There's no shortage of advice about neurodivergent students. Some of it is helpful. Some sounds convincing but turns out to come from one tiny study, an inspirational speech, or an author citing another author who never quite said that. So let's look at what the research actually tells us. A citation is useful, but it isn't a stamp of certainty.
First, a couple of thinkers who help us ask better questions. Then we'll look at recent reviews of college support and common ideas such as clear instructions, mentoring, Universal Design for Learning, and the double empathy problem. For each, the important question is the same: what was actually studied, and what can someone responsibly do with that information?
A quick rule for reading research: Ask who was studied, where, how the researchers measured success, and whether the results apply to your situation. A study of 12 students at one university can be informative without telling us what will work at every Community College. Research reviewed November 2026.
Two people worth spending time with
Not every useful idea comes from a study comparing two groups of students. Some of the most important questions are about who gets to define a problem and what happens when someone learns with the right kind of help. These two thinkers are where I keep coming back.
Orlando Fals-Borda
Fals-Borda helped develop participatory action research. The short version: people who live with a problem should help decide what the problem is, investigate it, and do something with the results. They aren't just the data.
For a college, that means more than sending students a survey after an initiative is designed. Involve them in choosing what needs fixing, give them a say in the decisions, and tell them what changed because of their work. Research shouldn't be something we do to people. [1]
Lev Vygotsky
Vygotsky's zone of proximal development is the space between what someone can do alone and what they can do with guidance or collaboration. Good teaching works in that space. The help is part of the learning, not evidence the learner has failed.
In practice, that can mean a model, a worked example, a peer who has done the task before, or an instructor asking the right question. The aim is to build toward greater independence without withdrawing support just to prove a point. [2]
What happens when you put them together? People should have a say in the support they receive, and that support should help them do things they couldn't yet do on their own. Fals-Borda is asking who holds the power. Vygotsky is asking what learning becomes possible with help. Neither one ran a study proving that a particular program raises graduation rates for neurodivergent Community College students. They're useful for designing the work, and the outcomes still need to be checked.
Voices beyond academic journals
The research journals matter. So do people writing from inside neurodivergent communities. Nick Walker's essay Neurodiversity: Some Basic Terms & Definitions is a good place to start when the vocabulary gets tangled. The book Neuroqueer Heresies goes further into identity, power, and what she calls the neurodiversity paradigm. Both offer ideas you can use to question assumptions. Neither should be presented as an intervention study. [3] [4]
That's not a downgrade. A community essay can give you a question that nobody thought to ask in a research study. The next step is to be clear about what kind of knowledge you're using. For lived experience and language, listen to people who use the language. For claims that an intervention improves outcomes, look for research that actually measured those outcomes.
What the recent reviews found
Study one: 37 studies on student wellbeing
Ross, Dommett, and Byrom looked for studies of programs intended to improve neurodivergent students' mental health, wellbeing, or experience of college. Their 2026 systematic review brought together 37 studies from seven countries. The approaches included coaching, counseling-based programs, mentoring, peer support, mindfulness, and help with organization. Their searches were updated in May 2025. [5]
Here's the catch. Seventeen studies focused on ADHD and fourteen on autism. Other forms of neurodivergence received much less attention. Almost half the studies involved fewer than twenty participants. And only two of the thirty-seven reported involving neurodivergent students in designing the support. Two were described as strengths-focused rather than primarily deficit-focused. That leaves some rather large holes in our understanding.
What we can take from it: There are several support approaches worth exploring, but the evidence is uneven. Students should help decide what gets offered, and more kinds of neurodivergent experience need to be studied.
What we cannot claim: That coaching, mentoring, mindfulness, or any other single approach is guaranteed to improve graduation or wellbeing for every neurodivergent student. Liking a program after six weeks is not the same thing as having a better outcome three years later.
What to do with that: If you start a support program, decide what problem it is supposed to solve, involve students in its design, and check whether it helped. A nice attendance count is not an outcome measure.
