https://journals.sagepub.com/doi/full/10.3102/0002831221999405.

Lorraine Blatt and Elizabeth Votruba-Drzal investigate how the rapid expansion of school choice – particularly in charter and magnet schools – is correlated with the test score gaps of White-Black and White-Hispanic students. The research paper uses data from the Stanford Education Data Archive (SEDA) and the U.S. Department of Education’s Common Core of Data. The authors argue that the expansion of school choice has negative implications for structural education inequity, which they define as the systematic oppression of Black and Hispanic students alongside the privileging of non-Hispanic White students in public education. For the White-Black test score gaps, Blatt and Vortruba-Drzal find that higher enrollment in charter schools is associated with larger test score gaps for both math and English language arts (ELA). In addition, higher enrollment in magnet schools is associated with larger test score gaps for White-Hispanic students, although this effect is less robust and only significant in certain models explored. The study has identified that this result is linked to a greater racial segregation at the district level, suggesting that these rising test score gaps can be attributed to an increase in segregation when there is an expansion of charter schools. This research, similar to the previous paper, suggests that a possible explanation for achievement gaps is related to school choice. However, where the last paper examined differences between public and private schools, this paper focuses on the differences in public schools (traditional, charter, magnet, etc.). What links these papers is a commonality in the idea that segregation influences academic achievement disparities. Since this paper uses the same dataset we are analyzing, the findings are incredibly relevant to our research. Despite this, it is important to take note of the focus on achievement gaps and how this relates to the previous paper, which criticizes this very focus (and of which much of the education literature falls victim). Despite this, the findings are still relevant and help to paint a clearer picture of this incredibly complex topic. 

https://doi.org/10.1177/2332858420939848.

Andrew D. Ho introduces the Stanford Education Data Archive (SEDA) as a groundbreaking tool for analyzing educational inequality across the U.S., emphasizing its capacity to track long-term trends in achievement by race, income, and geography. He explains how SEDA standardizes and links test score data across states to enable consistent district-level comparisons. His central argument is that while national averages may show modest improvement, vast local disparities persist and require more attention. This piece provides the foundational framework and justification for using SEDA in my project, offering both the rationale and methodology behind the dataset I rely on. It’s less focused on detailed findings than later articles but is crucial for understanding the strengths and limitations of the SEDA data that power both of my visualizations.

https://doi.org/10.1177/2332858419872459.

Heewon Jang and Sean F Reardon examine how state-level contexts influence the relationship between student socioeconomic status (SES) and academic achievement, arguing that the strength of this relationship varies widely by state. They show that some states, through policy or funding structures, moderate SES-based gaps better than others. Their findings suggest that geography matters not only at the district level but also in terms of state-level systems. This article supports my second research question, helping explain why certain high-achieving districts tend to cluster within particular states, as seen in my treemap. It complements the more district-focused approach of the Matheny paper by shifting the lens toward state-level structures that foster or reduce educational inequality.

http://www.jstor.org/stable/3876731.

Gloria Ladson-Billings examines the role that the achievement gap plays in education research, challenging the current literature that focuses on this metric. Ladson-Billings draws an analogy with the concept of the national debt – as opposed to the national budget deficit – to emphasize that there is a current education debt that is composed of years of achievement gaps. She argues that this education debt is a multifaceted problem that can be attributed to historical, economic, sociopolitical, and moral components, which have each contributed to minority groups generally having poorer academic performance compared to their white counterparts. The journal article concludes by explaining why addressing this education debt is crucial if progress is to be made in reducing educational inequity in America. This article helps provide a macro-view of educational disparities in America. Ladson-Billings highlights certain historical events and important policies that have potentially contributed to segregation and the widening of the advantages between racial groups. She also rightfully criticizes the obsession that much of the literature in the education space has with achievement gaps without providing definitive steps to tackle the issue. However, despite stating the importance of having actionable solutions to tackle the education crisis in the US, the article does not offer any tangible suggestions on confronting the education debt.

https://doi.org/10.1080/1369183X.2017.1315851.

