What does it mean when a school, having rejected a child who applied for admission, explains that he or she just “isn’t a good fit” (or “match”) with the school? In some cases, sure, the phrase would seem appropriate — for example, if there’s a marked discrepancy between the school’s and the family’s religious orientations, or if the school is committed to progressive education while the parents demand grades, quizzes, worksheets, and traditional discipline.
More commonly, though, it’s not clear at all how the decision to prevent a child from enrolling is best described as a lack of fit, particularly if the school’s goals and priorities (a) correspond to what most parents (including these) are looking for, and (b) can’t easily be distinguished from those of other schools. Try to imagine an admission director saying something like this to an applicant:
Well, you know, here at Tweedle-Dee School, we believe in “guiding our students to reach their optimum potential intellectually, physically, and socially” — so I’m afraid this really isn’t the right place for you. Perhaps you’d be happier at Tweedle-Dum Academy across town, which, in contrast to us, offers a “rigorous college-preparatory education in a caring and attentive school community.”1It’s hard to avoid the suspicion that “not a good fit” is often just code for “You’re not good enough for us.” And if that’s true, it’s not only the dishonesty of the language that rankles but the implication that the school is looking out for the interests of the child it has rejected. Be straight about what’s going on: She wants to attend your school and you decided she didn’t measure up. Don’t pretend you’re doing her a favor.
Euphemisms aside, there are two assumptions driving the admissions process. The first is that schools can predict with reasonable accuracy which applicants are most likely to be academically successful. There’s good reason to doubt whether that’s true. Standardized test scores, including those that supposedly measure intelligence, are a lousy bet (see sidebar below), but it’s not clear that anything works all that well. For example, social skills and the extent of misconduct in a child’s history are poor predictors of subsequent academic performance — “even among children with relatively high levels of problem behaviors.”2
| Standardized Tests for Admissions? Why? Having studied and written about the design and use of standardized tests for some years now, and having listened to heads, teachers, and admissions officers at more independent schools than I can count, I’m damned if I can find any legitimate justification for using such tests in the admissions process. Whether we’re talking about the Secondary School Admission Test (SSAT), the ERB’s Independent School Entrance Exam (ISEE) and Early Childhood Admission Assessment (ECAA), the Wechsler Preschool and Primary Scale of Intelligence (WPPSI) or Wechsler Intelligence Scale for Children (WISC), the Hawaii Early Learning Profile (HELP), or any of the others, I believe such tests have no useful role to play even for schools determined to distinguish wheat children from chaff children. It’s particularly painful when schools that think of themselves as progressive, child-centered, alternative, or otherwise enlightened continue to require prospective students to take one of these tests when they apply. Their rhetoric says, “We look at children as individuals and are committed to 21st-century education.” Their use of these tests says, “We still haven’t let go of standardized assessment that represents a throwback to early 20th-century beliefs about intelligence and sorting.” Here’s what we know about standardized tests in general:1 • Their results are highly correlated with socioeconomic status, to the point that they tell us less about the potential of the child than they do about the size of the house in which that child lives. • They measure the skill of test-taking itself, which often isn’t related to the intellectual qualities that most of us care about. In fact, three studies independently found a statistically significant negative correlation between scores on such tests and the depth of students’ thinking. • The worst such tests are those (1) whose format is multiple-choice, meaning that children cannot generate or even explain their answers, (2) that are norm-referenced, which means they’re designed primarily to answer the question “Who’s better than whom?” rather than “What characteristics does this individual possess?,” (3) given to children younger than about eight years old, or (4) used to measure aptitude rather than achievement. (The “A” in “SAT” used to stand for “aptitude” until the College Board was finally forced to admit that the test didn’t measure that at all.) Which leads us to the tests that are employed specifically for admissions purposes. There’s a reason that nearly 850 U.S. colleges and universities have stopped requiring SAT or ACT scores, with more joining that list every year.2 Scores on these exams have nothing interesting to tell us beyond what can be learned about an applicant in other ways, particularly when it’s possible to sit down with him or her. Even in purely statistical terms, and even if we’re interested only in predicting future grades and test scores (as opposed to more authentic outcomes), such tests are unnecessary at best, and misleading at worst. What’s true of tests used for college admissions applies even more strongly in the case of younger students. For example, one recent study found that SSAT scores predicted only about 15 percent of the variance in high school grades.3 And IQ-type tests given to preschool-aged children? Aside from the disturbing implications of using any method to sort tiny tots into winners and losers, these tests are generally useless at predicting academic achievement even just a year or two later.4 The test makers themselves acknowledge as much, explicitly or implicitly. The ISEE “is designed to measure the verbal and quantitative reasoning and achievement of students in grades 