Analyzing the Results of a New National Study: Why Some “Two-Dimensional Problems” Need “Three-Dimensional Thinking”
Dear Colleagues,
Introduction
Every day, I get e-mail briefings discussing or analyzing a wide variety of world, national, and professional issues, reports, or perspectives. One of my favorites is Smartbrief’s “While You Were Working” (WYWW).
Last month, WYWW had a story about an on-line personality assessment questionnaire with an invitation to take it.
While I typically pass on these opportunities (why would a psychologist want to know “why he ticks”? . . . actually, I’m just joking), I jumped on this one.
Fifteen minutes later—after completing the questionnaire using my “work persona,” I had my “psychological profile.” It included the following descriptors:
You like to think in plain and simple terms. Others describe you as down-to-earth, practical, and conservative. You do not react with intense emotions, even to situations that most people would describe as stressful. You set clear goals and pursue them with determination. People regard you as reliable and hard-working. You have a strong interest in others' needs and well-being.
If anything, relative to my work as an organizational and school psychological consultant in districts across the country (and world), I think that I am a research-to-practice straight-shooter.
That is, while it is never my goal to offend, when the facts are clear, I am not afraid to share these facts—even if they are counter to someone else’s psychoeducational beliefs, actions, practices, or interventions. Indeed, as a consultant, I truly believe that it would be inappropriate for me to withhold my opinions—for example, with a superintendent—even though I know said opinions may be counter to those of the superintendent him or herself.
As a bottom line: I am an advocate for students—working to ensure that their academic and social-emotional outcomes are maximized from preschool through high school—and beyond. I believe that I need to be “true” to these students—many of whose “voices” are not heard in their classrooms, within their schools, or across their districts.
_ _ _ _ _ _ _ _ _ _
Describing a Recent National Study Investigating Racial Discrepancies in Achievement and School Suspensions
In this context of advocating for all students, I want to review and analyze a study published last month on October 15, 2019 in AERA Open, a peer-reviewed journal of the American Educational Research Association.
The article was titled, “Are Achievement Gaps Related to Discipline Gaps? Evidence from National Data.”
[CLICK HERE for original article]
But then, I want to connect the results of this study with a topic that I have discussed in a number of Blogs this year: The Funding Inequity between Majority Black and Majority White schools.
While a “hidden” variable in the “Achievement Gap” study, I believe that the inequitable funding between schools that serve mostly white versus mostly students of color may triangulate with the racial achievement and racial discipline gaps found. In fact, I believe that it may be a primary predictor in the achievement and discipline gaps discussed in the AERA Open article. Indeed, in my past Blogs, we have already discussed how inequitable funding correlates with some of the negative outcomes for students of color—especially African-American students.
_ _ _ _ _
The Achievement/Discipline Gap Study in a Nutshell
This study, authored by Pearman, Curran, Fisher, and colleagues at Stanford University, the University of Florida, the University of Louisville, and Drexel University, analyzed the correlations between academic achievement and school suspension/disciplinary gaps for White, Hispanic, and Black students—from third through eighth grade—from over 2,000 school districts across the country. The data were pulled from the U.S. Department of Education’s Civil Rights Data Collection and the Stanford Education Data Archive for the 2011-12 and 2013-14 school years, and the districts were chosen to be nationally representative.
Overall, the statistical analyses showed that:
- Students of all races faced higher suspension rates in districts with bigger racial achievement gaps—but they took a particular toll on black students.
For every 10 percentage-point increase in a district’s gap in math and reading performance between white and black students, there was a 30 percent larger black-white gap in suspension rates than the national average for similar districts.
Likewise, a school district with a 10 percentage-point wider disparity in suspensions between black and white students would have a black-white achievement gap that was 17 percent larger than the average for similar districts nationwide.
_ _ _ _ _
- For every 1 percentage point increase in the discipline gap, there was a 0.02 standard deviation increase in the Black-White achievement gap, and a 0.03 standard deviation increase in the Hispanic-White achievement gap.
