Despite the many debates occurring within the education space, everyone can agree that students should come first. Within data work, putting students first means performing comprehensive, unbiased, and detailed analyses of data to understand how schools impact their students. Some of the most important factors to do this research include accessing the best data possible, making logical comparisons at the proper level, and having diverse data and research teams.
First and foremost, we need to acknowledge that not all data is created alike and the availability of data can often be limiting. Each state publishes differing amounts of data with various levels of specificity. Some states intentionally misuse and misinterpret FERPA and other data privacy policies to keep more data suppressed than necessary and others simply don’t release important data even if the data is in compliance with these policies. As of today, only about half of states have released school-level information on the number of students receiving free or reduced price lunch in 2016-17 and less than a third of states have released school level data on students with disabilities. These policies of withholding, delaying, and over suppressing data prevent researchers from producing and publishing the best possible analyses and policy recommendations. Researchers need thorough and disaggregated data to account for and advocate for the most disadvantaged students. With detailed data, we can continually identify and support groups of students struggling the most in each school and hold the school – district or charter – accountable for providing a top education to all students.
Each state has a different landscape for education; this is especially true for public charter schools—in which Montana doesn’t have any, Missouri limits them to two districts, and California has over 1,200. Thus, it is important to recognize the profound role that state landscape and policy context plays on interpretation and analysis. One of the strongest ways to assess charter schools, for instance, is the method from The Center for Research on Education Outcomes (CREDO) where they compare charter school students with similar district public school students. Comparisons between schools, school types, or districts need to take into consideration the best level of analysis. Broad claims on the state of education in the U.S., or even within individual states, can sweep the students most in need under the rug. Further, turf battles between charter school and district school proponents distracts from the most important analytical work. We need to focus on helping schools push each other to serve all students, increase student satisfaction, lower expulsions, and achieve better graduation rates. Despite the sometimes divisive educational rhetoric, researchers have a responsibility to make logical, thoughtful comparisons and analyses instead of partaking in detrimental educational partisanship.
Responsibility does not begin and end with how one analyzes the data, it’s also about who is on the team of analysts. Across numerous professions, studies have shown that diverse teams usually perform better than homogenous teams. The benefits of a diverse team within the research world go even further. Far too often, statistics are taken at face-value. However, every dataset is construed and interpreted through each team members’ reading of the data and analysts inherently bring their own perspectives and thought processes to the table. In addition to being more likely to form a comprehensive analysis with multiple understandings, a diverse team will also be more likely to check one another’s assumptions. Furthermore, the student population in the U.S. continues to become more diverse. In 2014-15, the percentage of white students in U.S. public elementary and secondary schools dropped below 50 percent for the first time, according to the National Center for Education Statistics (NCES). Increasingly in education there are calls for more teachers of color. This is a start—but we need to go beyond this. We need more diversity at all levels and within all systems of the education sphere, not just within the classroom. The need for diversity spans beyond just race though and includes class, gender, and political orientation. The interpretation of education data impacts every student in this nation, so it is imperative that these analyses do not come from a single-axis of understanding. We should strive for diversity not just in the classroom, but on the teams of analysts that help form perceptions of what occurs in public education.
Data and research are playing an ever-increasing role in education policy: it’s our duty as researchers to ensure these analyses are timely, nuanced, and high-quality. To do this properly, we rely on states to collect and publish comprehensive datasets that are disaggregated by demographic group, socio-economic status, and grade level (among other important variables). Similarly, the most beneficial reading of this data will account for differences in geographic location, school type, grade level, socio-economic status, and race/ethnicity. Therefore, it is only logical that top data teams should include team members from different backgrounds and with different perspectives. In all data work nuance is essential, but in education that nuance represents individual students, thus it is crucial that strongest analyses are being conducted and published.