New Paradigms for Assessment

University Testing

In an attempt to "rethink the process and content of selection" from a pro-diversity perspective, University of Pennsylvania Law School Professors Susan Sturm and Lani Guinier have written The Future of Affirmative Action: Reclaiming the Innovative Ideal, published in a recent issue of the California Law Review.


The authors, who sponsored a 1996 conference on the topic in which FairTest played a significant role, begin with a review of current selection techniques that leads them to charge that "a 'testocracy' masquerades as a meritocracy." Based on a meticulously footnoted analysis, including many FairTest sources, they build a compelling case that contemporary society's "stock narrative" equating merit with performance on standardized tests is flawed. In fact, they show how these exams create an "arbitrary barrier for many otherwise-qualified candidates," often those from poor and minority backgrounds, because the tests do not consistently predict school and job success. "Standardized tests," they note, "adopt a one-size-fits-all approach to measuring successful performance" while ignoring many other crucial abilities and skills.


Explicitly rejecting the notion that "the solution is to develop a new, less biased, equally universal test that more accurately predicts future performance," Sturm and Guinier offer what they call "A New Paradigm for Selection and Inclusion." Though still vague in some details, the central proposition of these new policies would be "ability to perform" as demonstrated in real-world situations, not artificial, multiple-choice exercises. Examples of this approach include the successful open-admissions enrollment program at the City University of New York (see Examiner, Winter 1996-97) and on-the-job training programs in some trades. Such mechanisms "should be structured to enable individuals to show what they can do," with selection decisions made on "an individual's capacity to perform."


"We are mired in a testocracy that, in the name of merit, abstracts data from individuals, quantifies those individuals based on numerical rankings, exaggerates its ability to predict those individuals' future performance, and then disguises under the rubric of 'qualifications' the selection of those who are more socio-economically privileged," they conclude. "Only by rethinking our assumptions about the current system and future possibilities can we move toward the ideals that so many Americans share."


* California Law Review, Volume 84, Number 4, July 1996.