4–11” — what students already have been taught, not what they’re capable of doing. (The word “aptitude” appears nowhere in the ERB’s guide.)5 The ECAA assessments, meanwhile, “do not predict long-range school success.”6 That’s why the ERB just tries to call attention to correlations with current academic performance. Some schools that rely on standardized tests for four-year-olds have become annoyed because gung-ho parents are now prepping their children for the exam in order to give them a leg up. “There’s nothing magical in getting a high score anymore,” one admissions officer complained to the New York Times.7 Notice the word “anymore” in that sentence, which implies that the test was — and, if it weren’t for those pesky parents gaming the system, would continue to be — a valid and appropriate way to evaluate small children. But it isn’t. The problem isn’t the parents, hypercompetitive though they may be; it’s the schools that use the tests.8 The result is both false positives and false negatives, which means that even at a traditional, grades-and-tests kind of school, many young children with high scores will not do particularly well on real classroom learning tasks, and many young children with low test scores will do just fine — and the predictive accuracy keeps dropping the farther into the future you look. This is true, among other reasons, because of the inherent limits of tests at predicting the future or assessing aptitude, because most tests focus on a very restricted set of cognitive skills, and because children develop at different rates. Thus, it won’t do to say, “But the test is just one of several factors we consider!” The burden rests with schools to show that they’re needed at all, that they’re valid and informative instruments, that we can vouch for the values and priorities that underlie their design, that we learn something from them about the most important characteristics of children that we couldn’t have learned in other ways. Think of it this way: If a test result just corroborates what we already knew from nonstandardized sources, then it’s not adding anything of value. But if there’s a sharp divergence between what the test tells us and what we’ve learned from all the other indicators, which are you going to believe? Either way, then, why funnel all that money to the test companies? — Alfie Kohn Notes1. For more detailed discussion of, and the research support for, these arguments, see my book The Case Against Standardized Testing (Portsmouth, NH: Heinemann, 2000).2. See www.fairtest.org/university/optional. 3. Elena L. Grigorenko et al., “Are SSATs and GPA Enough? A Theory-Based Approach to Predicting Academic Success in Secondary School,” Journal of Educational Psychology 101 (2009): 964–81. 4. A review of 44 studies found serious limits to the predictive validity of tests given in preschool and kindergarten even when the point was just to predict scores on achievement tests a year or two later (Juhu Kim and Hoi K. Suen, “Predicting Children’s Academic Achievement from Early Assessment Scores,” Early Childhood Research Quarterly 18 [2003]: 547–66). For readable discussions of this issue (and other references to relevant research), see the chapter “The Search for Intelligent Life in Kindergarten” in Po Bronson and Ashley Merryman, Nurtureshock (New York: Twelve, 2009); and the first half or so of Jennifer Senior, “The Junior Meritocracy,” New York Magazine, February 8, 2010. 5. http://isee.erblearn.org/pdf/01_Lower/guide.pdf 6. https://www.ecaatest.org/ERBRegistration/ParentPortal/GeneralInformation.aspx 7. Jenny Anderson, “Private School Screening Test Loses Some Clout,” New York Times, May 7, 2010, p. A1. 8. Any tests. The question isn’t whether to use those administered by individual schools as opposed to a standardized product bought from ERB or Pearson; it’s whether to test preschoolers at all. |
The second assumption underlying traditional admissions is that sifting out all but the most promising children (and then working hard to persuade their parents to enroll) is what schools should be doing. If the first premise is accepted too readily, the second is simply taken for granted — to the point that calling it into question will strike many people as bizarre. Why would we even have an admissions office if we weren’t trying to figure out whom to admit? And whom would we rather admit than the best?
Let me begin by pointing out the connection between the two assumptions. If we can’t reliably identify potential — that is, predict future impressiveness — in a 14-year-old, much less a four-year-old, then we end up evaluating children on the basis of what they’ve already done. And that disproportionately reflects the family’s socioeconomic status. Even when we think we’re selecting on other criteria, we’re mostly privileging privilege, and therefore reproducing it in another generation. The troubling truth is that selective schools help to perpetuate the deep inequities that define our society, not just failing to make things better but actively making them a little worse (notwithstanding piecemeal diversity efforts3). There are the elect, who merit the benefits of our talented faculty and terrific facilities; and then there’s everyone else — filed under “not a good fit.”
The problem is even worse if what we’re mostly interested in identifying for the purpose of admission is academic potential. Even if we were able to predict who will fare well in our classrooms, this is a very narrow definition of faring well. By trying to choose kids who seem to be smart, particularly according to traditional criteria like grades and test scores, we exclude an awful lot of applicants with other talents and interests, with admirable attributes (compassion, humor, courage, playfulness) and intriguing personalities — none of which necessarily goes hand-in-hand with verbal or mathematical proficiency. The result is that our schools become less interesting places and our students are exposed to fewer people who are meaningfully different from themselves.