Study two: 59 articles on student engagement
O'Neill and colleagues looked at 59 articles about how colleges support neurodivergent students' engagement. That included 25 quantitative studies, 25 qualitative studies, and nine that combined methods. The studies covered mentoring, ADHD coaching, strategy instruction, and other approaches. Many focused on autism, and the researchers were not all studying the same students or measuring the same things. [6]
A scoping review is a map of the available research. It's useful for seeing what people have tried and where the gaps are. It isn't a tournament that tells us which program won. The authors found that better, more inclusive research is still needed before recommending one approach across very different colleges.
What to do with that: Before borrowing another campus's program, ask who it was designed for, which problem it addressed, and whether that is actually your problem.
What both reviews are telling us
Structured support is worth considering. So is having students help design it. But we still have much more evidence about a relatively narrow range of students than about neurodivergent college students as a whole. And we need better follow-up: did a program help beyond the semester in which it was offered?
What supports the advice you keep hearing?
Here's a quick comparison of frequently recommended practices. Some are based on direct studies. Others come from educational frameworks, legal requirements, or strong practical reasoning. Those differences matter.
| Recommendation | Evidence or rationale | Important limitation | How to use it responsibly |
|---|---|---|---|
| Explain hidden college expectations | Karp & Bork's Community College interviews [7] | Not a trial focused on neurodivergence | Make procedures explicit; test whether students can act on them |
| Design accessible options in advance | CAST UDL 3.0 [8] | Framework, not one intervention effectiveness estimate | Match options to goals; ask who can use them |
| Offer structured mentoring or support | Ross and O'Neill reviews [9] [10] | Highly varied studies and outcomes | Pilot with defined goals, boundaries, student input, and follow-up |
| Include students in intervention design | Reviews reveal a major gap; participatory scholarship offers a method [11] [12] | Co-design alone does not prove effectiveness | Give affected students decision influence and evaluate results |
| Improve communication repair | Milton's double empathy theory [13] | Conceptual framework, not a universal treatment study | Check mutual understanding; do not infer intent from style |
| Avoid fixed learning-style matching | Pashler et al. critical review [14] | Doesn't deny individual needs or accessible formats | Choose formats for the content and learner's access |
| Protect academic adjustment processes | U.S. Department of Education guidance [15] | Legal scope and implementation vary | Follow institution-specific procedures and current law |
| Respect choice over disability disclosure | U.S. privacy and civil-rights guidance [16] [17] | Rights differ across contexts | Ask what is necessary and use the designated channel |
| Audit digital accessibility | W3C WCAG 2.2 [18] | Technical compliance alone is not usability | Test keyboard, contrast, semantics, and real user experience |
None of these is a guaranteed recipe for getting students to graduate. They're starting points for making better decisions and checking what happens.
Hidden rules are still rules
Karp and Bork interviewed students and staff at three Community Colleges about what it takes to be a 'successful college student.' They found that colleges expect students to understand rules and roles that haven't necessarily been explained. This wasn't a study about neurodivergence in particular. Unwritten college rules are a broader problem, and neurodivergent students may face additional difficulty interpreting them. [19]
What might help? Plain-language directions, examples of assignments, and a clear explanation of when and how to ask for help. Those are sensible responses to the problem. But the original interviews didn't test whether a particular assignment template raises grades. If you're trying something new, ask students whether the directions are clearer and look at how their work changes.
Even a great explanation of a problem can lead to a bad solution. You still have to check.
Universal Design for Learning: helpful, not magical
CAST's Universal Design for Learning (UDL) Guidelines 3.0 encourage educators to plan for differences among learners. They emphasize access, meaningful choices, and different ways of engaging with material or demonstrating learning. [20]
UDL can help you ask useful questions. It does not mean every assignment must be available in seventeen different formats. Learning objectives still matter, and some formats are themselves essential. A thoughtful instructor identifies the goal, removes incidental barriers where possible, and designs legitimate ways to demonstrate the target competence.
A course can follow UDL guidance and still fail students in practice, especially if feedback is inconsistent or the technology isn't accessible. Conversely, a particular practice may be useful without being formally named in the framework. Ask students what works. Watch where they get stuck. Change the course if the changes make sense.
