Paper Review: ADHD and Technology Research -- Investigated by Neurodivergent Readers

Paper: Spiel, K., Hornecker, E., Williams, R. M., & Good, J. (2022). ADHD and Technology Research – Investigated by Neurodivergent Readers. In CHI ‘22: Proceedings of the 2022 CHI Conference on Human Factors in Computing Systems, April 29-May 5, 2022, New Orleans, LA, USA. Article 547, 21 pages. ACM. https://doi.org/10.1145/3491102.3517592

Affiliations: TU Wien (Spiel), Bauhaus Universitat Weimar (Hornecker), Purdue University (Williams), University of Amsterdam (Good).

Open access: CC BY-NC-ND 4.0. Available via UvA-DARE repository.

Status: This is perhaps the most important paper available for understanding the political and philosophical foundation of tool sovereignty for neurodivergent users. Not because it has data (Sankesara does that), or task mappings (ICSE 2025 does that), but because it names the problem at the root level: the field of ADHD technology research is built on deficit assumptions, excludes the people it claims to serve, and produces tools that enforce neurotypical compliance rather than support neurodivergent self-determination. The concept of “tool sovereignty” – neurodivergent people building and controlling their own tools – is the practical answer to this theoretical critique.

TL;DR

Four neurodivergent HCI researchers spent two and a half years reading every computing/HCI paper about ADHD technology they could find (52 core papers + 48 extended corpus = 100 papers total). They read these papers not as neutral academics but as people who are described by – and affected by – the research. They found three things:

1. Technologies mostly exist to “mitigate” ADHD behaviors that neurotypical society finds disruptive. Not to support ADHD people. To make ADHD people behave more normally.

2. People with ADHD are almost never involved in designing the technologies built for them. Only 12% of papers even interviewed ADHD participants. Only 5 projects genuinely included ADHD people as co-designers.

3. Participant resistance is visible in the researchers’ own data, but they don’t recognize it. ADHD participants hide devices, subvert interventions, trigger feedback systems for fun rather than “compliance.” The papers frame this as failure. It’s actually the users telling you your design is wrong.

This Is the Theoretical Backbone

Every other paper in the ADHD-and-technology literature tells you what to build or how ADHD manifests in digital behavior. This paper tells you why the entire field is building the wrong things and provides the critical framework for doing it differently. If you only read one theoretical paper on neurodivergent tool design, read this one.

Who Are These Researchers?

This matters because positionality is load-bearing in this paper. The authors are explicit about who they are:

• Four HCI researchers. Two non-binary, two women. Two tenured, two not.
• All neurodivergent. Three formally diagnosed with ADHD.
• Their work spans gender, disability, neurodivergence, children, games, wearables, tangibles, virtual reality.
• Their research approach draws from Critical Disability Studies, feminist methodologies, and crip technoscience.

Katta Spiel specifically is an Assistant Professor for “Critical Access in Embodied Computing” at TU Wien, directing the ERC Starting Grant ACCESSTECH. They are a certified sign language interpreter, won the 2020 SIGCHI Outstanding Dissertation Award, were elected to the Young Academy of the Austrian Academy of Sciences in 2024, and serve on Austria’s Monitoring Committee for the UN Convention on Rights of People with Disabilities. This is not a fringe researcher with an ax to grind. This is someone at the center of HCI accessibility research, with institutional power and recognition, choosing to use that position to challenge the field.

Methodology: Cripping the Literature Review

The methodology is as important as the findings. This isn’t a standard systematic review. It’s a deliberately positioned critical reading.

Positionality

The authors dissolve the subject/object split. They are both researchers and the population being researched. They follow standpoint theory – the idea that marginalized perspectives produce knowledge that dominant perspectives cannot. They explicitly state that reading the corpus “often affected us emotionally” and that “reading the corpus as a whole left us with a sense of disquiet, pain, and even fear.”

This is not a weakness. It’s an analytical tool. When you read papers that describe your neurotype as a burden, your children as suffering, and your traits as invasive – and you feel something about that – that feeling is data.

Corpus Construction

• Core corpus: 52 papers from ACM Digital Library + hcibib, searched February 2019. Computing and HCI literature only.
• Extended corpus: 48 additional papers from mid-2018 to end of 2020, to check whether trends persisted.
• Total: 100 papers analyzed.
• Excluded: Clinical/medical journals, workshops, PhD theses.
• Scope: Intentionally limited to how computing/HCI frames ADHD, not the broader medical literature.

Analysis

All four authors read every paper in the core corpus, each with a specific lens:

1. Participants – who is included, how are they positioned?
2. Disability – how is ADHD conceptualized and explained?
3. Researchers – what framing, what disciplinary origin?
4. Technology – what artifacts, what design processes?

