It is too often the case that media outlets will cite a research paper, cherry-pick information, and make misleading claims about what the findings entail.
Unfortunately, this is precisely what our friends at NC Policy Watch have done in their recent piece about a study from Duke University Sanford School of Public Policy that examined group segregation within North Carolina schools. The article provides a synopsis about specific information denoted in the study; however, it leaves out relevant information about the nature and extent of the study. Furthermore, a closer inspection reveals their proposed policy recommendations, insofar as they are interested in predicating their recommendation on the study, completely misses the point of the study. This is why the piece is grossly misleading: their conclusion for what should be done does not follow from the study.
The researchers explicitly express their aim with the study. They pose the following research questions: “the extent of racial and ethnic segregation in North Carolina over time, with special attention to the relationship between segregation within and between schools…how white/black segregation compares to white/Hispanic segregation…whether there exist systematic differences in the courses taken by students in different racial and ethnic groups.” In the pursuit of answering the aforementioned questions, the research team found that “segregation at the classroom level was consistently higher between white and Hispanic students than between white and black students.” This is mainly found to be the case in middle and high schools, but not elementary schools. The research team also found “black and Hispanic students were less likely to be enrolled in advanced courses in 7th and 10th grade, particularly math courses…they also tended to take less rigorous courses than their white peers.”
Now, I will briefly mention some relevant information about the study that was not mentioned by NC Policy Watch.
The study was designed around middle and high school students that took English and Mathematics courses. To measure group imbalance (i.e., group segregation) in North Carolina schools, they used a dissimilarity index. This method is used for finding an imbalance between two data sets, and it is commonly used in the social sciences to find disparities. Alternatively, the team could have used a difference-in-differences (DiD) approach to measure the imbalance. However, the results between the two models in terms of finding group disparity would have been nominal. Thus, the model selected by the team was appropriate for their descriptive study.
While the model used to measure group imbalance is valid, it is a superficial measure. That is to say, it does not tell you anything other than differences exist. Unfortunately, NC Policy Watch wishes to tee up the study to suggest systematic discrimination of minority students is the reason for the imbalance. However, the study’s measurement cannot speak to discrimination, and thus the results are merely descriptive. Disparity does not imply discrimination.
The research team did not perform hypothesis testing but instead focused “primarily on the analysis of basic empirical questions related to segregation in public schools.” Additionally, the research team did not investigate the students’ socioeconomic backgrounds. While these are meaningful considerations that, I believe, would provide better insights into the root cause of group imbalance, it is not required to answer the research questions established by the team at the onset.
Indeed, the results demonstrate there is evidence of within-school group imbalance. However, NC Policy Watch seeks to advocate policy changes despite the study’s limitations. They are making a hasty judgment about what should follow to address the patterns noted in the study. It is more likely that the within-school group imbalance is due to socioeconomics—assuming our aim for a balanced classroom is proportional to group population sizes—in which case, the type of policy changes that they are promoting would not solve anything.
However, NC Policy Watch is all too eager to pick up where the research team put their pencils down. They argue, “[e]ducators, policymakers, and other stakeholders must continue to push for policies and practices that harbor real integration at the classroom level.” Their solution is something called the 5Rs of Real Integration, which is an approach adopted in NYC. The policy recommendation, which they call “truly integrated,” calls for the following:
- Schools must provide a diverse and inclusive environment.
- Funding systems must provide equitable distribution of resources across schools.
- Schools must be considerate and empathetic of the identities of all students, focus on the power of different backgrounds, and build relationships between students across group identities.
- All high schools should be free of police and military presence, safe, free of metal detectors, protective of the integrity and humanity of each student, and help build student leaders.
- School faculties should be inclusive, elevating the voices of communities of color, immigrant communities, and the LGBT community so that student identities and experiences are reflected in the leadership.
I read this, and none of these recommendations address why black and Hispanic students are not enrolling in advanced math courses. (The LGBTQ community was not even mentioned in the study.) NC Policy Watch claims black and Hispanic students “continue to be denied the same curriculum and opportunities being offered to white students.” But this is not in the study. Moreover, AP and honor courses are an election made by kids and their parents. This is not something schools push on some children over others. What drives kids to be in courses like advanced mathematics and honors English comes from the individual student and parental involvement in the child’s education.
In fact, none of these bullet points will lead to more diversity within a given school. What influences diversity in a school is the ability to spread out a diverse finite population with variation in group sizes such that you tend to have proportionate representation within a given school. It is effectively a constraint problem that is influenced by choice and socioeconomics. Hence, throwing money at schools will not reduce group homogeneity in schools.
What the study is adverting and what NC Policy Watch is demanding are two separate things. The research team expressed that their findings indicate current policies designed to reduce segregation are not working. However, NC Policy Watch wants to continue with the same old rhetoric-driven policy approaches to address something that has nothing to do with skin color or ethnicity.
Joshua Peters is a philosopher and social critic from Raleigh, NC. His academic background is in western philosophy, STEM, and financial analysis. Joshua studied at North Carolina State University (BS) and UNC Charlotte (MS). He is a graduate of the E.A. Morris Fellowship for Emerging Leaders.