This means that the racial achievement gap in a district that suspended 10% of its Black or Hispanic students but only 5% of its White students would be predicted to have a Black-White and Hispanic-White achievement gap that was 0.10 and 0.15 standard deviations larger, respectively, than a district that suspended the same proportion of minority and White students.
_ _ _ _ _
- Academically, the racial discipline gap in a district whose White students scored, on average, 1 standard deviation higher than its minority students would be predicted to have a Black-White and Hispanic-White discipline gap that was 3.67 and 1.24 percentage points larger, respectively, than a district whose minority and White students achieved at similar levels academically.
Districts with higher levels of achievement for Black students had lower suspension rates for Black students.
_ _ _ _ _
- Even after controlling for unobserved district-level characteristics (see a discussion of these below), a 1 percentage point increase in the Black-White discipline gap was associated with a 0.01 standard deviation increase in the Black-White achievement gap.
In other words, two districts that were otherwise equivalent on observable and time-invariant unobservable characteristics but that suspended differing shares of Black relative to White students would also differ, on average, in racial achievement gaps, with the achievement gap being larger in the district that suspended greater shares of Black relative to White students.
Moreover, a 1 standard deviation increase in the achievement gap between Black and White students is associated with a 2.2 percentage point increase in the Black-White discipline gap after accounting for observable and time-invariant unobservable differences between school districts.
_ _ _ _ _
- One reason for the positive adjusted association between the Black-White discipline gap and the Black-White achievement gap is that Black students perform poorer, on average, in districts that suspend them at elevated rates compared with their White counterparts. There is no evidence, however, that the racial discipline gap is predictive of White students’ achievement.
In other words, the Black-White discipline gap’s positive relationship with the Black-White achievement gap is attributable, in part, to the Black-White discipline gap being predictive of lower achievement for Black student—but unpredictive of White students’ achievement.
_ _ _ _ _
Significantly, Pearman and his colleagues were very aware of the district, student demographic, and other variables that might confound the results of their study. Thus, relative to the results documented above, they statistically controlled (to neutralize) these different variables—in essence, factoring their potential influence (or bias) out of the empirical analyses.
Indeed, the authors stated:
(P)rior research has shown that out-of-school factors, especially those related to poverty, influence student achievement and disciplinary problems, with poverty-related factors being inversely related to student achievement and positively related to disciplinary problems (Gregory et al., 2010). Consequently, this study controls for a set of factors at the community level that approximate the amount of disadvantage within district neighborhoods.
The following characteristics were gathered from the 2009–2014 American Community Survey and aggregated to the level of the school district: median income, percentage of adults with bachelor’s degree or higher, percentage of households that are female headed, unemployment rates, and percentage of residents living at or below the federal poverty line.
It is also possible that discipline and achievement disparities may arise as the result of a broader racial climate in a school district that disproportionately favors White students over Black or Hispanic students, such as attitudes that lead teachers and administrators to differentially suspend students by race or that lead teachers to differentially cater to the academic needs of one group of students over the other (Mattison & Aber, 2007).
This study therefore controls for two proxies of a district’s broader racial climate: differential special education assignment by race and differential gifted and talented assignment by race.
In addition to differential treatment by race, a potential relation between racial achievement gaps and racial discipline gaps might also be confounded by racial disparities in other key educational inputs. For instance, discipline gaps and achievement gaps may arise because students of color are disproportionately of lower income or because they are in bigger classes.
Consequently, we also control for racial differences in free and reduced-price lunch status, and racial differences in student-teacher ratios.
_ _ _ _ _
As is evident in the results above, the links between suspensions and test performance remained significant for black students even after the researchers controlled for other district characteristics, such as parents’ education levels, the concentration of poverty among students, and the level of racial segregation among districts.
_ _ _ _ _
Summarizing the Study
The authors summarized the results as follows:
We find evidence that districts with larger racial discipline gaps have larger racial achievement gaps (and vice versa). Though other district-level differences account for the positive association between the Hispanic-White discipline gap and the Hispanic-White achievement gap, we find robust evidence that the positive association between the Black-White discipline gap and the Black-White achievement gap persists after controlling for a multitude of confounding factors (e.g., parents’ education levels, the concentration of poverty among students, the level of racial segregation among districts). We also find evidence that the mechanisms connecting achievement to disciplinary outcomes are more salient for Black than White students.