The Children Who Need Us Least…
Selectivity isn’t an accident. It may not be in the best interest of individual children or our society more generally, but obviously someone thinks it benefits the institution that’s doing the selecting. So let’s ask an embarrassingly basic question: Why do schools try to admit only the crème de la crème if they’re in a position to do so?First, there’s the prestige factor: An institution is seen as more desirable if it can boast about its selectivity, which then allows it to be even more selective. Staff, parents, and students may simply enjoy being associated with a club that lots of people want to join but can’t. That’s why colleges are often judged less by the quality of the education they provide than by the number of applicants they manage to avoid admitting. Some even engage in expensive marketing campaigns to encourage more students to apply… so the school can reject them. Has this mentality trickled down to PK–12 schools?
Second, a more elite group of children coming in is believed to result in a more impressive placement list going out — that is, a larger group of students accepted by fancy colleges, which means another round of bragging rights. But here we confront an interesting dilemma. The premise of such bragging is that the school, with its top-flight teachers and tireless commitment to excellence, is largely responsible for all those Ivy-covered acceptance letters. The implication is that our remarkable school could do wonders with just about any child, which is why you should send yours here or write us a fat check so we can propel even more students to highly coveted universities. Yet the admissions process is committed to not admitting “just any child”; its mission is to select exceptional children who are extremely likely to succeed. (Indeed, it sometimes seems that the most advantaged and precocious youngsters could spend 14 years playing video games and still graduate with flying colors.) So which is it? If you’re boasting about your placements, you shouldn’t be admitting selectively — and vice versa. (See James Moffett’s lacerating remark that serves as the epigraph for this essay.)
Third, if smarter (or richer) children are admitted, they’re more likely than other applicants to make — and perhaps donate — a lot of money when they’re alumni.
Finally, the kids whom we regard as the most able, those who have been certified as (here comes another unsettling phrase) “ready to learn,” are really just the easiest to educate. Consider a conversation that the education theorist Martin Haberman reported having with his grandson’s kindergarten teacher at a selective school. “Wouldn’t it make more sense to admit the children who don’t know their shapes and colors, and teach them these things?” he asked. The teacher looked at him as if he were “leftover mashed potatoes,” but he persisted:
Next year, my grandson, who is already testing in your top half, will have had the added benefit of being in your class for a whole year. Won’t he learn a lot more and be even further ahead of the four-year-olds who failed your admission exam and who have to spend this year at home, or in day care, without the benefit of your kindergarten? Will the four-year-old rejects ever catch up?
This question did even less to endear him to the teacher, but Haberman by now had realized what was going on more generally, and he summarized his epiphany as follows: “The children we teach best are those who need us least.”4
Some years ago, in thinking about college admissions, I was weighing the predictive power of high school grade-point average against that of the SAT or ACT. Some critics emphasize (correctly) that these exams are much less useful than grades at predicting performance, but I was at pains to point out that grades have their own problems and, in any case, it would be more sensible to lump them together into a compound variable called gradesandtests, which fails to predict anything other than future gradesandtests; it tells us nothing about who will be creative or a deep thinker or excited about learning or happy or successful in his or her career.
But I came to realize that even this reframing of the discussion failed to challenge the premise that I, too, apparently shared with more conventional participants in the colloquy about admissions. The eminent psychologist David McLelland, known for his theory of achievement motivation, delivered a public lecture at the Educational Testing Service in 1971. This talk was devoted primarily to raising pointed questions about the value of intelligence tests (Do such tests predict “who will get ahead in a number of prestige jobs where credentials are important”? he asked rhetorically. Sure. And so does “white skin.”).
In an almost offhand way, McClelland then issued what struck me as a truly provocative and profound challenge. Why, he asked, do we spend time trying to figure out which criteria best predict educational success? Why are schools looking for the most qualified students? “One would think that the purpose of education is precisely to improve the performance of those who are not doing very well,” he mused. If schools were “interested in proving that they could educate people, high-scoring students might be poor bets because they would be less likely to show improvement in performance.”5
Of course, institutions that get to choose whom to admit tend to look for the applicants who are good bets to succeed: those who seem smart and compliant, will require the least time and effort, and are most likely to make the school look good. And that means those who most need what your school has to offer are turned away.
