"WALTER LIPPMAN AND ME" ....
Rebecca Zwick's letter to The Nation responding to my review of her book and my reply.
January 1, 1970To the Editors:
Eighty years ago, the journalist Walter Lippman took on the standardized testing enterprise in The New Republic, addressing such broad issues as the effects of education, opportunity, and heredity
on test scores.
For example, Lippman dismissed the claim that IQ tests measure hereditary intelligence as having "no more scientific foundation than a hundred other fads, vitamins and glands and amateur psychoanalysis and correspondence courses in will power." His articles on testing continue to be valued today not merely because he could turn a phrase, but because he had a firm grasp of the complex technical and political issues surrounding the use of test scores.
Alas, Peter Sacks is no Walter Lippman. To Sacks, who reviewed my book Fair Game? The Use of Standardized Admissions Tests in Higher
Education (June 24), the issues are simple: Tests are evil; eliminating them is good.
Sacks has undoubtedly been aware of my work because I have pointed out errors and omissions in his writings on testing; in fact, I do so
in my book. He ignores large portions of the book in order to characterize it as "a defense of the hegemony of gatekeeping exams."
A reader of the review might be surprised to find that my book proposes a new consumer agency to monitor admissions testing, discusses the perils of relying too heavily on test scores in admissions decisions, and describes research, including some of my own, in which test scores did not do a good job of predicting subsequent grades.
Rather than attempt to address every inaccuracy, I will focus on a central feature of Sacks's review his belief that the existence of score disparities among ethnic and economic groups proves that
admissions tests are biased. In Fair Game? I point out that determining whether tests are biased is complex and requires a willingness to look beyond patterns of average test scores.
In Change (March/April 2001), I commented on Sacks's earlier Change article, "Standardized testing: Meritocracy's crooked yardstick": "[Sacks] cited several studies to prove that SAT scores and socioeconomic status are related, and alluded to [a study conducted by the National Center for Education Statistics].What he neglected to mention is that this study showed that socioeconomic status was also related to high school grades ... [and to course background, teacher evaluations, and extracurricular activities.] In particular, 24 percent of the high-SES group, compared to only 10 percent of the low-SES group, had high school [grade-point averages] of at least 3.5... What the GPA and the SAT have in common is that they are indexes of previous achievement and therefore reflect past inequalities in educational opportunity. In The Nation (June 5, 2000), Pedro Noguera
and Antwi Akom noted that "[e]xplaining why poor children of color perform comparatively less well in school is relatively easy: Consistently, such children are educated in schools that are woefully
inadequate on most measures of quality and funding."
Sacks omitted the findings on grades and other achievement measures from his book as well as his Change article. Presenting the complete
results would have undercut his position that some inherent property of tests causes the scores to be related to economic factors. (Including all the findings might have also required him to abandon his pet phrase, "the Volvo effect," which he uses to refer to the association between family income and standardized test scores.)
In addition, Sacks is incorrect in implying that class rank admission plans like the Texas 10% plan, which involve consideration of high
school grades, but not test scores, have uniformly led to greater campus diversity. The Dallas Morning News, for example, reported on
June 19, 2002 that at Texas A&M, the percentages of Black and Latino students have decreased since the initiation of the Texas plan. As I point out in my book, the plan is structured so that diversity benefits are likely to accrue to the state's flagship institution, UT Austin.
Finally, in response to Sacks's criticism that my writing style is textbookish, I readily concede that I lack his ability to generate catchy phrases like "Volvo effect" and "crooked yardstick." But
clever labels are a poor substitute for thoughtful consideration of the controversies that surround the use of standardized tests.
Peter Sacks replies to Zwick:
In response to my criticisms of her new book, Rebecca Zwick takes aim at the reviewer. She says I believe that "tests are evil; eliminating them is good." It's not surprising she'd make up this straw man, since attacking it also sums up the entire marketing strategy behind her book.
Zwick -- a former researcher at the Educational Testing Service, the firm that produces such standardized tests as the SAT -- and her publisher have touted Fair Game? as a source of objective information about testing, positioned to clear up all this testing fuss with common sense and straight
facts. If one chooses to look at a different or broader set of facts than does she, or to interpret them with a non- ETS spin, Zwick seems to imply
that one must then be a simpleton and an ideologue.
Zwick tries to make hay of the finding that high school or college grades, just like test scores, also correlate strongly to socio-economic status.
Not recognizing this, as Zwick takes pains to do in her book, is to unfairly single out standardized tests as punitive to poor and minority kids, Zwick claims.
Like so much of her book, Zwick seems to miss the big picture. The thrust of my entire critique of the testing culture -- and her book -- is that
gatekeeping tests give questionable weight to one-time performance on highly abstracted testing exercises, which by definition are mere
approximations of genuine work. And mostly poor approximations, at that.
Given this, it's no wonder that test scores are such feeble predictors of later success, whether in school or in work. Just as Bates College and other institutions have done with great success in their efforts to reduce the importance of admissions tests, I'll take classroom performance as measured by grades, portfolios of student work, and other documentation of student accomplishments both in and out school any day over test performance as an indicator of how a student will perform in real life, not the tested life.
Regarding the Texas 10 percent plan, Zwick says that I'm incorrect in implying that de-emphasizing the SAT has led to greater diversity for all
state institutions. In fact, I'm not implying any such claim in the context she quotes. I draw on data only from the University of Texas at Austin. Zwick speculates that the plan has merely reshuffled the deck in terms of statewide enrollments of minorities. If Zwick wants me or another reviewer to take her seriously on this point, she'd better offer up something of substance or do some real analysis. In her book, Zwick could only muster up this: "Data on the statewide (itals) effect of the Texas 10 percent plan are hard to come by." (Zwick, p. 136).
What can she possibly mean with such a vague statement? That university officials are trying to hide some dirty little secret? Does it mean that
there are no campus specific enrollment data broken out by race and ethnicity? Seems improbable. Or could it mean that Zwick could find no readily available studies by credible researchers that support her claim
that enrollments have merely been redistributed from other state campuses to Austin?
But even a boatload of data needs a theory, an explanation of what the data mean. Alas, Zwick offers readers no theoretically plausible explanation whatsoever as to why minority enrollments might be expected to
decline across the state by reducting the emphasis on SAT scores.
In fact, there's every reason to expect just the opposite. As for textbookishness, that in isolation is certainly no major offense to the sensibilities of thoughtful readers. Sign me up any day for a dry but forthright book about testing in America.
Regarding Zwick's curious reference to me and Walter Lippman, I won't touch that one with a 10-foot, number 2 pencil.