Give an autistic employee the quiet, private workspace they asked for, and there is a documented chance you will make things worse. Not because the quiet was wrong. Because being moved to a room away from everyone else read as exclusion rather than support.

That finding sits inside a systematic review of physical workplace adjustments for neurodivergent workers by Clara Weber and colleagues, published in Applied Psychology: An International Review. It is worth sitting with, because it breaks the logic that most neurodivergent workplace adjustments are built on: identify the sensory problem, apply the matching fix, close the case.

This post covers what that review found, why its single most generalisable result is the one nobody puts on an accommodations list, and how to run the conversation differently if you are a neurodivergent researcher asking for something, or the person being asked.

Twenty studies is the entire evidence base

The review searched five databases across two decades and screened 319 records. Twenty studies survived. Those twenty covered 916 neurodivergent participants in total, with individual samples running from 5 people to 856.

The skew is worse than the headline number. Eighteen of the twenty studies looked at autism. One looked at ADHD alone. One used a mixed sample that included dyslexia. Dyspraxia and DCD are close to absent. Study quality varied sharply: the qualitative studies scored well on the CASP appraisal tool, averaging 88 percent, while the mixed-methods studies averaged 47 percent and the single quantitative study scored 20 percent.

The most important methodological point is easy to miss. No study tested whether an adjustment actually worked. There were no intervention designs and no experimental comparisons. Every finding describes what neurodivergent workers reported wanting, or reported having received, not what measurably improved their outcomes.

Set that against how confidently sensory workplace design is sold. Flooring manufacturers, architecture practices and workplace consultancies publish neurodiversity design guidance with a certainty the research does not support. If you have ever been handed an adjustments list that felt suspiciously tidy, this gap is why.

The most consistent finding is not an adjustment at all

More than half the included studies found that having control over one’s own physical environment mattered independently of the specific adjustment involved. Lighting, sound, temperature: the ability to change them was doing work that the settings themselves were not. One study found autistic workers rated environmental control as significantly more important than a neurotypical comparison group did.

More than half the included studies found that having control over one’s own physical environment mattered independently of the specific adjustment involved.

This is the result that survives the review’s weaknesses, and it is the one that cannot be turned into a purchase order. A department can buy noise-cancelling headphones. It cannot buy the experience of being the person who decides when to wear them.

It also reframes what an adjustment request is. If control is the active ingredient, then an adjustment decided for someone has already lost most of its value, however well chosen. That is a plausible reading of the study finding that opened this article: a private office granted from above is a different object from a private office chosen.

What the review found, sense by sense

With the caveats above firmly attached, here is the shape of the evidence:

• Auditory (18 studies). The most studied domain by a wide margin. Noise-cancelling headphones and earplugs appeared in nine studies; quiet or private locations away from open-plan space appeared in ten. These were the most consistently reported facilitators of both staying in work and performing in it.
• Visual (15 studies, autism only). Natural or incandescent light preferred over fluorescent. Desk lamps preferred over overhead lighting. The ability to dim or switch lighting off. Blinds, sunglasses, and screens used to cut visual distraction.
• Tactile (1 study). Almost nothing. Temperature control, plus fidget tools and a decompression room, and notably those were suggested by supervisors rather than by neurodivergent workers themselves.
• Olfactory (1 study). Effectively no evidence. One mention of personal fragrance as a possible facilitator.
• Social. Private offices, quiet corridors, and the ability to withdraw dominated. This is also where the counter-finding sits: some participants given isolated workspaces reported feeling stigmatised rather than helped.
• Non-sensory. Adequate furniture, treadmill and expandable desks, and the single most frequently cited item across nine studies: the ability to work from home.

Notice the collapse from left to right. Sound and light look like a body of evidence. Touch and smell are one study each. If your own sensory profile sits mainly in the under-researched columns, published guidance has close to nothing for you, and you should treat its confident silence accordingly.

Adjustment fatigue

The review introduces a concept that deserves wider use: adjustment fatigue. The idea is that an accommodation can exist, function as designed, and still cost the person using it ongoing effort that the default arrangement costs everybody else nothing.

The wheelchair ramp is the analogy the authors reach for. The ramp is there. It works. It is also longer, round the side, and takes more time and energy every single day than the four steps everyone else uses without thinking.

An accommodation that exists is not the same as an accommodation that is free to use.

This explains a pattern that confuses a lot of supervisors: the adjustment was granted, the paperwork is closed, and the person still seems to be struggling. Nothing has failed. The adjustment is simply not costless, and the running cost was never counted. In coaching conversations, this is one of the most common sources of mutual frustration I see between researchers and their supervisors.

The direction nobody studied

Every included study addressed hypersensitivity. None addressed hyposensitivity.

That is a substantial hole. Understimulation is a documented problem, and one study in the review found it hampered concentration. Yet the entire adjustment vocabulary points one way: reduce, quieten, dim, screen off, withdraw. If you focus better with noise, movement, or more sensory input rather than less, the literature has not looked at you, and the standard adjustments menu may actively work against you.

