EBD Blog

DISPROPORTIONAL IDENTIFICATION OF EBD

The Problem of Disproportionate Representation in Special Education for Students with EBD

Some classifications of individuals are disproportionately represented in some categories of special education. That is, in some categories of disabilities created for special education purposes, students may be either over- or under-represented based on their proportion of the school population. For example, although girls represent close to 50% of the general school population, they are disproportionately under-represented in special education for students with emotional and behavioral disorders (EBD). And although African-American students comprise approximately 17% of the general school population they comprise about 27% of those identified as having EBD, which means they are disproportionately over-represented. Keep in mind that slightly less than 1% of public school students are identified as having EBD.

Disproportional representation is a well-established fact in special education, and it is a serious problem. It is a more serious problem in some schools than in others. But I think it is a problem that is poorly understood and about which few reliable data, other than those demonstrating the fact of disproportionality, inform us. I have encountered myths about the problem, and I do not think the level of rational analysis related to the topic has been consistently high. One of the myths I’ve encountered is that all “children of color” (i.e., those not of white European ancestry) are over-represented in special education. The data resoundingly disconfirm this myth. Another myth is that about 27% of African-American students (mostly males) are identified as having EBD, whereas the truth is that African-Americans constitute approximately 27% of those identified as having EBD (i.e., a closer approximation is that about 0.27% of African-American students are identified as having EBD for special education purposes). African American students are about 1/4 of the students in special education who are identified as having EBD, whereas African American students are less than 1/4 of the school population. True, African-American students comprise about 27% of those receiving special education for EBD, whereas African American students are about 17% of the general student population, meaning that they are seriously disproportionately over-identified (see National Research Council, 2002).

Another mistaken notion about disproportionality in special education for EBD students is that those in some groups are identified in higher proportions than is justified by prevalence data. No group, as far as I know, is identified anywhere near the proportion of students that studies of prevalence suggest. That is, reasonable estimates of the prevalence of serious emotional and behavioral disorders run, conservatively, at about 5% of the total population of children and youth (see Kauffman, 2005 for a review), whereas no group (e.g., African Americans, boys, Latinos, American Indians) is identified anywhere close to that percentage.

Perhaps this bears repeating in a different way: Although the percentage of African American students identified as having EBD for special education purposes is higher (by about a factor of 1.5 to 2.0) than the percentage of European American students so identified, neither African American nor European American students are identified at anywhere close to the percentage of students that prevalence studies suggest have serious emotional or behavioral disorders (see Kauffman, 2005 and U. S. Department of Health and Human Services, 2001). The disconnect between prevalence research and the facts about disproportionality suggests to me that people are concerned about something other than identifying students for special education in the EBD category when they need it, regardless of their ethnicity or other group identity (Kauffman, 2004). You might also go to the following address for another perspective on the lack of services for students with EBD (www.kansascitykansan.com/articles/2005/04/28/news/opinion/opin1.txt). Overwhelmingly, the data suggest that students of all ethnic groups are under-identified and under-served in the EBD category.

Explanations of Disproportionality

For some cagtegories and for some disorders, we readily accept disproportionality. We readily accept the fact that not every disorder comprising those we know as EBD will be found in the general population in the same proportion. That is, if there are two types of disorders, we do not necessarily expect that each will be identified 50% of the time; if there are 10 different disorders, we do not necessarily expect that each will comprise 10% of the disorders that are identified. We understand that some disorders are more prevalent than others.

Neither do we expect that all illnesses or problems will be identified in the same proportion in all groups (that is, we recognize that they may vary with groups differing in age, gender, SES, ethnicity, or national origin). Breast cancer is found disproportionately (though not exclusively) in women. Tay Sachs and sickle cell anemia are found in some ethnic groups more often than in others. Alzheimer’s disease and vehicular deaths are both correlated with age (i.e., occur disproportionately in some age groups).

We understand that some people are at higher risk for many things than are others because of the causal factors to which they are exposed, which include genetics, a host of environmental conditions, and the individual’s own behavior. For example, we don’t consider it a mystery that people who smoke are at higher risk for lung cancer than are nonsmokers. We don’t question the link between obesity and a variety of health problems, including high blood pressure. We understand that alcoholism and obesity, as well as a variety of other problems, run in families, probably for both genetic and environmental reasons.