We also found evidence that the association between the Black-White achievement gap and the Black-White discipline gap was attributable, in part, to the tight coupling of achievement and discipline for Black students in particular, who experience higher suspension rates in districts with larger achievement gaps and who experience higher achievement in districts that suspend them less frequently. Notably, this tight coupling of discipline and achievement was not observed for White students. This pattern indicates that the mechanisms connecting achievement and discipline (e.g., teacher biases, peer effects, feelings of belonging) are more salient for Black than White students.
While the results of our study do not speak directly to solutions to discipline and achievement gaps, they do suggest that interventions aimed at addressing one gap may have potential to influence the other. First, recent research has found that teacher professional development focused on improving instructional capacity and other classroom practices improves classroom behavior and reduces racial disparities in discipline (Gregory, Allen, Mikami, Hafen, & Pianta, 2015), which may be particularly the case for curricular and pedagogical approaches deemed successful at raising the academic achievement of ethno-racial minorities, such as culturally relevant teaching and ethnic studies programs (see, for instance, Dee & Penner, 2017; Ladson-Billings, 1995; Milner, 2010). Second, a body of evidence is emerging on alternative disciplinary practices that can reduce the time students spend out of the learning environment.
_ _ _ _ _ _ _ _ _ _
Show Me the Money: Does Discrepant Black-White School Funding Triangulate with These Achievement and Disciplinary Results?
The Pearman and Colleagues study is both impressive and significant to our greater understanding of the relationship between academic achievement and disciplinary discrepancies among white students and students of color.
And yet, it did not seem to emphasize a third, triangulated factor—the funding discrepancies between majority white versus non-white schools.
To make sure that I was not misreading the article, I e-mailed Dr. Pearman to ask if I missed anything in the treatment of school funding in their analyses.
Here is his response:
Thanks for the note, Howie—and thanks for your interest in the study.
Our study did not focus on spending so I would be cautious about extending our findings in that direction (we would have modeled things a little differently had spending been a focus).
Nevertheless, I think you can find a partial answer to your question in Columns 6 of Tables 1-4, which provides coefficient estimates for included covariates in the preferred model (district FEs), including per pupil spending.
Bear in mind that because we didn’t focus on spending, we didn’t do anything with it besides simply control for it. Consequently, the only thing the models reveal is whether school spending is related to either the discipline gap or the achievement gap, but not whether school spending is related to the relationship between the discipline gap and the achievement gap, which sounds like what you may actually be interested in.
At any rate, school spending does show up as significantly (and negatively) related to the Black-White and Hispanic-White achievement gap, as might be (optimistically) expected. Interestingly, and somewhat counterintuitively, school spending shows up as positively related to the Black-White discipline gap, as shown in Column 6 of Table 3, despite being negatively related in Models 1-5. Unfortunately, I have not thought much about this coefficient (it wasn’t of interest in our study), so I do not have any theories for this pattern at the moment.
_ _ _ _ _
So. . . in essence, Pearman’s research found that, as school spending increased in the districts that they studied: (a) Black-White and Hispanic-White achievement gaps decreased; and (b) Black-White discipline gaps increased.
And yet, while Pearman’s research controlled for school spending, it did not—as he noted—differentiate between and explicitly analyze the funding gaps that have been previously established between majority white and majority students of color schools (see below).
Had they run these analyses, Pearman and his colleagues may have confirmed some of the research-to-practice that we have previously discussed.
_ _ _ _ _
Summarizing What We Know about the Impact of School Funding Discrepancies
Earlier this Spring, near the 65th anniversary of the landmark Brown v. Board of Education case, I wrote two Blog messages addressing the national issue and reality of how students and schools are inequitably funded relative to their students’ psychoeducational and multi-tiered academic and behavioral needs.