Jennifer Lee and Min Zhou challenge the common notion that the reason for Asian-American academic achievement can be reduced to aspects of Asian culture, such as “hard work” or a “passion for education.” The paradox is that Asian Americans, even from various socioeconomic backgrounds, make up a disproportionately large portion of students in prestigious universities, being 6% of the population but as high as 40% of a university’s population, such as at the University of California, Berkeley. In their collection of data, they surveyed various Asian and non-Asian ethnic groups in Los Angeles, later focusing in-depth on Mexican, Chinese, and Vietnamese groups for their prominence in LA’s immigrant and refugee populations. The authors assert that a large yet overlooked influence lies in the U.S. 1965 Immigration and Nationality Act, which tended to be “hyper-selective” toward Asian immigrants who are highly educated. They argue that this explains why there seems to be no distinct difference in the college graduation rates for Asian Americans of different socioeconomic backgrounds. The cultural frames and mindsets historically generalized to Asians were more specifically attained by the highly educated Asians that the law selected toward. The article continues by discussing the education and social psychology effects that the “success frame” has on both in-group and out-group individuals. While this data analysis successfully provides insight into challenging cultural generalizations, the article lacks explanations for how that specific act was selective toward a subgroup of Asians. There is also room for further research into whether the findings are consistent across other U.S. Asian populations outside of LA and non-East Asian ethnic groups. Keeping these questions in mind would allow research to leave the restrictions of historical generalizations and examine how political influences may propagate across subgroups differently.

https://doi.org/10.1037/h0099861.

Murray Levine and Adeline G Levine present a historical overview of events in Black education and analyze data to explain the persisting but narrowing gap in academic achievement between White and Black students. Since the early 17th century, it was illegal to educate slaves, and a small portion of them only eventually attained some form of education to perform skilled operations on the plantations or adopt Christian beliefs (which to owners, increased a slave’s value), or from playing “school” as children with the master’s children. Increasing Black education, which was primarily privately funded, was to improve their skills in low-wage jobs rather than advance them socially. In 1896, the Plessy v Ferguson allowed “separate but equal” schooling, but Black students still experienced unequal treatment. In 1954, the Brown v. Board decision allowed Black teachers to teach in white schools. The researchers compared data about enrollment, length of school year, attendance, attainment, and illiteracy, assuming these as benchmarks necessary to learning. Despite having less of each of these characteristics in their education, Black and White students both show improvement, with an attainment gap that is narrowing between them. The researchers briefly raise other examples of systemic inequity, including financial disadvantages, job opportunities, and unemployment that make socioeconomic status disproportionately work against Black students. They argue that it is unreasonable to expect “quick fixes” to remedy achievement gaps given these historical challenges, and they applaud the gradually narrowing gap despite the systemic disadvantages against Black students.

https://doi.org/10.1076/edre.6.4.337.6935.

Xin Ma evaluates achievement disparities among socioeconomic status and whether varying subject domains have a positive/negative impact on widening/narrowing the socioeconomic achievement gap. The analysis is conducted utilizing collected data covering approximately 6,883 middle school students spanning 148 schools in New Brunswick, Canada, in order to effectively evaluate the structurally rooted and institutionalized nature (hierarchy) of academic achievement disparities by socioeconomic status (SES). Ma refers to studies from Lee and Smith (1995) to infer that there may be widespread inconsistency within the SES achievement gaps across the different subjects, specifically that “within-school SES gaps could be different across subject areas depending on the way a school is organized and operated” (339). Ma emphasizes that the discrepancy is elevated and strongly correlated within-academic discipline areas such as math/science (STEM) and reading/writing (RLA) as compared to the weaker correlation across-academic discipline areas such as between mathematics and writing. Therefore, the specific subject area does have a significant effect on the socioeconomic achievement gap within schools, especially between academic discipline areas such as between reading and math. A key finding of the analysis is that “schools appeared relatively consistent in socioeconomic gaps in the major areas of sciences (mathematics and science)” (352), directly implying that the historical systematic precedent of school organization is disproportionately affecting the lower socioeconomic status students (economically disadvantaged) consistently at disadvantageous levels, distinctively by the varying subject field.  This is explicitly evident, as Ma found through the analysis that elevated levels of parental involvement within STEM subjects such as mathematics further prolong and extend the wide gap in socioeconomic achievement by the proximity to resources at the disposal of affluent families, worsening the inequities and providing context for the average (Math – RLA) achievement discrepancy across specific historically marginalized racial demographics, a key parameter of interest in our SEDA dataset. One major limitation in the applicability of the study is the lack of generalizability to racially diverse, heterogeneous school districts contained within the SEDA dataset in comparison to the study in which “New Brunswick has a homogenous population with a few people from visible minorities” (341). Subsequently, the socioeconomic measurement level (SES) criteria to separate out the “not economically disadvantaged” from the “economically disadvantaged” was solely based on self-reported material resources and participation, including “possession of education-related items at home … and their participation in cultural activities related to education” (342), which inherently “underestimate the socioeconomic effects at the student level” (342) as Ma acknowledges. The fundamental assertion behind Ma’s argument is that these internal socioeconomic achievement gaps are oftentimes resistant to change oftentimes irrespective of interventions. Ma is weary that the usual reform approaches mistakenly focus on educational quality, rather than educational equity – a desire for increased achievement levels at the expense of an increased SES achievement gap, claiming “current school policies and practices may not be formulated for the purposes of reducing socioeconomic gaps in academic achievement” (353). 

https://doi.org/10.3102/00028312221134769.

