Research about us, not with us

Autistic Mirror (Outside view/surroundings, research)

The most common mechanism is that the advisory board format itself channels participation in a specific direction, regardless of the intentions of all involved parties.

Structural Asymmetry in Knowledge Exchange

An advisory board that "gives feedback" is conceptually a feedback instrument, not a design instrument. This means: research questions, operationalisations, data collection instruments, and theoretical frameworks are developed by the team before the advisory board sees them. Autistic advisory board members then comment on decisions in which they were not involved. This is particularly significant in autism research, because operationalisations often implicitly embed assumptions of the non-autistic majority perspective without being noticed. A classic example: measurement instruments for "social competence" implicitly operationalise neurotypical communication as a reference point. An advisory board involved only after instrument development can still identify this, but rarely correct it, because revision at that point jeopardises the project timeline.

The Double Empathy Problem in Research Communication

The Double Empathy Problem describes a reciprocal barrier to understanding: when autistic and non-autistic people communicate, friction arises in both directions because both sides operate with different processing styles and communication expectations. In advisory board settings, this manifests structurally. Autistic advisory board members often provide precise, detailed feedback that points directly to conceptual gaps. This feedback is sometimes categorised by non-autistic research teams as "too specific", "too critical", or "not constructive", because it does not correspond to the expected style of expertise communication. Conversely, research teams often communicate project goals within an implicit framework that is not fully transparent to outsiders. Both sides leave the meeting with the impression of not having been fully heard. The result is that feedback is recorded, but not translated into the logic of the project.

Phatic Communication as a Barrier to Participation

Advisory board meetings typically run in formats that rely heavily on phatic communication: rounds of introductions, small talk before the actual agenda, implicit rules of conversation about how agreement and criticism are expressed. These rituals serve the social cohesion of the group and signal willingness to cooperate, but they do not convey substantive information. For autistic advisory board members, who process language primarily referentially, these channels consume capacity without yielding benefit. The consequence is that a significant portion of available energy goes into navigating the format, rather than into substantive participation.

When Co-Research Begins

Genuine co-research begins before the first research question, not after it. This includes: autistic co-design of the research question and theoretical framework, control over operationalisations, the ability to reject a data collection instrument rather than just comment on it, and access to decisions, not just their outcomes. An advisory board can take on this role if it is involved in the conceptual work from the start of the project, not just in the review phase. Whether this is the case can be checked with a specific question: Would an advisory board member have had the power to discard a central concept of the project?

Autistic Mirror (Outside view/surroundings, research)

The difference runs deeper than methodological design. It lies in what counts as valid knowledge.

Consultative involvement presupposes that the knowing subject is already fixed: The researcher asks, the autistic person answers, the researcher interprets. The authority of interpretation remains non-autistic. Neurologically, this corresponds to a model in which one side brings the prediction structure and the other side provides input data. The prediction error, i.e., the deviation between prediction and input, is calculated and evaluated on the researcher's side. Autistic experience appears in this process as a data point, not as an instance of knowledge.

This is not a criticism of intentions. It is a structural problem in the processing model: Whoever sets the prior determines which deviations are considered significant and which are considered artefacts. If research designs are implicitly aligned with the male-phenotypic or neurotypical default, data that deviates from this is systematically treated as noise, not as a signal. The bias is not in the data point, it is in the prediction model.

Shared decision-making power shifts precisely this structure. Autistic co-researchers bring their own priors, their own prediction models about what is relevant in which situation, what constitutes a meaningful operationalisation, what counts as a valid answer. This changes which questions are asked before the first sample is drawn. It changes construct validity at the conceptual level, not just at the level of analysis.

Epistemologically, this is the difference between triangulation and epistemic justice. Triangulation adds perspectives without shifting the authority of interpretation. Epistemic justice, a term from philosophy of science, asks: Is this person's competence to know recognised, or is it systematically discredited before they speak? In autism research, this is not an abstract question. Masking research, operationalised without autistic co-researchers, for years coded camouflaging as a social skill, not as an exhaustion strategy. This is not a measurement error. This is an error at the level of the epistemic framework.

Neurologically, there is a second difference that is rarely made explicit. Monotropism, i.e., the tendency to focus attention deeply on a few channels rather than distributing it broadly, influences how autistic co-researchers work through research designs: they tend to identify inconsistencies in operationalisations that broadly distributed attention overlooks. This is not a poetic statement of strength, it is a consequence of the processing architecture. Studies that take this contribution methodologically seriously often report that autistic co-researchers identify conceptual contradictions between the research question and the measurement instrument early on, which are only noticed late in peer review.

The question of whether a research project operates consultatively or with shared power is therefore not decided in the ethics application. It is decided by who has decision-making weight in disagreements about operationalisations.