One of the “bottom lines” discussed was that:
While segregated educational facilities were deemed by the Supreme Court to be inherently unequal, the quality of instruction and the availability of resources and money in today’s schools—for many students from poverty and students of color—is unequal.
Indeed, at the root of this statement was an April, 2019 Report by the Shanker Institute, The Adequacy and Fairness of State School Finance Systems. This Report demonstrated that there is a cumulative state education funding gap in this country of $23 billion per year favoring white over non-white districts—a gap that is experienced by approximately 12.8 million of our nation’s students.
[CLICK HERE for Part I of this Series:
Solving Student Crises in the Context of School Inequity: The Case for “Core-Plus District Funding” (Part I). When Schools Struggle with Struggling Students: “We Didn’t Start the Fire.”]
_ _ _ _ _
[CLICK HERE for Part II of this Series:
The Journey toward Real School Equity: Students’ Needs Should Drive Student Services … and Funding (Part II). The Beginning of the Next School Year Starts Now: The “Get-Go Process”
_ _ _ _ _
In Part I of this Spring Series, we provided data from a number of sources showing that high-poverty non-white schools in this country receive significantly less money per pupil each year than high-poverty white schools and middle or upper class dominated schools, respectfully. While this involves approximately 12.8 million students—many of them attending schools in urban settings—this is a nationwide problem.
Because of the financial inequity, these high-poverty schools have fewer resources than middle or upper class-dominant schools, and they are typically staffed by less experienced teachers who have more skill gaps, and who resign from their schools more often and after fewer years in-rank. In addition, the students in these schools typically have less access to high level science, math, and advanced placement courses, and less access to needed multi-tiered academic and social, emotional, and behavioral services, supports, programs, and interventions.
Correlated with the poverty, many of these students exhibit health, mental health, academic, and social, emotional, and behavioral challenges, that also triangulate with stress and trauma—including the impact of hunger and poor nutrition, parental incarceration and loss, abuse and neglect, and the exposure to violence and drugs.
From a school perspective, all of this translates into lower numbers of academically-proficient students, and schools that are either in their state’s ESEA-driven school improvement programs or that are rated at the low end of their state’s school report card scale.
From a student perspective, all of this translates into negative effects on students’ school attendance and expectations, classroom engagement and motivation, academic readiness and proficiency, emotional self-control and prosocial interactions and, ultimately, their high school graduation and readiness for the workforce.
Part I ended with a plea for systemic changes relative to federal, state, and district funding policies, principles, and practices. We recommended a “Core-Plus Funding” process whereby all schools in a district receive the core funding needed for student success, but where the schools with additional or significant student needs receive, annually, the additional funds and resources needed for their success.
_ _ _ _ _
In Part II of this Series, we discussed a new report showing the impact of poverty and the importance of inequitable school funding, and revisited our Part I recommendation that districts and their school utilize a Core-Plus Funding process.
Core-Plus Funding is when every school in a district receives the “Core” funding that it needs to implement a sound educational program for its students. The “Plus” funding involves the additional funds that each school receives based on the intensity of the academic and social, emotional, and behavioral needs of its students.
The new report was released on May 15, 2019 by Georgetown University’s Center on Education and the Workforce. The report was titled: Born to Win, Schooled to Lose: Why Equally Talented Students Don’t Get Equal Chances to Be All They Can Be.
[CLICK HERE for this study]
This study used data from a number of national longitudinal databases to investigate the impact of students’ socio-economic status in kindergarten on their college and career outcomes. Among the databases used were: the U.S. Census Bureau’s American Community Survey, the Consumer Expenditure Survey, the National Longitudinal Study of Adolescent to Adult Health (to determine students’ socio-economic status); and the Early Childhood Longitudinal Study: Kindergarten Class of 1998-1999, and the Education Longitudinal Study of 2002.
The Early Childhood Longitudinal Study’s 1998-1999 kindergarten cohort involved a representative sample of 21,260 kindergartners from across the country. The Education Longitudinal Study of 2002 sampled a nationally representative group of over 15,000 10th graders from over 750 schools beginning in 2002.
