A stroll down the slippery slope of Sub-group Norms
A stroll down the slippery slope of Sub-group Norms: What do boy/girl developmental differences have to do with Affirmative Action?
“Girls develop more quickly than boys.” That seems like a harmless statement. In fact, not only do most parents agree with the statement, but developmental standardized tests also find that, overall, girls do indeed develop many skills more quickly than boys. Hold on tight because we’ve just stepped on a very slippery slope. I have just introduced you to the concept of “Sub-group Norms,” which refers to first taking a large group (in this case children), then dividing that group into two smaller groups (in this case boys and girls), and finally examining differences between the groups. Sub-groups often differ, with one group being more advanced than the other. For example, if we examine sub-group norms for language development, we will indeed find that among toddlers and preschoolers, girls develop skills more quickly than boys. As another example, consider height. If we examine sub-group norms for height among 18-year olds, we find that the young men are taller (on average) than the young women. So far no controversy, right? The problem is that when we agree to use sub-group norms, we implicitly agree with the premise that we are willing to hold the lower functioning group to a lower standard—in effect, we are willing to expect (and accept) less from one group than the other. It is an unintended consequence of educational assessment and public policy that in its origin was intended to be kind. You see, use of sub-group norms opens a Pandora’s Box of other sub-groups. What if standardized tests found differences between children who are White, Black, and Asian. If we agree to use sub-group norms for boys and girls, doesn’t that mean we should use sub-groups norms for different racial groups? What about using sub-groups norms for poor, middle-class, and rich families? If you have any doubts about the politically charged connotations of this topic, just ask Richard Herrnstein and Charles Murray, whose 1996 book, “The Bell Curve: Intelligence and Class Structure in American Life” earned them both national acclaim and scorn (e.g., Peter Brimelow of Forbes wrote, “Like Darwin's Origin of Species -- the intellectual event with which it is being seriously compared -- The Bell Curve offers a new synthesis of research...and a hypothesis of far-reaching explanatory power,” while CBS ran a 60-Minutes segment that was highly critical). In the world of early childhood education and assessment, there are those who would charge racism (or at the very least “bad form”) at the suggestion that one group of children should be held to a lower standard than another, yet by way of analogy, in the adult world some of these same people may staunchly defend Affirmative Action.
Let’s re-visit the landmark 1954 U.S. Supreme Court case of Brown v. the Board of Education of Topeka, which overturned earlier rulings (most notably Plessy v. Ferguson from 1896). The Warren Court voted unanimously (9 to 0) that “separate but equal” is “inherently unequal.” Therefore, racial segregation of schools was deemed to be a violation of the Equal Protection Clause of our Fourteenth Amendment. This was a major milestone on the journey to integration, and a major victory for the Civil Rights Movement. But here’s the (rhetorical) question as the 1954 Brown decision applies to sub-group norming: If “separate but equal” is considered “inherently unequal,” then when we hold children to unequal educational standards, wouldn’t that be, by logical extension, also considered “unequal”? While sub-groups norms may indeed exist in certain areas, why would we, as a matter of educational policy, expect any less (or any more) from any sub-group of children?
How has history addressed this issue? We find a powerful example by Robert L. Williams from 1972, when he created the Black Intelligence Test for Cultural Homogeneity with content derived from black culture and inner city slang. As Williams suspected, black students scored better than white students, and he used his results to illustrate his point about cultural bias in testing. As an interesting footnote, Williams purposefully called his test the Black Intelligence Test for Cultural Homogeneity so that he could make his point even more powerfully with the it’s acronym, the “B.I.T.C.H.” Williams’ “B.I.T.C.H.” test prompted an important and productive professional discussion about which tests which were culturally loaded, and which were culturally biased.
How have the courts spoken on this issue? In a landmark case, “Larry P. v. Riles” (1979) the California Ninth Circuit Court of Appeals had to decide what became known as “the case of the six-hour mental retardates.” To summarize, based on results of IQ tests, there was disproportionately high placement of black students in special education classes (know as classes for the “educable mentally retarded”). Judge Robert Peckham wrote that standardized intelligence tests “are racially and culturally biased, have discriminatory impact against black children, and have not been validated for the purpose of essentially permanent placement of black children into educationally dead-end, isolated, and stigmatizing classes for the so-called educable mentally retarded.” In his dissenting opinion, though, Judge William Enright noted that a correct placement in a special education classroom (i.e., educable mentally retarded) is beneficial, not a stigmatizing dead-end, and that there was no better alternative in establishing educational placement and predicting academic outcomes than IQ tests. Do you agree with Judge Peckham or Judge Enright? In spite of Enright’s dissenting opinion, the court resolved that the proportion of black children in classes for the mentally retarded must closely match their proportion in the population. That sounded like a nice little solution on the surface, but the decision left an 800 pound elephant in the living room, as there are some obvious logical flaws that we must consider. First, what if there is a disproportionately low percentage of black children placed in special education classes? Would that mean that black children are being under-served? Or, would that mean that black children in that particular school district were outperforming the norm? Either could be true, and there would be no way of knowing for sure. Second, let’s consider a different scenario. If the proportion of black children in special education matched the proportion of black children in the general population, the court’s requirement would be satisfied, but surely we could envision a case where there are many black children in a particular community who are eligible for (but are not receiving) special education service. Thus, black children would be under-served, in spite of the fact that the Ninth Circuit Court’s requirement of “demographic matching” would be satisfied. Of course, the most obvious way to remedy this would be to create a test designed to reliably and validly assess individual children, regardless of race and cultural factors.
In this ironic journey, we see how the kind-hearted intentions about applying sub-group norms to children can have striking, unintended consequences (“All roads to Hell are paved with good intentions”). The good news is that we have learned from our history. We can happily report that today in the field of early childhood assessment professionals strive to navigate these subtle issues and avoid having an “all-or-none” stance. At the same time we strive to be “color-blind” and “gender-blind” while remaining culturally sensitive and in touch with our common-sense. We strive to measure developmental status with equal expectations across all children. Indeed, "We hold these truths to be self-evident: that all people are created equal."