Gustatory and vestibular modalities were not addressed at all.

How to run the conversation instead

If you are the one who needs the adjustment

Ask for the decision rights, not only the object. “I need to be able to change the lighting at my desk” holds up better over time than “I need a desk lamp,” because it survives the day the lamp turns out to be wrong. Be specific about direction, including upward: if you need more input rather than less, say so explicitly, because nobody will guess it.

Then check the running cost before you accept. If using the adjustment requires you to explain yourself weekly, sit somewhere visible, or ask permission each time, it will erode, and it will look from the outside like you stopped needing it.

If you supervise or work alongside someone

Resist the reflex to solve it well. The review’s clearest practical warning is against deciding on someone’s behalf, however good the decision. Offer options and hand over the choice.

Watch for the two failure modes: an adjustment that increases isolation while technically meeting the request, and an adjustment that quietly runs on the other person’s effort. Both look like success on paper. In my work with doctoral supervisors, the second one is almost always the one that goes unnoticed.

The review is also blunt that changing the physical environment alone, without psychosocial or policy support, risks what it calls smoke-and-mirrors inclusion. A quiet room does not compensate for a culture in which asking for one is a career risk.

Where this reading has limits

This is a thin evidence base, and I have tried not to inflate it. Twenty observational studies with no efficacy testing cannot tell you what will work for you. They can tell you what neurodivergent workers consistently report needing, and they can tell you which confident recommendations are running ahead of their evidence.

The autism skew matters too. If you have ADHD, dyslexia, or dyspraxia, most of what is summarised here was studied on someone else. Read it as a set of hypotheses to test on your own week, not as findings that transfer.

And “control matters most” is not a universal solvent. Control over your environment does nothing about workload, supervision quality, contract precarity, or a department that treats adjustment requests as a performance signal. Those are different problems and need different conversations.

What this changes

The useful move is not to memorise which adjustments have the most studies behind them. It is to notice that the finding which generalised best was about authorship, not equipment. Who decides, who can change their mind, and who absorbs the ongoing cost.

That is a smaller question than “what accommodations does a neurodivergent researcher need,” and a much more answerable one. It is also the version you can act on this week, without a budget or an institutional policy change.

Frequently Asked Questions

What are the most evidence-supported workplace adjustments for neurodivergent employees?

Auditory and visual adjustments have the widest support, though “widest” is relative in a field with twenty studies. Noise-cancelling headphones or earplugs appeared in nine studies, and quiet locations away from open-plan space in ten. On the visual side: natural light over fluorescent, desk lamps over overhead lighting, and the ability to dim or switch lighting off. Working from home was the single most frequently cited non-sensory adjustment.

Why did giving someone a private office sometimes backfire?

One study in the review found that participants placed in isolated workspaces reported feeling stigmatised and isolated rather than supported. The quiet itself was not the problem. Being relocated away from colleagues carried a social meaning that the adjustment was not designed to account for, particularly when the decision was made for the person rather than with them.

What is adjustment fatigue?

It describes the ongoing effort of using an accommodation that technically works. The comparison the authors use is a wheelchair ramp: functional, compliant, and also longer and more effortful every day than the steps everyone else takes without noticing. It matters because employers tend to treat a granted adjustment as a closed case, and stop counting the cost from that point on.

Does this apply to ADHD and dyslexia, or only autism?

Mostly autism. Eighteen of the twenty studies focused on autistic workers, with one on ADHD and one mixed sample including dyslexia. Dyspraxia and DCD are essentially unrepresented. If you are ADHD or dyslexic, treat the findings as untested hypotheses rather than transferable evidence, and give more weight to your own observation of your working week.

What if I need more sensory input rather than less?

Then the published literature has not studied you. Every study in the review addressed hypersensitivity; none addressed hyposensitivity, despite evidence that understimulation harms concentration. Standard adjustment menus assume the “reduce stimulation” direction throughout. You will likely need to name your requirement explicitly and specify the direction, because no template will do it for you.

Should I ask for a specific adjustment or something broader?

Ask for the decision rights where you can. Requesting the ability to control lighting at your desk is more durable than requesting one particular lamp, because it survives the discovery that the lamp was the wrong choice. Specific requests are still useful as concrete examples; the difference is whether the arrangement can be revised without a new negotiation each time.

When is a physical adjustment not the answer?

When the actual constraint is workload, supervision, job insecurity, or a department where asking is treated as a warning sign. The review is direct that changing the physical environment without psychosocial and policy support produces what it calls smoke-and-mirrors inclusion. A quiet room does not offset a culture in which using it is read as not coping.

How can academic career coaching help with this?

Coaching is useful at the point where the problem stops being informational. You know what you need, and the difficulty is deciding what to disclose, to whom, how to frame a request so it is heard, or how to tell whether an adjustment is genuinely working or just quietly costing you. For supervisors, it is often about learning to ask rather than solve. Those are individual questions, which is exactly why a list cannot answer them.