This study identifies and analyzes major disparities in how academic achievement has changed across U.S. school districts between 2009 and 2018. The authors focus on the cohort slope of test scores to highlight which districts are improving, declining, or remaining stagnant, finding that while some districts show significant progress, others are seeing sharp declines. Using SEDA’s longitudinal measures, they reveal how socioeconomic and regional differences correlate with these trends. This directly informs my first research question about improvement or decline across districts and supports the dual-bar chart I created to visualize top positive and negative growth trends. Unlike broader overviews like Ho’s article, this study dives deep into district-level variation, offering both data and interpretation I can use to explore regional educational change.

http://cepa.stanford.edu/wp17-07.

Richard J. Murnane and Sean F. Reardon investigate the long-term trends in private school enrollments based on family income to see if changes in private elementary school enrollment from 1968 to 2013 reveal any notable trends. Their research uses data from several national surveys. Firstly, they discovered that the gap between private school enrollment rates of high and median-income families (what they term the 90-50 gap) had grown substantially during these five decades, largely due to a considerable drop in the private school enrollment rate of middle-income families, while the rate for high-income families remained stable. This is due to a shift in school types, as the stable enrollment rate for high-income families masks a shift from religious private schools to nonsectarian private schools, whereas middle and low-income families have reduced enrollment in Catholic schools, accounting for much of the growth in the 90-50 gap. Furthermore, conditional on family income, private school enrollment rates are much higher for families in cities than in suburbs. Private elementary school enrollment rates are lower for black and Hispanic families than for white families, but this difference is largely owed to differences in income. They also identify private school enrollment trends that differ between the South & West and the Northeast & Midwest. In both former regions, the private school enrollment rates from high-income families increased, while they declined in the other two regions, resulting in the 90-50 gap increasing much more in the South than other regions. This research paper provides a possible explanation for the disparities that are seen in academic achievement outcomes between racial groups. Whilst Murnane and Reardon point out that income largely accounts for the lower private school enrollment rates for Black and Hispanic families, as we have read in previous papers, all of these factors are strongly interrelated due to historical and sociopolitical reasons. The reasons for the income disparity may be associated with racially discriminatory policies that have negatively impacted generations of minority groups. While this paper does offer many useful insights, there are several limitations that the authors have made clear. Firstly, they acknowledge the lack of an ideal dataset, with certain parts of the dataset having few data points available. They are also unable to judge the relative effectiveness of different school types and how this has changed over time. In addition, in comparison to the dataset that we chose to explore, there are only a few grades that overlap, since this research paper only covers elementary school enrollment (which encompasses grades 1-5, whilst our dataset spans 3-8). This limitation must be acknowledged when this secondary source is used to support possible explanations or hypotheses that are proposed as causes for educational disparities.

https://doi.org/10.1016/j.stueduc.2024.101367

Max Nachbauer evaluates the influence of historically structured school teaching practices and school-based extended day programs on both the average student academic achievement, denoting the education quality, and the differences in achievement based on socioeconomic status, denoting the education equity, simultaneously, throughout German secondary schools. The overarching mission of Nachbauer’s analysis is to determine the specific educational practices & school-based interventions which directly improve educational achievement outcomes on average, while also mitigating (narrowing) the disparity in achievement levels across students from varying socioeconomic backgrounds concurrently. As such, certain key instructional practices such as effective cognitive activation which is measured as “assignments that the teacher gives to students (… assignments that require time to think)” (6) and efficient classroom management which is defined as “proportion of time that the teacher on average spends with maintaining order and dealing with interruptions” (6), generally tend to enhance overall student academic achievement on average, meanwhile other prominent school-based teachings such as heterogeneous grouping in classrooms with “groups of students with different achievement levels” (6) across varying socioeconomic statuses (high & low) has directly assisted in reducing/narrowing the disparity in achievement levels across the differing socioeconomic statuses. A key distinction is also made by Nachbauer in which he demonstrates that certain extended-day programs such as enrichment programs alongside homework supervision generally tend to not only diminish overall student academic achievement on average, but furthermore they also widen (exacerbate) the socioeconomic-based achievement gap despite the preconceived notion of narrowing/reducing the gap, whereas tutoring in comparison, instead effectively narrows/reduces the socioeconomic achievement gap in the strive towards educational equity. The major limitation in the applicability of the study in a broader, generalizable context of educational equity, is due to the dependency of self-reported data stemming entirely from “teacher questionnaires or principal questionnaires” (10) directly. This introduces a potentially significant layer of inherent/intrinsic bias on both an individual level in terms of the teaching staff intending to only offer a “one-sided perspective on the school learning environment” (10) in order to maintain their reputation and on a systemic level within the schooling system in preserving a cleaner image by framing a narrow socioeconomic-based achievement gap in disparity, putting academic integrity and falsifiable results into question. The variation in certain school-level classroom practices such as cognitive activation and classroom management when accounting for the unavoidable differences in conditions/procedures across geographic school districts over the states will similarly have an adverse effect on the reproducibility and generalizable applicability of the educational equity in context. 

