Psychological risk assessment: why “everything is okay” does not mean it is okay

Autistic Mirror (Outside view/Surroundings, role employer)

Standard procedures for assessing psychological workload systematically capture autistic load profiles incompletely. That is not because of the employee, but because of the methodology.

Why standard procedures fail

Most instruments (surveys, workshops, observation methods) measure loads that are considered relevant in the neurotypical working world: time pressure, conflicts, lack of latitude. Constant sensory load, communication effort and masking pressure do not appear in these categories. On top of that: masking means that the employee actively hides from the outside how high her load is. That is not a statement about her actual load, but a survival strategy that itself costs capacity. The neurological exhaustion you notice is typically what is left after a long masking effort, not its beginning.

What makes up autistic load

Three sources are frequently involved that standard procedures overlook:

First, constant sensory load. Light, noise, smells, changes of room, the number of people present at the same time. These are not questions of comfort but neurological sources of load. Sensory gating, the brain's ability to automatically dampen irrelevant stimuli, works with less automaticity in many autistic people. What others filter out unconsciously is processed consciously and therefore permanently costs cognitive capacity.

Second, predictability and change. Unplanned meetings, short-notice task changes, unclear expectations and social situations without recognisable structure create predictive coding load: the brain constantly updates its internal model of the situation because the predictions do not fit reliably. That costs energy regardless of whether the task itself is difficult.

Third, communication effort. Every meeting with implicit social expectations, small talk, nonverbal codes that have to be decoded, creates processing effort that does not exist for neurotypical colleagues. That is the double empathy problem: the communication styles do not fit together structurally, not because the employee communicates worse, but because she performs an additional translation effort that no other person in the room performs.

What an assessment that captures this can look like

The assessment itself needs a different form, not just different questions.

In writing, not orally, and anonymous if wished. Open individual questioning instead of a group format. The employee should be able to decide whether she communicates directly with you or via a questionnaire. That reduces the effort that a conversation about exhaustion would itself require.

Ask about concrete situations, not categories. Not "Do you feel burdened by noise?", but "Which concrete situations at work cost you particularly much energy?" and "Are there times of day, rooms or tasks after which you feel clearly more exhausted than after others?" The employee often knows her own sources of load very precisely. She is just rarely asked about them directly.

Go through the workplace objectively. Regardless of what the employee says, the sensory environment can be observed and named: light sources and whether they can be regulated, noise level and origin, room changes per day, number of unplanned interruptions, length and structure of meetings. These are measurable factors that can feed into the assessment without the employee having to out herself.

Involvement without pressure

Disclosure, that is whether the employee officially names her autism, is her decision. The assessment does not have to presuppose it. The entry point can be the working conditions, not the person: "We are looking at which conditions create which load for whom." That creates a frame in which specific needs can be voiced without a diagnosis having to be named.

If the company has a disability representative body (SBV), it should be involved if the employee wishes. It has participation rights in the risk assessment under the German Social Code IX.

The results of the assessment lead to measures that are discussed with the employee, not decided for her. She defines which adjustments actually relieve her. You create the structures, she fills in the content.

Autistic Mirror (Outside view/Surroundings, role employer)

Because "okay" and "burdened" are both true for him at the same time, and that is not a contradiction.

Masking runs in the open-plan office as a continuous background task. Automatic autistic reactions to noise, movement, parallel conversations and unpredictable interruptions are suppressed prefrontally so that behaviour stays inconspicuous on the outside. This runs alongside the actual work, without him necessarily experiencing it as "stress" or being able to name it. From his perspective everything really is okay in the office: he functions, gets tasks done, does not stand out. That is his honest feedback on what he consciously perceives in the office.

The exhaustion afterwards is sensory filter exhaustion combined with the overall load of masking. Background noise that other brains filter out preattentively, that is without effort and without conscious control, is not automatically dampened in many autistic people. Instead the filtering continues in a controlled way and continuously uses capacity in the prefrontal cortex. After a full working day this budget is used up. The feeling of emptiness afterwards is not an assessment of the work situation, but a physiological result of the day's load.

The problem for your risk assessment: he describes two different things, and you only ask about one. The question about load in the office triggers his assessment of the situation in the moment. The exhaustion in the evening he may not see as workload, but as a normal consequence of a working day, or he does not connect it with the office because it only sets in later. This pattern is called restraint collapse. Over the day masking holds together what collapses abruptly in the first safe environment. He sees the collapse in the evening, not the cause in the morning.

For the risk assessment this means: a binary okay/not-okay query structurally does not capture this form of load. More useful are questions about the state after the working day, about the time until recovery, about the comparison with working days in another environment. The employee knows the answers, but he needs the right questions to be able to give them.