Predictability of Typical Graduate Admissions Criteria: A Report to CAPCSD

William J. Ryan, Ph.D.
Texas Christian University

In 1998, Ryan, Morgan, and Wacker-Mundy published an article in Contemporary Issues in Communication Sciences and Disorders based on an on-going study whose purpose is to quantify the relationship between GRE scores, undergraduate grade point average (UGGPA) and other potential pre-admission predictors, and graduate student success as measured by performance on the PRAXIS (formerly known as the NESPA or National Examination in Speech Pathology and Audiology).  For purposes of comparison with other studies, GGPA is also being used as a separate criterion measure.  A secondary objective is to gather preliminary data on additional pre-admission criteria such as undergraduate major and personal interviews, in an attempt to identify variables that might be more effective in predicting success when compared to standard measures such as GRE scores.

The PRAXIS score was chosen since it is currently the “gold standard” against which all certified clinicians are compared and is the only quantitative value used by the American Speech-Language-Hearing Association in the certification process.  Also, a number of institutions have abandoned traditional comprehensive examinations in favor of the student’s PRAXIS score as the final arbiter of successful completion of the graduate program.  Therefore, an individual’s PRAXIS score might be viewed as a final outcome measure of the degree to which the institution was able to impart the knowledge and skills that the profession tacitly agrees are required for a clinician to be viewed as competent.

Records were originally available for 96 graduate students who completed their degree requirements between 1993 and 1997.  A stepwise multiple regression analysis was conducted in which the dependent variables (PRAXIS score or GGPA) were regressed on the independent variables: GRE verbal score (GRE 1); GRE quantitative score (GRE 2); GRE analytical score (GRE 3); GRE total verbal and quantitative score (GRE 1&2); GRE total score (GRE Total); GPA in speech-language course work (SLP-GPA), and; (UGGPA).  In addition, t tests were used to test for differences between mean PRAXIS scores for four subgroups (1) undergraduate major vs. non-major, (2) personal interview vs. no interview, (3) students with GRE total verbal and quantitative scores > 1000 vs. those with scores < 1000, and (4) students with GRE Total scores > 1400 vs. those with scores <1400.

Results were in general agreement with those of studies reported in other disciplines.  GRE scores failed to predict either PRAXIS scores or GGPA, accounting for only 16% of the variance in the PRAXIS scores and even less in the GGPA.  Somewhat surprisingly, UGGPA was no better than chance at predicting either PRAXIS score or GGPA.  However, SLP-GPA was moderately correlated with GGPA (r = .60).  When means for the four subgroups were analyzed, significant differences were found as follows: (1) students with pre-admission GRE total verbal and quantitative scores > 1000 had significantly higher mean PRAXIS scores than those with scores < 1000, and (2) students with pre-admission GRE Total scores > 1400 had significantly higher mean PRAXIS scores than did those students with GRE Total scores < 1400.  Students who majored in a field other than speech-language pathology and students participating in a pre-admission interview obtained PRAXIS scores that were higher than those obtained by undergraduate SLP majors and those who did not participate in interviews, respectively.  Although the mean differences approached significance, they did not meet the required .05 alpha level.

The purpose of this report is to provide an update on our on-going study of predictors of performance in SLP graduate applicants.  We now have 130 students in our database whose demographics have remained essentially unchanged.  The same protocol for data collection, reduction, and analysis has been used throughout the on-going study (see Ryan, Morgan & Wacker-Mundy [1998] for a detailed description of methodology).  Overall, there has been very little change in results compared to our original findings.

In 1998 we wrote:

Based on the findings of this study, it appears that traditional pre-admission criteria such as GRE scores, UGGPA, and SLP-GPA, alone or in combination, have little or no relationship to the magnitude of the student’s NESPA score or GGPA.  Therefore, to the extent that these two outcome measures are reflective of an applicant’s potential, independent of the education and training provided by the graduate program, the pre-admission criteria presently in use are of little predictive value. (p. 60).
After adding approximately 50 students to our database and re-analyzing our data, we found nothing to change the aforementioned statement.  Of special concern is the exceedingly low predictive value of UGGPA that was noted in Study 1 and has remained constant in Study 2.  In essence, use of UGGPA is no better than chance for predicting the outcome measures of interest.  Yet many if not all graduate programs include UGGPA in their admissions decisions, often as a screening device to make the first “cut” before they have a chance to consider other variables such as letters of reference, letters of intent, etc.

So where do we go from here?  On one hand, we have very strong evidence that traditionally used pre-admission criteria have little or no value in predicting quantitative measures such as GGPA and PRAXIS scores.  This is especially troublesome for graduate admissions committees that often have 10 or more applicants for each available position.  Unfortunately, the findings presented here offer no additional insight into this vexing problem, although a possible alternative might be the development of some type of standardized undergraduate evaluation similar to the PRAXIS examination currently in use for graduate students.  On the other hand, we must not lose sight of the finding that, in spite of widely divergent GRE scores, UGGPA’s, and SLP-GPA’s, the overwhelming majority of our students successfully complete and often far exceed the requirements for ASHA certification.

Reference

Ryan, W.J., Morgan, M., &  Wacker-Mundy, R.   (1998).  Pre-Admission criteria as predictors of  selected  outcome measures for SLP graduate students.  CICSD, 5, 54-61.