https://doi.org/10.1016/j.ssresearch.2023.102938

The authors of this article explore the correlation between inaccurate teacher judgements of primary school student achievement and gender-based achievement gaps across subjects. In England, Germany, and the U.S., girls generally outperform boys in reading than boys, and boys generally outperform girls in math. They review current findings that biased teacher judgements based on gender stereotypes can lead to verbal feedback and behavior that produce self-fulfilling prophecies. To collect new data on the topic, they had teachers rate students on a scale from 1 to 5 on general and written language skills and math skills compared to other students of the same grade or age. The researchers then compared these to each country’s various tests and tasks, including nonverbal cognitive abilities and working memory, as an objective measure. Family characteristics were also taken into account, such as the highest education of parents, how many parents lived with the student, immigration status of the parent(s) or student. The results confirmed the previous theories, where teacher judgements of language skills were an overestimation for girls and an underestimation for boys, while judgements of math skills had an opposite pattern. Consistently in all three countries, teacher judgements were also able to predict students’ achievement at the end of primary school, demonstrating the theory of a self-fulfilling prophecy. Since this study relies on correlational data analysis, it is difficult to determine if there may be confounding variables. Examples of these unknowns, which the researchers acknowledge, include the degree of gender inequality in a society and educational accountability approaches.

https://doi.org/10.1080/19345747.2024.2342850

Tyler A Womack, Gregory J Palardy, and Soojin Park evaluate how the intersectionality in disparities across race, gender, and socioeconomic class for mathematics achievement specifically in US high schools are foundationally regulated and shaped by two specific systemic, structural school-based factors, being the opportunity to learn (OTL) alongside the school socioeconomic status (SES). The authors are interested in determining the effect of equalizing either one of these two particular institutionalized factors at the intersection of the various racial, gender, and socioeconomic groups, and whether a natural consequence of such implementation would entail a significant reduction in educational achievement disparity. The potential resolution of equalizing either the Instructional Opportunity to Learn (OTL) “between marginalized and non-marginalized groups” (17), which is categorized by “whether or not students have had an opportunity to study a particular topic” (2), or the School Socioeconomic Status (SES) which is measured as “PISA’s index of economic, social, and cultural status” (8), across the students of all demographics, has an intrinsically beneficial outcome in reducing achievement disparities in mathematics, however this will unequally affect each group. The data exemplifies this such disparity most prominently in widespread improvement with Hispanic students and Black students across both genders noticeably benefiting positively from the “equalizing school SES across intersectional groups” (20), in addition to the White students benefiting more positively as compared to the Asian students from the increased accessibility in opportunity to attend higher-level mathematics instruction. The racial and gendered achievement gap is reversed (increased) most noticeably (largest effect) to the detriment of Asian Female students by 124% of an increase, while in comparison, equalizing the School Socioeconomic Status (SES), results in a reduction (decrease) most noticeably in favor of both genders of Hispanic students (26.6% for females, 25% for males). Therefore, a compromise based resolution would require opposing those universal policies that are targeted across students of all demographic groups, and instead effectively focusing on racial and gendered specific school-based interventions. One focal limitation in the applicability of this study in a broader context is the origin of the data. The data is collected from an international cross-section exam (PISA 2012), essentially restricting the possibility of causal inferences/analysis. An additional pertinent limitation of this study is regarding the lack of inclusion of differing subgroups within the racial demographics, excluding “native americans”, “other”, and “biracial” / “multiracial” students as these categories had “small samples with low statistical power” (7). The dataset instead only includes black, white, hispanic, and asian students, therefore not offering the perspective of a wider array of diversities and demographics, effectively narrowing the scope of the generalizability.