College Admissions and the SAT I

WE APPLAUD RICHARD Atkinson’s article [May 2005, “College Admissions and the SAT”]. We hope that the changes in the SAT will remedy the problems that Atkinson described.

One author [Baron] had an experience with his son similar to Atkinson’s experience with his granddaughter. When Baron’s son was in kindergarten, he was in a small group of children who could already read. Instead of using this wonderful skill to expand the children’s world, the teacher wasted considerable time teaching them to take multiple choice tests “to prepare for the SAT.”

We published a study in 1992 on SATs, achievement tests, and high school class rank as predictors of college performance, that reached conclusions similar to the Geiser and Studley study cited by Atkinson. Of course, we were not in a position to take action as Atkinson was, but Baron, then chair of the admissions committee of the University of Pennsylvania’s school of arts and sciences, tried and failed to persuade the school to drop the SAT I from its admissions criteria, while retaining SAT II and high school class rank.

The greatest difficulty in carrying out our study was obtaining the data. It was almost as if Penn did not want to know how much or how little the SAT I was contributing to the quality of its admissions process. We conjecture that there is a lack of information about the incremental validity of the revised SAT I at Penn and elsewhere. We will not know whether the new test is an improvement unless someone analyzes the data, and, in our experience, we cannot count on the College Board to do this.

— Jonathan Baron and M. Frank Norman
University of Pennsylvania

College Admissions and SAT I Data Bias

ATKINSON clearly explained why he believes that achievement tests (the old SAT II) are better than general intelligence tests (the old SAT I) for the admissions process in the University of California system [May 2005, “College Admissions and the SAT”]. As scientists, we cannot help but be impressed by the data he presented, namely, that SAT II scores correlate with freshman GPA better than SAT I scores do. It is impossible not to agree with him that the University of California will benefit from the changes in the SAT that will give more weight to knowledge obtained in high school, including writing skills.

Unfortunately, Atkinson does not seem to realize that his data are very biased. By mandate, the University of California accepts students at the upper 12.5 percent of statewide high school graduates. This means that almost 90 percent of high school graduates (many of whom will attend college elsewhere) never become UC freshmen and are not included in Atkinson’s data. Thus, the data used to compute correlations are rather biased. Of course, the point here is that correlations within a narrow range of the variables under study are expected to be rather low. Indeed, even the combination of high school grades and SAT II scores “accounts for 22.2 percent of the variance in first-year college grades” at the University of California. If we decide to accept that UC will benefi t from the changes in the SAT, we still have no idea whether other universities will.

Atkinson acknowledges that “one of the clear lessons of history is that colleges and universities, through their admissions requirements, strongly infl uence what is taught in the schools.” This means that the changes in the SAT will affect not only thousands of universities but also, most likely, the whole grade school system. To propose changes with such widespread consequences based on admissions data from one university (even if a large one) is, at the least, irresponsible. It is quite possible that general intelligence (the old SAT I) provides a better predictor of college performance (especially in junior and senior years) than previous knowledge does. Only data from a large number of universities can help us elucidate the issue, and Atkinson did not provide these data.

— Roberto Refinetti
University of South Carolina

In Sync With APS

ROBERTA KLATZKY, IN THE April 2005 issue of the Observer [“The Case for Changing Our Name”], suggested that APS change its name. She separated the previous arguments of the membership that appeared in the Observer from February to November, 1999, into three categories: “In Favor; Opposed; and Other opinions (thankfully, a small minority).”

I believe I authored the first example listed in her third category, “the thankful small minority,” but she got it all wrong. I really belong in the Opposed category, “Those that need to proclaim they are scientific aren’t,” which I would modify to: “Those who need to proclaim psychological science suffer from physics-envy.” Klatzky also truncated and revised my suggestion for a name change, which had a certain ragtime syncopation to it that was lost in her abbreviated version: The American Association for By God, Real and We Mean Really Real, Scientific, Real Really Scientific, By God, Psychological Science. Said aloud enough times by enough members, and soon everyone would know just how scientifically we regard APS.

So, yes, let’s change the name to the Association for Psychological Science, but let’s keep Klatzky to her promises. After the name change, she says: 1) we become as recognized and respected as “the other Ay-Pee thing;” 2) international membership will increase; and 3) we will get more funding and respect on the Hill, all measurable goals. In the spirit of really scientific social science, anyone for evaluation research of that program?

— Frederick Meeker
California Polytechnic State University

Comprehensive Final Exams

FOR SEVERAL YEARS OUR DISCIPLINE members have discussed the need for a Final Exam in our Introductory Psychology classes since we are a multi-campus community college and employ over 40 adjunct instructors each year.

Our faculty contract requires that the catalog description topics be taught by all instructors but allows great latitude in decisionmaking about selection and emphasis. We believe that a comprehensive final exam would give us an indication of content and retention, and would be a valuable bit of information for future assessment.

If you haved used an exam or have experience or opinions that would be helpful, we’d appreciate hearing from you.

— Don Nichols
Oakland Community College
ddnichol@oaklandcc.edu