This study is important to our discussion of inequity because it (a) demonstrates that socio-economic inequity for students in kindergarten has dire, long-term effects; and (b) that schools with large numbers of poor students (many of whom are students of color) need more (not fewer—as shown in Part I of this Blog series) resources and funding to address these students’ needs.
Here are the Key Findings from the Born to Win, Schooled to Lose study.
1. In America, it is often better to be rich than smart. Among the affluent, even a kindergartner with test scores in the bottom half has a 7 in 10 chance of reaching high SES among his or her peers as a young adult. But for similarly talented White, Black, Latino, and Asian children from low-SES families, the meager material supports available along the way to adulthood subvert nature’s generosity. Across racial and ethnic groups, a disadvantaged kindergartner with test scores in the top half has approximately a 3 in 10 chance of being high SES by the age of 25.
2. Even at an early age, environmental disparities by class, race, and ethnicity are evident in measures of children’s achievement. Only about a quarter of lowest-SES kindergartners have top-half math scores, compared to around three-quarters of highest-SES kindergartners. Children’s early scores also vary by race, in part because Black and Latino children are twice as likely as White children to come from lowest-SES families.
3. As children progress through primary school, they can improve on measures of achievement, but their chances of improvement correlate to their class status. Becoming high achieving is less likely for low-SES kindergartners with bottom-half math scores. By the eighth grade, fewer than 1 in 5 lowest-SES kindergartners with bottom-half math scores will score in the top half, compared to more than 2 in 5 highest-SES kindergartners with bottom-half math scores.
4. A child from an advantaged class is more likely to maintain high scores than one from a poor family, and White and Asian children are more likely to do so than Black or Latino children. For low-SES students with top-half math scores, staying at the top throughout their academic journeys is difficult. In addition, Black and Latino students with top-half math scores in kindergarten are less likely than their White and Asian peers to persist in earning top scores.
5. Achievement patterns are largely set by the time children enter high school. This is particularly evident for students with the lowest scores: students with bottom-quartile scores have difficulty improving their scores once they reach high school. Most tenth graders who score in the bottom math quartile will still score in the bottom quartile in twelfth grade.
6. High school achievement sets the stage for college attainment—but family class plays an even greater role. The highest-SES students with bottom-half math scores are more likely to complete a college degree than the lowest-SES students with top-half math scores.
7. Class mobility in America is limited—but education can be a lever for change. The lowest-SES tenth graders with top-half math scores are twice as likely to become high-SES (top-half) young adults as their peers with bottom-half math scores. Disadvantaged students who show promise can achieve, but their chances are better with interventions—and while lowest-SES tenth graders with bottom-half scores can become high SES, their chances are very slim.
And here are some additional data from the study:
- According to report, more-affluent students are often provided with more resources both in and out of school, which may benefit their education.
- White and Asian tenth graders are more likely than Black and Latino tenth graders to earn a college degree in 10 years, no matter their high school math scores.
- Most tenth-graders who score in the bottom math quartile maintain their low grades through twelfth grade.
- Improving scores in high school is uncommon, but highest-SES students are twice as likely as lowest-SES students to move into a higher math quartile.
- Lowest-SES tenth graders with top math scores are less likely to immediately enroll in a college than highest-SES tenth graders with bottom math scores. These lowest-SES tenth graders also are less likely than highest-SES students to complete college 10 years later—regardless of their high school math scores.
- Affluent children with low test scores have a 71% chance of becoming affluent adults at age 25, while poor children with high test scores only have a 31% of chance of becoming wealthy in adulthood.
- The disparity becomes more severe when broken down by race. Fifty-one percent of black and 46% of Latino 10th graders with high math scores were more likely to earn a college degree within 10 years than similar students with low scores, but they were still less likely to early a college degree than their white and Asian high-scoring peers. Among the latter groups, 62% and 69%, respectively, received degrees.