The double empathy problem: misunderstanding goes both ways
Damian Milton suggested that when autistic and non-autistic people misunderstand one another, it may not be because one person is socially deficient. Different assumptions and communication styles can create a mismatch in both directions. Instead of asking 'What's wrong with this person's communication?' try 'What did each person think the other meant?' [21]
Crompton and colleagues (2020) reported effective autistic-to-autistic information transfer in a particular experimental task. That work offers a challenge to claims that autistic communication is globally ineffective. It does not establish that all autistic people understand one another, that mixed-neurotype groups are inherently disadvantaged, or that structured communication always succeeds. [22]
In a classroom: Have groups agree where decisions will be written down. When confusion happens, ask each person to explain what they thought had been decided. Then check whether the new process makes misunderstandings easier to resolve.
Belonging is not something you can declare by email
Belonging is important, but it is not an item a campus can order in bulk. Being admitted, receiving accommodations, feeling respected, participating in learning, and identifying with a community are related but different experiences. A student can have access without belonging; a student may feel socially accepted but still lack effective academic accommodations.
Benjamin Kaupp's essay on neurodivergence makes the case that being allowed into college isn't the same as being treated like you belong there. If students spend all their energy decoding social expectations, they have less left for learning. That's an argument about how we should design institutions, not a claim that a particular program has been proven to fix retention. [23]
Institutions can document conditions more easily than they can compel feelings: are expectations clear, can students ask questions without stigma, is there a route to repair conflict, are support processes trustworthy, and do students have meaningful influence over decisions? These are promising evaluation targets, not a substitute for listening to students about their experience.
Where these ideas come from
Some scholars are especially useful for asking better questions about power, disability, and the idea of 'normal.' Their work helps us understand problems, even when it doesn't test a particular teaching method.
Tobin Siebers: His work in disability theory challenges overly narrow accounts of embodiment and the separation of disability from social life. It informs how we think about disabled identity, rights, and knowledge. [24]
Rosemarie Garland-Thomson: Her concept of misfitting directs attention to the relationship between bodies and environments. The same person may encounter a barrier in one arrangement and greater fit in another. This is a theoretical frame for auditing design choices. [25]
Lennard J. Davis: His historical work examines how “normal” became a powerful category and standard. It invites institutions to ask whether a requirement tests competence or simply conformity to a preferred norm. [26]
Michel Foucault: His writing about discipline and normality helps explain why schools care so much about whether people behave the 'right' way. He wasn't testing college accommodations. His value here is in helping us question assumptions about normal behavior. [27]
Orlando Fals-Borda and Lev Vygotsky: The two examples at the top of this page explain why student ownership and supported learning matter. Their work helps shape our questions; we still need to evaluate the practical results. [28] [29]
Contemporary neurodiversity scholarship and writing: Nick Walker offers clear language and community-based perspectives; Dwyer offers a nuanced map of neurodiversity approaches; Botha and colleagues correct a simplified origin story; Zaneva and coauthors present an annotated reading pathway developed collaboratively by a neurodiverse research group. Taken together, they make it harder to treat a varied community as one tidy doctrine. [30] [31] [32] [33]
Be careful with labels and generalizations
Language preferences vary among individuals and communities. A research sample labeled “students with ASD” should not automatically become “all neurodivergent students” in a summary. An autism-specific study may illuminate a related barrier, but it is not direct evidence for every neurotype. A self-selected study of people engaged in advocacy may capture important perspectives while not representing the preferences of an entire population.
It's fine to take ideas from an autism-specific study and consider whether they might be helpful elsewhere. Just don't pretend the researchers tested every neurodivergent student. People may describe themselves as autistic, disabled, neurodivergent, several of those, or none. Listen to them rather than declaring which label is correct.
The history of neurodiversity language is collective. Botha and colleagues' 2024 correction and Zaneva and colleagues' 2024 reading guide are useful starting points for understanding how communities, scholars, and activists have contributed to the field. [34] [35]
Eight questions to ask before trusting a headline
Before citing a claim, ask eight questions:
- Who was studied? Autistic students, students with ADHD, students registered with disability services, or people who self-identify as neurodivergent?
- Where? A selective university, Community College, online program, secondary school, or workplace?
- How many people? A small pilot can show feasibility, but rarely establishes a robust population effect.
- What was compared? No comparison, waitlist, existing practice, or randomized control?
- What outcome? Satisfaction, immediate symptom scores, grades, retention, completion, or long-term wellbeing?