They used Boyatzis’s thematic analysis (inductive + deductive coding), deliberately choosing not to use a rigid quantitative codebook. The intent was to “favour multiple perspectives in our analysis, and strengthen our results through negotiating divergence among us.”

Crip Time

The project took two and a half years (early 2019 to mid-2021). They explicitly invoked “crip time” – Kafer’s concept of “bending the clock to meet disabled bodies and minds” – as a methodological principle. They built in space for grief, for the pandemic, for individual illness. This is itself a statement: research conducted on neurodivergent time rather than institutional time.

The Deficit Framing Critique: What Exactly Is Wrong

This is the core of the paper and it’s devastating. Three specific rhetorical strategies dominate the corpus:

1. ADHD as Individual Suffering

Authors use the language of suffering and affliction: “children around the world suffer from ADHD,” ADHD is “a common cognitive disorder afflicting many children and adults,” its consequences are “devastating.” ADHD traits are called “undesirable,” “excessive,” and “invasive.”

Two rhetorical moves:

• Framing ADHD as deviant other. The condition is positioned as alien, separate from the person.
• Crystallizing ADHD as detached entity. The condition is treated as an invading force rather than a fundamental aspect of how the person perceives and processes the world.

This runs directly counter to ACM SIGACCESS’s own accessible writing guidelines, which recommend treating disability as one aspect of a person, not their defining characteristic.

2. ADHD as Burden to Others

Children with ADHD are “at risk” of underachievement, substance abuse, and criminality. ADHD is “a significant burden on those affected, their families and society,” “a big threat for public health,” and “can have a huge emotional and economic impact on families.”

The critical move: these papers frame the family unit as suffering because of the ADHD person, legitimizing intervention on the individual rather than addressing the social structures that create the burden. ADHD becomes “an urgent societal problem solved only by intervention on the individual, rather than on the society which problematises their embodiment.”

3. ADHD as Interesting Argument

Four papers use ADHD as a keyword/sales-pitch without actually addressing ADHD. Studies on interruption effects or modality uses invoke ADHD to make the research seem relevant, without interest in or regard for ADHD people. “ADHD is utilized here as a sales pitch and attention grabber, with little interest or regard for the people behind a catchy diagnostic criterion.”

Who Gets to Be a User?

The participation analysis is where the paper becomes genuinely uncomfortable reading.

The Numbers

Of 52 core corpus papers:

• 12 papers (23%) had no participants at all
• 25 papers (48%) did not report participant gender
• 32 papers (62%) focused on children
• 7 papers (13.5%) focused on youth/teenagers
• 10 papers (19%) studied adults
• Only ~12% of papers interviewed people with ADHD about their experiences

The Pattern: By-Proxy Design

Of 19 papers claiming user-centered design:

• Only 5 actually included ADHD people as co-designers
• 6 engaged teachers, therapists, or parents – but not the ADHD people themselves
• Projects that did consult children sometimes ignored their input when it conflicted with researchers’ agenda (e.g., children said they didn’t want data shared with parents/teachers – this was overridden)

The Power Structure

The paper identifies a consistent pattern: researchers consult proxy stakeholders (parents, teachers, clinicians) whose interests often conflict with the ADHD person’s interests. Teachers want quiet classrooms. Parents want compliant children. Clinicians want measurable symptom reduction. None of these are the same as what the ADHD person wants.

The sharpest line in the paper: “if people with ADHD are not systematically engaged in designing and making meaning of the technologies built to operate in their lives, the resulting artefacts will very likely embody and materialise neurotypical expectations more than self-determined support.”

Participant Resistance: The Users Fighting Back

This is the most underappreciated finding. The authors identify embodied resistance – ADHD participants actively subverting the technologies imposed on them.

Examples from the corpus:

• Tactile feedback meant as corrective (to suppress certain behaviors): ADHD participants instead enjoyed the sensation and deliberately triggered more of it, “resulting in paradoxically higher ‘scores’ on the game while rating lower on ‘performance’ in terms of the intervention’s behavioural aims.”
• Timer interventions for meal routines: caused “stress and frustration.” One participant “protested the intervention by hiding the timers during meal times.”
• Wearable sensors for body movement tracking: “the quality of the ADHD image is worse than that of the control image” because ADHD children’s bodies literally resisted detection – moving during acquisition, fiddling with electrodes, disrupting the signal.

The papers frame these as failures of compliance. Spiel et al. read them as evidence that the design is wrong. The users are telling you something. You’re just not listening because you’ve already decided what the correct behavior is.