Most of us have no difficulty understanding why “Increasingly, dental disease is a problem of economically disadvantaged children. The amount of tooth decay, the association [American Dental Association] notes, ‘is inversely related to income level’” (Levine & Wilgoren, 2005, p. B10). Most of us accept the explanation that poor kids tend to have less and poorer dental care and dental hygiene than do kids whose families have sufficient money to buy dental care, dental insurance, healthful foods, and dental hygiene products. Consider a report of bringing poor children to Howard University to address their needs for dental care:

Last year, nearly 100 children treated during Give Kids a Smile [an American Dental Association program, with the event reported being headed at Howard University by dentist Sally Cram] needed additional work after the day was over. Cram sent referral forms home with them and followed up repeatedly.
Only five parents called to ask about making an appointment with one of the dentists willing to donate additional services.
Only one parent actually made an appointment for her child.
She never kept it. (Levine & Wilgoren, 2005, p. B10)

Although we may consider it unfortunate that some children (in this case, poorer children) have more dental problems than others, we do not consider the problems to be manufactured by the service providers. Nor do we consider it unfair that someone is providing dental care for poor children; in fact we consider it unfair that poor children do not get the dental care they need, even if they need more such care than other children. We don’t want children to have dental caries or gum diseases, but don’t want these problems to be ignored or glossed over or explained away by some sort of philosophical speculation about disproportionality. We seem to agree that dentists generally do more good than harm and that children who have serious dental problems but do not go to a dentist regularly are at a serious disadvantage, even though the children who most need dental care may be disproportionately poor or disproportionately of African descent. Our concern is not about a child’s skin color or ancestry but about his or her need for the services of a dentist.

In special education, however, we have no firm data-based explanations for disproportional identification, especially for the EBD category. We have to piece together the few data we have with a rational analysis of the problem. This means, of course, that any given “take” or position on the issue could be proven wrong by well-designed studies and replications. These facts should not keep us from using whatever cognitive powers we have to try to figure out the conditions under which disproportional representation is seen as a problem and how we might establish, through careful research, why it exists.

Necessary and Sufficient Conditions for Objecting to Disproportionality

Disproportional over-representation is not objected to in cases in which the identification or inclusion in the program is assumed to be beneficial (as I just mentioned in the case of dental caries). If an activity or program is deemed beneficial (e.g., being identified as gifted or talented, achieving membership on a prestigious athletic team, being chosen for a coveted award, serving in a high-visibility role in government, receiving medical care for a disease), then it is under-representation (not over-representation) that is said to be disadvantageous. In short, I can think of no instance in which disproportionate over-representation is protested by the over-represented group if identification for and inclusion is thought to be beneficial or advantageous. Nor can I think of any instance in which disproportionate under-representation is protested by the under-represented group if identification for and inclusion is thought to be harmful or disadvantageous.

I am led to the conclusion that a necessary and sufficient condition for objection to disproportional over-representation in any program by an over-represented group is the assumption that, on balance, identification for and inclusion in the activity or program is disadvantageous. To me, it seems extremely likely that those objecting to the disproportional over-representation of African American students in the EBD category of special education raise objections based on the assumption that special education, on balance, does more harm than good–that identification for and participation in special education is disadvantageous for the students involved.

One more point needs clarification: If an individual is thought to be falsely identified for a service (e.g., if a child’s tooth is drilled unnecessarily by a dentist, an individual is treated medically for a nonexistent disease or disorder, a student is falsely identified as gifted or talented), then nearly everyone agrees that identification or inclusion in the program or treatment is disadvantageous. This is the problem of false positives–false identification. In special education for students with EBD, this is the problem of identifying for special education students who do not actually have EBD. False identification is a necessary and sufficient condition for objections to disproportional over-representation if and only if false positives occur (or are assumed to occur) in numbers sufficient to cause the disproportion.

Responses to Disproportionality

We may respond to knowledge of disproportionality in a variety of ways. Two possible responses attack disproportionality itself, seeking to arrive at proportionality simply by requiring idenfication of more or fewer of the disproportionately represented group.

