The results demonstrate that the total score on Nelson-Denny and scores on the comprehension subset of the test correlate moderately with academic success.
Multiple regression analysis shows that the vocabulary score on the NDRT is a significant predictor of academic achievement. Other variables such as age, gender, number of years since attending school and so on, did not predict academic success of social work students. These findings indicate that vocabulary is the single most important predictor of comprehension within the sample.
Based on the findings of this study, NDRT results should not be used strictly for admission purposes. They should also be used by teachers and counselors for handling individual differences in guiding adult learners. The data revealed a moderate correlation (r=.38) between vocabulary and comprehension scores on the NDRT, suggesting that a high score in vocabulary should translate into a high score in comprehension (word power). Remedial attention to low scores in vocabulary would seem in order in planning remedial instruction such as vocabulary building for students with marginal scores on the vocabulary section of the NDRT. This effort will result in high student retention.
The data collection strategy employed in this study limits the examination of the inclusion of other institutional variables such as class size or non-institutional variables such as family stability. Future research should take these into consideration.
Akin (Bob) Adebayo is Research Coordinator at the Alberta Vocational College in Lac La Biche. This article is abridged from an original submission. The full text with complete references is available from the author.
• The views expressed by the authors are those of the authors and do not necessarily reflect those of The College Quarterly or of Seneca College.
Copyright © 1993 - The College Quarterly, Seneca College of Applied Arts and Technology