This is the first post I’ve made since completing my Master’s in Education coursework; it’s great to write an entry without the pressure of a deadline, specific content, or grammar checking. It reflects my ongoing thinking about equity and inclusion in computer science education, and how both the research and my Christian faith push me to ask questions that aren’t always comfortable.
I believe every student deserves to belong. Every student must be to be seen, supported, and given a fair shot. That’s not up for debate in my classroom. But wanting something good and actually getting there are two very different things. What if some of the ways we’re pursuing diversity, equity, and inclusion (DEI) are quietly working against us?
As a CS teacher who follows Jesus, I think I owe my students honest reflection, not just comfortable agreement.
Why DEI Sounds So Right (and Usually Is, on the Surface)
The intent behind DEI is hard to argue with. Historically, computer science has had a serious access problem. Women, Black, Latino, and Indigenous students have faced cultural, structural, and social barriers that kept them out of classrooms and careers. DEI efforts were designed to fix that. Representation matters. Belonging matters. Creating a welcoming environment matters.
And on paper, it works beautifully. Hire diversely. Celebrate different cultural contributions to technology. Design curricula that speak to a wider range of experiences. Reduce bias in admissions. These are legitimately good ideas.
So what’s the problem, bro?
When the Remedy Becomes the Disease
Research is increasingly showing that many common DEI interventions, particularly mandatory diversity training, don’t just fail to help; they can actively make things worse.
A study by Legault, Gutsell, and Inzlicht (2011), published in Psychological Science, found that when anti-prejudice messages feel like external obligations rather than internal values, they produce what the researchers call “ironic effects”, participants actually showed more explicit and implicit prejudice than those who received no intervention at all. The study used anti-prejudice brochures and word primes rather than formal training sessions, but the implication is significant: when external pressure replaces internal motivation, the intervention can backfire.
A peer-reviewed meta-analysis by Paluck, Porat, Clark, and Green (2021), synthesizing over 400 randomized experiments on prejudice-reduction programs, found the overall effect size to be near zero, with what the authors called “troubling indications of publication bias.” The Aristotle Foundation (2024), a Canadian policy organization, summarized these findings in a public brief. Frank Dobbin, a sociologist at Harvard, has concluded from his own research that “hundreds of studies dating back to the 1930s suggest that antibias training does not reduce bias, alter behaviour or change the workplace.”
More recently, a 2025 scoping review by Mihaylova and Rietmann in the Journal of Sustainable Business mapped the landscape of workplace DEI backlash across 28 studies. They identified three recurring drivers: a gap between DEI research and practice, bias and inequality in how programs are actually implemented, and insufficient organizational and managerial support. The pattern that emerges is well-intentioned programs, when imposed without buy-in or grounded in evidence, tend to generate resistance rather than belonging (Mihaylova & Rietmann, 2025).
The Silo Problem
But the deeper issue, the one I keep coming back to as a teacher, is how DEI can fragment a classroom into identity camps rather than building community.
When every initiative is organized around group identity, separate mentorship programs for women in STEM, separate affinity groups by race, tailored curriculum tracks by demographic, we’re inadvertently sending a message: your group defines your experience here. We mean to create belonging, but we end up reinforcing the very separations we hoped to dissolve.
There’s also the phenomenon of stereotype threat. Steele and Aronson’s foundational research (1995), replicated many times since, showed that simply making students aware that their demographic group is stereotyped to underperform in a subject can cause them to underperform, even when they’re fully capable. Their experiments studied brief race-salience prompts rather than sustained DEI programming, but the finding invites a question I keep returning to as a teacher: could a DEI effort that constantly emphasizes which groups are underrepresented in CS unintentionally prime students with exactly the message we don’t want them to carry into an exam?
This is the paradox. The more we spotlight the gap, the more we may be reinforcing it. Morgan Freeman was once asked how to get rid of racism. He said, “Stop talking about it.” The point isn’t to ignore injustice, but to recognize that the way we frame it can have unintended consequences (https://www.facebook.com/reel/1209992131018789).
Mogilski et al. (2025), writing in Theory and Society as part of an adversarial collaboration between DEI critics and supporters, raise a concern both sides share: much of DEI’s institutional programming is “poorly documented,” and some amounts to a “crowdsourced amalgam of unstandardized procedures” rather than evidence-based practice. When programs aren’t grounded in what actually works, we risk reducing students to demographic representatives rather than seeing them as individuals.
What This Looks Like in CS Education Specifically
In computer science, the stakes are high. A 2023 analysis by Taylor, Drucker, Alvin, and Sultan examined completion trends for underrepresented students in CS and found something striking: despite years of DEI investment, Black students showed an alarmingly sharp decline in CS completions, a decline not observed in other fields of study during the same period. The authors trace this to Simpson’s Paradox: a small number of large institutions drive the aggregate numbers down, masking more varied patterns elsewhere. But whatever the mechanism, the gap wasn’t closing (Taylor et al., 2023).
Now, this doesn’t prove DEI caused the decline. But it does suggest that our interventions aren’t producing the outcomes we expect. A 2025 exploratory study by de Souza Santos, Magalhaes, Wessel, and Barcomb examined how software engineering organizations are responding to DEI backlash and found a complex picture: companies are restructuring, scaling back, or quietly continuing programs, while professionals show varied responses such as anxiety, frustration, and in some cases hope. Notably, the authors conclude that DEI “is evolving, not disappearing,” and emphasize the resilience of inclusion values even under pressure (de Souza Santos et al., 2025). That resilience suggests the problem may be less about the goals of DEI than about how specific programs are designed and executed.