Admission by Lottery
A radical critique deserves a radical solution — and I use the word radical in its original Latin sense to mean “of the root.” Switching from one admissions test to another, or appointing someone to oversee diversity issues, or adding “emotional intelligence” or “character” to the list of admissions criteria are not responses that go to the root of the problem.In its purest form, the response that makes the most sense is simply to stop being selective, which means choosing your student body randomly from those who apply. It’s clean; it’s fair; it’s socially responsible; it can be used at any level, from preschool to graduate school;6 it gives you a chance to show just how amazing your school really is; … and it will be rejected as hopelessly idealistic by the very schools that claim to be committed to the values that recommend it.
My sense is that objections to admission by lottery fall into two categories: those that can be addressed by tweaking the basic idea, and those that say more about the people who offer them than about actual flaws with the plan. To begin with the first:
• Do you want to give priority to siblings, or to minority students or those from other underrepresented groups? Go ahead: Put your thumb on the scale. We’ll call it a modified lottery, and it will still be an enormous improvement over the status quo.
• Are you worried about students with profound disabilities whom you’re unable to accommodate — or are unwilling to spend the resources to accommodate? We’ll give you a pass on that, too, so you can exclude those you’re sure you can’t help and limit the number of those who would be extremely challenging to help.
• Is there a legitimate “fit” concern based on educational philosophy? As long as you’re not trying to be all things to all people — as long as you’re willing to describe your school accurately rather than taking advice from marketing consultants whose goal is to make everyone in the world want to apply — families will tend to self-select.
The other kind of resistance to a lottery proposal is more disturbing. I’ve heard people object on the grounds that it would compromise their school’s “culture of excellence,” which is predicated on accepting only “people who share our values.” Nobody comes right out and says, “Some kids and families are inferior and we don’t want them in our building.” Rather, it’s phrases like these by which old-fashioned snobbery identifies itself and by which the powerful keep what they have and pass it on to those who look and sound as much like themselves as possible.
A lottery doesn’t compromise excellence, but the absence of a lottery does compromise democracy. After all, if the school has a culture of excellence, which applicants wouldn’t benefit from it? The only inference to be drawn is that some children, from a very young age, are thought to lack excellence (or “our values”), and that this deficit is irreversible.
Honestly, now, if your school admitted a random sample of all those who applied and those children then spent years learning from you and your colleagues, how do you think they would fare in life? If you’re not sure, conduct an experiment next year. Choose a fraction of the spaces you have — NAIS President Patrick Bassett suggests 10 percent; I’d propose half — and accept children for those spots by lottery while using your usual selection standards for the rest. See what happens, and remember to assess the results by criteria more meaningful than test scores.
The School at Columbia University, a K–8 school in New York City, accepts half its students randomly from applicants in the area. The Atrium School, a pre-K–6 school in Watertown, Massachusetts, has used a modified lottery for all its students since it opened its doors 30 years ago. Its founder, Ginny Kahn, told me recently that her point of departure was the belief that “each child is special, and [because] students have a variety of learning styles, it’s desirable to combine a range of students” rather than excluding the sorts of youngsters who are “overlooked by many private schools.”
Take a look at your school’s admissions practices. Then look at your school’s core values and the reason you personally became an educator. How’s the fit?
Copyright 2011 by Alfie Kohn.
Notes
1. These are actual quotes from the websites of two independent schools chosen virtually at random. At a speech for a group of admissions officers, I read a series of such excerpts, each describing a local school that was presumably familiar to everyone in the room, and asked if anyone could match each description to its school. I had no takers.2. Greg J. Duncan et al., “School Readiness and Later Achievement,” Developmental Psychology 43 (2007): 1428–46.
3. In the words of Steve Nelson, head of the Calhoun School in New York City, “Most of our private school ‘diversity’ comes in the form of kids culled out of underclass communities and headed on a track of upward mobility — programs like Prep for Prep, Oliver, ABC, Early Steps. This is our own version of creaming and doesn’t create authentically diverse school communities. Our schools should offer admission to some kids of no privilege, who have not been identified, prepped, or selected“ (personal communication, October 2010).
4. Martin Haberman, Star Teachers of Children in Poverty (W. Lafayette, IN: Kappa Delta Pi, 1995), p. 80.
5. David C. McClelland, “Testing for Competence Rather Than for ‘Intelligence,’” American Psychologist, January 1973, pp. 6, 2.
6. I first encountered a version of this proposal decades ago in the context of college admissions. Robert Paul Wolff, a philosopher, argued that all places in the freshman class, at least at the most elite institutions, should be awarded randomly to applicants who had met a minimum standard that was worked out jointly by the colleges. Interestingly, his motivation wasn’t just to address class and race bias but to rescue the intrinsic value of a high school education, something that’s lost when those years are devoted mostly to college prep. See Wolff’s The Ideal of the University (Boston: Beacon, 1969).