- Who defined success? Researchers and professionals, students, families, or the institution?
- What did the authors actually conclude? Was the result statistically uncertain, narrowly scoped, or mixed?
- What would change our minds? A replication, longer follow-up, broader sample, or direct student feedback?
Example of restrained wording: “A small pilot found the mentoring approach feasible and acceptable to participants.” Not: “Research proves this mentoring method increases graduation for neurodivergent students.”
What we still need to know
We need more Community College research, more work involving neurodivergent students beyond autism and ADHD, and studies that don't only count students registered with disability services. We also need students to have more influence over what gets studied. We also need to distinguish immediate satisfaction from sustained participation, wellbeing, transfer, employment, and completion. [36] [37]
Other important questions concern the interaction of race, language, gender, financial insecurity, and disability; how students experience the burden of repeatedly explaining access needs; whether institutional policy changes reduce procedural friction; and what happens when students control more of the support design.
And please document what didn't work. If an initiative increases paperwork, demands disclosure, or makes students feel observed rather than supported, that evidence belongs in the record. A report that explains why something failed can be more useful than another glowing brochure.
Sources and further reading
You can follow the source links below to read the original work. The notes explain why each source matters and where its limits are. Some articles are paywalled; their links lead to the publisher or an available public record rather than copied text.
Sources and further reading
Want to go further? These are the sources behind the information above, with a short note about what each one covers.
Explains why difference and disability can both be real, and why being admitted is not the same as belonging. This is an essay, not a test of a particular support program.
Read the source ↗Explains what people mean by different approaches to neurodiversity. It does not test a classroom technique.
Read the source ↗Challenges single-founder accounts using community history and archival evidence.
Read the source ↗A useful starting point for exploring the field, not a study of student outcomes.
Read the source ↗Reviews 37 studies, mostly involving autism or ADHD. Only two reported involving students in program design. The findings are mixed, not a verdict that every support program fails.
Read the source ↗Surveys 59 articles on coaching, mentoring, and other approaches. They studied different things, so there is no single winner.
Read the source ↗Helps plan accessible learning experiences. Following a checklist is not proof that students are doing better.
Read the source ↗Shows how community colleges leave important expectations unstated. The interviews were not specifically about neurodivergent students.
Read the source ↗Suggests that misunderstanding can go both ways. This is a way to understand communication, not proof about every conversation.
Read the source ↗Finds no adequate evidence for sorting people into fixed learning styles and teaching to those labels.
Read the source ↗General guidance on academic adjustments, essential requirements and differences from high school; institutions and individual situations vary.
Read the source ↗Postsecondary attendance or turning 18 generally transfers FERPA rights to students. Specific exceptions allow some disclosures.
Read the source ↗Restrictions on pre-offer disability questions with limited exceptions; not individualized legal advice.
Read the source ↗Explains the reasonable accommodation process. Particular obligations depend on circumstances and governing law.
Read the source ↗Specific web accessibility criteria, not a guarantee of usability for all readers.
Read the source ↗Theory of disability, bodies, identity, and politics.
Read the source ↗A framework for understanding disability through person-environment fit.
Read the source ↗Examines the historical construction of statistical and cultural normalcy.
Read the source ↗An approach to producing knowledge with people affected by a problem, not only about them.
Read the source ↗Explains how research, action, and participation can be combined. This is a research approach with an explicit social-change commitment, not a trial of college supports.
Read the source ↗Introduces the zone of proximal development and the role of guided, social learning. Originally about development and schooling, not a measured effect on Community College completion.
Read the source ↗Explains influential vocabulary used in neurodiversity communities. Useful for identity and terminology, but it does not test whether an education program works.
Read the source ↗Goes deeper into the neurodiversity paradigm, autistic empowerment, and neuroqueer theory; a source of ideas rather than evidence of intervention effectiveness.
Read the source ↗A lens on disciplinary institutions and normalization; not neurodivergence-specific.
Read the source ↗Supports rethinking one-sided deficit narratives, but findings are tied to the particular experimental task.
Read the source ↗Describes a co-design process; it is not proof of broad long-term outcomes.
Read the source ↗Looking at the same problem from a different perspective?
Understanding neurodivergenceResearch and evidence