Technologies in the Corpus: What’s Actually Being Built

Dominant Technology Types

• Wearables (10x): hugging vests, wristbands, belts, smartwatches
• EEG (4x): neurofeedback training, media consumption monitoring
• Smartphones (4x): general-purpose or time-structuring apps
• AR/VR (3x): skill training, time perception, brain-computer interfaces

Dominant Purposes

• Training focus/concentration/attention span (10 projects, mostly children)
• Self-regulation (emotional control, breathing, “motor excess” suppression)
• Suppressing impulsive action (“Blurtline” – a belt that detects when a child is about to speak impulsively by monitoring deep inhaling)
• Monitoring (daily activity tracking, in-situ prompts)
• Diagnosing (fMRI, eye tracking, movement data, touchscreen patterns)

What’s Missing

• Technologies supporting adults with ADHD
• Technologies supporting self-given tasks (not externally assigned)
• Technologies supporting ADHD strengths (hyperfocus, creativity, cognitive dynamism)
• Technologies oriented toward self-determined support rather than behavioral compliance
• Research on how ADHD people already use and appropriate existing technologies

The Crip Technoscience Alternative

Drawing on Hamraie and Fritsch’s Crip Technoscience Manifesto (2019), Spiel et al. speculate on what ADHD technology research could look like if it centered disabled people as knowers and makers. Four principles:

1. “Crip technoscience centers the work of disabled people as knowers and makers”

ADHD people are experts on their own experience. Research should position them not as data sources or test subjects but as collaborative partners in generating knowledge and designing technology. The authors invoke Design Justice principles: “prioritize design’s impact on the community over the intentions of the designer.”

2. “Crip technoscience is committed to access as friction”

Access is not about removing all friction to make things smooth. It’s about sitting with friction when it appears, asking why it exists, and understanding what dynamics produce it – rather than reflexively smoothing it over. This is anti-solutionism. When an ADHD participant resists a timer intervention, the crip response is not “fix the timer” but “why does this timer exist, whose norms does it enforce, and what would this person actually want?”

3. “Crip technoscience is committed to interdependence as political technology”

Everyone depends on others. This is especially visible for disabled people. Technologies should recognize and support interdependence rather than enforcing independence as the goal. Research should explicitly acknowledge the politics shaping its design.

4. “Crip technoscience is committed to disability justice”

ADHD research has real consequences: increased depression, increased suicide rates among people subjected to behaviorist interventions. Technologies that “solely aim at disciplining people with ADHD into acting more neurotypically” have “fundamentally negative impacts on their lives largely due to associated stigmas as seen in increased depression as well as suicide rates.”

Implications: What the Paper Recommends

6.1 Reconceptualize ADHD

Stop treating ADHD as pathology. Treat it as neurological variance. Recognize that many interventions aim to “train the ability to focus attention on a task” based on the assumption that “neurotypical modes of ‘paying attention’ are helpful for people with ADHD in the same way.” This ignores hyperfocus, ignores context-dependency of attention, and ignores that lab-trained attention skills don’t transfer to real-world contexts.

“Move away from behaviourist interventions that solely aim at disciplining people with ADHD into acting more neurotypically.”

6.2 Involve ND People as Partners

Advocate with IRBs and ethics committees for participatory approaches. Argue that ADHD children and adults are not “higher risk” than neurotypical participants. Use Design Justice principles. Use appreciative and respectful language.

6.3 Attend to Technology

Build tools that:

• Account for neurodivergence and ADHD as “a mere difference”
• Attend to existing strengths (ability-based design)
• Allow self-determined engagement
• Support flexible daily structuring with options, not mandates
• Support executive functioning on self-given tasks (not externally imposed compliance)
• Support ADHD people’s existing strategies rather than replacing them with researcher-designed interventions

6.4 Research Gaps to Fill

• Technologies for adults (the field is overwhelmingly child-focused)
• How ADHD people already appropriate existing technologies
• ADHD strengths (enhanced creativity, connections between domains, abstract thinking) – research “did not consider supporting any strengths in ADHD, given these were conceptually absent”
• Self-determined support vs externally defined task regimes

What Tool Sovereignty Looks Like in Practice

This is not a paper that informs tool design in specific ways. This is a paper that validates the entire paradigm of tool sovereignty – neurodivergent people shaping their own tools rather than accepting tools shaped for them.

Tool Sovereignty = Crip Technoscience in Practice

Spiel et al.’s critique generates one overarching demand: people with ADHD should be the knowers and makers of their own technologies, not the passive recipients of technologies designed by others to enforce neurotypical behavior.

Consider what this looks like concretely: a person with ADHD who cannot code uses AI to build bespoke tools shaped entirely to their own patterns and needs. No product manager, no clinician, no teacher deciding what’s correct. Every tool reflects the user’s workflow, not a generalized workflow.

A task management system, a knowledge base, an editorial pipeline, a conversation archive – each shaped to one person’s cognitive patterns. That is what crip technoscience looks like when you hand someone the means of production. Not “how can we accommodate ADHD in existing tools?” but “what tools does this person build when nobody else is deciding what the tool should do?”

