NotesFAQContact Us
Search Tips
ERIC Number: ED336048
Record Type: Non-Journal
Publication Date: 1991-May
Pages: 20
Abstractor: N/A
Reference Count: N/A
A Technique for Producing a Double-Entry Expectancy Nomograph from Observed Proportions without Distributional Assumptions. AIR 1991 Annual Forum Paper.
Saupe, Joe L.
This paper describes a procedure for smoothing the proportions of a double-entry expectancy table which might be used in higher education admissions advising or other functions. The product of the smoothing procedure is a nomograph, a medium for displaying expectancies which provides pairs of individual values for two predictor variables rather than pairs of ranges, as in the usual expectancy table. The paper demonstrates the procedure and the resultant nomograph, using the high school class percentile ranks, achievement test composite scores, and freshman year grade point averages of first-time freshmen of five consective entering classes at the University of Missouri, Columbia. The number of students available for developing the nomograph was 12,835, a large number. If it could be shown that the nomograph solution was satisfactory for that sample size, the question, "For what smaller sample sizes would it produce acceptable results?" could be raised. In a further step the study investigated the effects of varying sample size and minimum group size and found that samples smaller that 12,000 produce satisfactory expectancy nomographs but that the smallest sample possible may lie between 1,000 and 3,000. However, altering the minimum group size had little effect on the nomograph curves. The study concludes that valid expectancy nomographs can be produced from large samples using minimum group sizes of 50 and that these nomographs may be easier to use than the expectancy tables for which they are intended to substitute. Two figures, five tables and four references are included. (JB)
Publication Type: Speeches/Meeting Papers; Reports - Research
Education Level: N/A
Audience: Researchers
Language: English
Sponsor: N/A
Authoring Institution: N/A