NotesFAQContact Us
Search Tips
ERIC Number: ED081807
Record Type: Non-Journal
Publication Date: 1971-Jun-25
Pages: 17
Abstractor: N/A
Reference Count: N/A
The Utility of a Standard Composite for Predicting Freshman Average Grade in Eight Liberal Arts Colleges.
Wilson, Kenneth M.
To determine whether it is necessary for eight College Research Center (CRC) member colleges to use 16 distinct sets of weight in combining four admissions tests scores into predictive composities, a standard set of weights were developed and tested for effectiveness in predicting Freshman Average Grade (FAG). The basic sample in which weights were derived were women entering CRC-member colleges in Fall 1965; complete data were available for 2,523. Within each college, students with complete data were ordered from high to low with regard to FAG, and those in the upper and lower 27% were assigned to High Achieveing and Low Achieving groups, respectively. A multiple discriminant analysis resulted in four discriminant functions: the principal discriminant function exhausted approximately 91% of the discriminating power provided by the four test scores, with raw-score weighting as follows: .22 (SAT-V) + .13 (SAT-M) + .62 (Rank) + .74 (Ach Av); the second function accounted for an additional 6%, with raw-score weightings as follows: .44 (SAT-V) + .23 (SAT-M) + .20 (Rank) - .84 (Ach Av). Means and standard deviations of the two discriminant scores for each of the 16 college/level of achievement scores are tabulated. Results indicate that, within each college/school-type subgroup, a composite score based on application of a standard set of weights to the four admission scores correlates as highly with FAG as composites based on group-specific weights for the respective subgroups. (DB)
Publication Type: N/A
Education Level: N/A
Audience: N/A
Language: N/A
Sponsor: N/A
Authoring Institution: College Research Center, Princeton, NJ.