_ _ _ _ _
These results crystallize the long-term impact of poverty. And while the Georgetown University report did not analyze whether low-SES students attending better resourced schools had better outcomes, past research suggests that quality education and sound multi-tiered supports can mediate the impact of poverty. This, then, reinforces the critical need for funding for schools by student need when they have high numbers of students from poverty.
_ _ _ _ _ _ _ _ _ _
Summary
High-poverty non-white schools in this country receive significantly less money per pupil each year than high-poverty white schools and middle or upper class dominated schools, respectfully. This disproportionate funding affects approximately 12.8 million students.
Because of this financial inequity, these high-poverty schools have fewer resources than middle or upper class-dominant schools, and they are typically staffed by less experienced teachers who have more skill gaps, and who resign from their schools more often and after fewer years in-rank. In addition, the students in these schools typically have less access to high level science, math, and advanced placement courses, and less access to needed multi-tiered academic and social, emotional, and behavioral services, supports, programs, and interventions.
In their October 15, 2019 AERA Open study, Pearman and Colleagues found that:
- Districts with larger racial discipline gaps had larger racial achievement gaps (and vice versa).
- The positive association between the Black-White discipline gap and the Black-White achievement gap persisted after controlling for a multitude of confounding factors.
- The association between the Black-White achievement gap and the Black-White discipline gap was attributable, in part, to the tight coupling of achievement and discipline for Black students in particular, who experience higher suspension rates in districts with larger achievement gaps and who experience higher achievement in districts that suspend them less frequently.
- This tight coupling of discipline and achievement was not observed for White students. This pattern indicates that the mechanisms connecting achievement and discipline (e.g., teacher biases, peer effects, feelings of belonging) are more salient for Black than White students.
And while Pearman and Colleagues controlled for school spending in their study, they did not differentiate between and explicitly analyze the funding gaps established in the Shanker Institute and Georgetown University studies, respectively.
Nonetheless, Pearman’s research did find that, as school spending increased in the districts that they studied: (a) Black-White and Hispanic-White achievement gaps decreased; and (b) Black-White discipline gaps increased.
But until they re-analyze their data to differentiate the funding in their majority white versus majority students of color districts, we will not know whether there is a significant funding-achievement-discipline triangle, and whether the seemingly contradictory results immediately above are real or an artifact of the analyses conducted.
_ _ _ _ _
What we do know from the entire discussion above is that:
- There is a relationship between the White-Black achievement and discipline gaps (Pearman and Colleagues);
- There is a significant funding discrepancy between majority white and majority students of color schools (Shanker Institute);
- The funding gap does significantly impact the academic achievement of students experiencing socio-economic inequity who attend schools with large numbers of poor students; and
- The funding gap correlates with factors whereby students demonstrating social, emotional, and behavioral (including disciplinary) challenges are not receiving the multi-tiered prevention, strategic intervention, and intensive need/crisis management services, supports, programs, and interventions that they need.
_ _ _ _ _
In the end, districts and schools have more permission and flexibility (especially under the Elementary and Secondary Education Act) to distribute their funds to differentially address the students and schools with the greatest need (i.e., Core-Plus Funding).
But funding aside, there are ways to implement effective, evidence-based approaches that address disproportionate curriculum, instruction, achievement, and disciplinary practices.
For begin this process, feel free to review two of our free monographs:
Implementing Project ACHIEVE at the School and District Levels: School Improvement and Positive Behavioral Support Systems/Social-Emotional Learning Overview
and
A Multi-Tiered Service & Support Implementation Blueprint for Schools & Districts: Revisiting the Science to Improve the Practice
[CLICK HERE and Look at the Bottom of the Page]
While evidence-based practices may long out-impact increases in funding, we still need to “level the funding playing field,” and there are still too many districts—especially in the rural areas of our country—who simply need more funding.
_ _ _ _ _
As always, I appreciate those of you reading these thoughts. If you have comments or questions, please feel free to contact me.
And please also feel free to take advantage of my standing offer for a free, one-hour conference call consultation with you and your team at any time.
Meanwhile, have a Happy Thanksgiving—filled with good food, fellowship, and the warmth of your families celebrating together.
Best,