First, we could place limits on the number of individuals in the disproportionately over-represented group(s), such that fewer individuals in the over-represented group(s) are identified. It is also possible to increase the identification of the disproportionately under-represented group(s), such that more individuals in the under-represented groups(s) are identified. Depending on one’s view of the consequences of being identified, the outcome may be either greater or less social justice. My personal view, based on the existing research on prevalence of EBD, is that the disproportional identification of African American students could be addressed very productively by identifying many more European American students than we now do. I suggest that this would be a good idea for two reasons: First, we have good reason to believe that many more European American students could and should be identified as having EBD; second, on balance, I think identifying students for special education does more good than harm. If I didn’t think so, I’d want to get out of the field of special education asap (I’d rather not work at something that, on balance or typically, does more harm than good, and I’d recommend against funding or promoting such a line of work). Of course, failing to identify more African American students as having EBD does nothing to address under-service to these children, and I don’t think that can actually be justified. All I’m saying is that identifying more European American students is one way, and not a bad way, to address the problem of disproportional over-identification of African American students.

Second, we could claim that whatever tests or other assessments are used to identify problems are biased against the disproportionately represented group (in the case of either over- or under-representation). Coupled with this response is the argument that individuals are misidentified (i.e., that the identification procedure produces many cases that are false positives in the over-represented group). We could also claim that the treatment offered after identification is ineffective, such that identification serves no useful purpose, resulting only in spoiled identity (only in the case of over-representation).

We might examine as objectively as possible all of the possible causes of disproportional representation. Although bias is possible in any case, it is also possible that there are other causes of disproportionate over- or under-representation. In the case of something like dental caries or HIV infection, we are likely to find causal links between such factors as poverty and the condition.

For example, the National Public Radio’s news program All Things Considered for February 7, 2005, carried a segment regarding the over-representation (compared to U.S. citizens of other ethnic groups, particularly those of European descent) of African Americans, especially female African Americans, among those who have HIV/AIDS. As the reporter noted on the program, HIV/AIDS is a stigmatizing disease, so many individuals are reluctant to admit having it or identify others as having it. Because of the stigma involved, and because of the belief that HIV/AIDS and its treatment is a plot of White Americans to further abuse or decimate Black U.S. citizens, attempts at treatment and prevention are complicated and too often rendered ineffective.

Given that the NPR story is correct about the disproportionate over-representation of Black individuals, especially females, in the HIV/AIDS population, would social justice be improved by attempts to limit the identification of persons of African descent in order to reduce their over-representation? In my opinion, certainly not; in fact, I think such nonidentification would reduce social justice dramatically. In my opinion the only socially just way to respond to the disproportional over-representation of African Americans with HIV/AIDS is to (a) offer the best possible treatment to all of those in need, regardless of the proportion of cases represented by their ethnic group and (b) address the factors that lead to HIV infection.

But, of course, identification for special education for EBD is considerably more subjective, and therefore considerably more subject to bias, than is identification of HIV (I know of no claim that testing for HIV is biased, although there may be such).

Needed Research in Special Education

We need to know more about what heightens risk of EBD. Then we need to know whether different groups of students disproportionately experience those risks. If we find that different groups of students disproportionately experience things that put them at risk for EBD, then we need to focus on how to reduce those risk factors.

We also need to know whether disproportional identification is a matter of bias. We must not assume that in the absence of other explanations bias must be the culprit, as that is to blame people by default or out of ignorance of what causes disproportionality (we did that in the case of parents’ behavior and autism, and the results were ugly caricatures of parents). However, if we do find that disproportional representation in special education is a matter of bias, then we need to find ways of eliminating that bias.

From my perspective, there are at least two possible, unacceptable conditions related to the disproportionality issue: (a) kids being identified for services they do not really need because the people doing the assessments or the instruments they use are biased and (b) kids being denied the services they need because they are in a disproportionately over-represented group (i.e., they are denied services because of their group identity).

JK

References

Levine, S., & Wongoren, D. (2005, February 5). Chipping away at poverty’s ill effect: Event gives hundreds of youths a much-needed visit to the dentist. The Washington Post, B1, B10.

Kauffman, J. M. (2004). The president’s commission and the devaluation of special education. Education and Treatment of Children, 27, 307-324.

Kauffman, J. M. (2005). Characteristics of emotional and behavioral disorders of children and youth (8th ed.). Upper Saddle River, NJ: Merrill Prentice-Hall.

National Research Council (2002). Minority students in special and gifted education. Committee on Minority Representation in Special Education. M. S. Donovan & C. T. Cross (Eds.). Division of Behavioral and Social Sciences Education. Washington, DC: National Academy Press.

U. S. Department of Health and Human Services. (2001). Report of the Surgeon General’s conference on children’s mental health: A national action agenda. Washington, DC: Author.

Sphere: Related Content