Part of the problem is structural: DEI efforts in CS often focus on representation in recruitment and admission, while under-investing in the classroom experience. A student can be recruited into a CS program with great fanfare and then left to sink in an intro course that hasn’t changed at all. Or worse, they’re enrolled in a “special” section with lower expectations, which communicates something damaging: we don’t think you can do the real thing.
When DEI becomes about optics (diversity posters, checkbox workshops, etc.), it stops being about learning. And CS students notice.
A Word from Faith
This is where my Christian faith gives me both a critique and a vision for something better.
The DEI framework, as commonly practiced, tends to organize the world into categories: the oppressed and the oppressor, the represented and the underrepresented, the privileged and the marginalized. These categories aren’t invented, the inequalities they describe are real. But when group identity becomes the primary lens through which we see each other, we’re not seeing the full picture.
Scripture speaks to this directly. In Galatians 3:28, Paul writes: “There is neither Jew nor Greek, there is neither slave nor free, there is no male and female, for you are all one in Christ Jesus” (ESV, 2001). This isn’t a denial of difference, Paul knew very well the differences between Jews and Greeks and male and female. It’s a declaration that our deepest identity isn’t our demographic category. God looks at the heart, not the external (John 7:24, 1 Samuel 16:7, James 2:1). In the body of Christ, what unifies is more fundamental than what divides.
And Paul’s image of the body in 1 Corinthians 12 is instructive here. Every part matters. Not because of its group identity, but because of its function, its gifts and contributions. “If the whole body were an eye, where would be the sense of hearing?” (1 Corinthians 12:17, ESV, 2001). A truly inclusive community isn’t one where every group gets its own silo. This unity transcends race, gender, and social status (Galations 3:28).
Neil Shenvi, writing on the intersection of DEI and Christian theology (Shenvi Apologetics, 2023), makes this distinction well: the problem isn’t diversity, it’s when DEI becomes ideologically bound to a framework that reduces human identity to group membership and defines people primarily as oppressors or victims. That framework cuts against the imago Dei, which insists that every person has inherent worth and unique personhood, not worth derived from belonging to the right social category.
This matters deeply for CS education. If I organize my classroom around group identities, I’m teaching my students to see each other as representatives of demographics first. If I organize it around shared curiosity, shared challenge, and shared purpose, while actively working to remove structural barriers, I’m teaching something closer to what Paul envisioned.
So What Do We Actually Do?
I’m not arguing for abandoning the goal of equity. I’m arguing for interrogating the methods.
What the research suggests, and what the bible affirms, is that genuine inclusion grows from seeing people fully. It means designing courses that genuinely meet diverse learners where they are without signaling low expectations. It means building classroom cultures of belonging through shared challenge, not separate tracks. It means addressing structural barriers (financial, cultural, pedagogical) directly rather than wrapping them in workshops. And it means being honest when our good intentions produce bad outcomes.
The Apostle Paul didn’t say “be diverse.” He said “love one another.” That sounds simple, but it’s actually harder and more demanding than any DEI checklist. Love sees the whole person. Love pursues their genuine growth. Love doesn’t flatten difference or make it the whole story.
Trust in discernment that comes from God over systems or policies that lean on human understanding (Romans 12:2, Proverbs 3:5-6, ESV, 2001).
References
Aristotle Foundation. (2024). What DEI research concludes about diversity training: It is divisive, counter-productive, and unnecessary [Policy brief by David Millard Haskell]. https://aristotlefoundation.org/reality-check/what-dei-research-concludes-about-diversity-training-it-is-divisive-counter-productive-and-unnecessary/
de Souza Santos, R., Magalhaes, C., Wessel, M., & Barcomb, A. (2025). From diverse origins to a DEI crisis: The pushback against equity, diversity, and inclusion in software engineering. arXiv. https://arxiv.org/pdf/2504.16821
The Holy Bible ESV: English Standard Version. (2001). Crossway Bibles.
Legault, L., Gutsell, J. N., & Inzlicht, M. (2011). Ironic effects of antiprejudice messages: How motivational interventions can reduce (but also increase) prejudice. Psychological Science, 22(12), 1472–1477. https://doi.org/10.1177/0956797611427918
Mihaylova, I., & Rietmann, K. (2025). Diversity, equity and inclusion at a crossroads: A scoping review of the characteristics of its workplace backlash. Journal of Sustainable Business, 10, 18. https://link.springer.com/article/10.1186/s40991-025-00122-5
Mogilski, J. K., Jussim, L., Wilson, C. O., & Love, H. A. (2025). Defining diversity, equity, and inclusion (DEI) by the scientific (de)merits of its programming. Theory and Society, 54, 1173–1186. https://link.springer.com/article/10.1007/s11186-025-09646-y
Paluck, E. L., Porat, R., Clark, C. S., & Green, D. P. (2021). Prejudice reduction: Progress and challenges. Annual Review of Psychology, 72, 533–560. https://doi.org/10.1146/annurev-psych-071620-030619
Shenvi, N. (2023). DEI done right: Disentangling Christian community from critical theory. Shenvi Apologetics. https://shenviapologetics.com/dei-done-right-disentangling-christian-community-from-critical-theory/
Steele, C. M., & Aronson, J. (1995). Stereotype threat and the intellectual test performance of African Americans. Journal of Personality and Social Psychology, 69(5), 797–811. https://doi.org/10.1037/0022-3514.69.5.797
Taylor, J. M., Drucker, R., Alvin, C., & Sultan, S. F. (2023). Simpson’s paradox and lagging progress in completion trends of underrepresented students in computer science. arXiv. https://arxiv.org/pdf/2311.14891