Specific Design Principles from Spiel’s Critique

1. No behaviorist conditioning. No streaks, no rewards for “good behavior,” no penalties for missed days, no gamification that creates shame cycles.

2. No proxy stakeholders. The user is the designer. There is no gap between “who builds” and “who uses.” This eliminates the entire proxy-design problem Spiel identifies.

3. Friction is informative, not a bug. When a tool doesn’t work, it means the tool is wrong, not the person. Resistance is signal. When inbox items pile up, that’s information about task structure, not evidence of failure.

4. Support self-given tasks. Not “what should you be doing?” but “what do you want to do next?” Surfacing self-prioritized items, not externally imposed obligations.

5. Attend to strengths. Hyperfocus is not a disorder symptom. It’s a capability. Context-aware interruption management respects hyperfocus rather than interrupting it. AI-assisted workflows can rely on hyperfocus as the primary productivity mode.

6. Flexible temporal structures. No fixed schedules. No “overdue” language. Items wait. Sessions start when they start. Crip time as operational principle.

7. Self-determined support. Session-resumption tooling (e.g., “Where Was I?” hooks) is precisely the kind of tool Spiel envisions – supporting the person’s own re-entry strategy rather than imposing one.

Connection to the ICSE 2025 Curb-Cut Finding

Newman et al. (ICSE 2025) found that 80% of neurotypical developers also use the strategies ADHD developers rely on. The curb-cut effect: design for ADHD, benefit everyone.

Spiel’s paper provides the critical dimension this data point lacks: the curb-cut framing, while useful for the business case, is politically dangerous. If the justification for supporting ADHD developers is “it helps NT developers too,” you’ve subordinated neurodivergent needs to neurotypical benefit. The reason to design for ADHD is that ADHD people deserve tools that work for them, full stop. The curb-cut effect is a happy side effect, not the justification.

This is a tension in any project that designs for neurodivergence. Building for one person avoids the problem – but if findings ever generalize (“ADHD-informed tooling is good tooling”), both truths need to be held together: yes, it benefits everyone, AND the primary reason to do it is neurodivergent self-determination.

What Would Spiel Say About a Project Like This?

Reading the paper carefully and extrapolating from the crip technoscience framework, here is what Spiel would likely say about a project where an ADHD non-programmer uses AI to build bespoke tools:

Positive:

• This is exactly the kind of self-determined tool building the paper calls for. ADHD person as knower and maker.
• The absence of behaviorist patterns (no streaks, no shame language, no forced compliance) aligns with their recommendations.
• The AI-as-tool-builder paradigm represents a genuine paradigm shift that addresses the fundamental exclusion they document: you don’t need to be a programmer to build your own tools anymore.
• The research aspect (reviewing literature, extracting principles, building on evidence) models the “ADHD person as researcher” role they advocate for.

Critical:

• One person’s self-determined design is a case study, not a movement. The crip technoscience commitment to interdependence means tools should also support connection and community, not just individual optimization.
• Generalizing from n=1 – even self-determined n=1 – without engaging the broader ND community has limits. Spiel would push back on that.
• The “tool sovereignty” framing is individualistic. It could be enriched by the interdependence principle: what would tools look like that support neurodivergent collectives, not just individuals?
• The AI dependency itself is a power structure worth examining. AI coding tools are built by corporations. True tool sovereignty would interrogate that dependency.

Limitations of This Paper

1. All authors are white, Global North, with access to diagnosis. They acknowledge this directly. Access to diagnosis is itself a privilege shaped by class, race, and geography.

2. Corpus is limited to computing/HCI. The medical literature, clinical psychology literature, and education literature on ADHD technology are not covered. This is intentional (they want to critique their own field) but means the analysis doesn’t address clinical ADHD technology.

3. No constructive prototypes. This is pure critique – no alternative designs are built or tested. The “speculations on crip technoscience” section is explicitly speculative.

4. The extended corpus analysis is lighter. Only two of the four authors analyzed the 48 additional papers, using the existing codebook rather than fresh inductive coding.

5. The emotional methodology is double-edged. Reading as affected parties produces insights neutral reviewers miss, but also produces motivated reasoning risks. The authors are honest about this but it remains a limitation.

Open Questions

1. Has the field changed since 2022? The extended corpus showed some improvement (more participatory design, more adults), but the fundamental framing had not shifted. The explosion of ADHD + developer research since 2023 (Liebel, Newman, etc.) may represent a real shift, or may reproduce the same patterns in a new domain.

2. What does crip technoscience look like in the AI era? Spiel wrote before the LLM revolution. The possibility of non-programmers building their own tools is a qualitatively different situation from the one they analyzed. Does AI-assisted tool building count as self-determination if the AI embodies neurotypical assumptions?

3. Can the resistance data be formalized? The “embodied resistance” finding – ADHD participants subverting interventions – is fascinating but anecdotal. Could you design studies that specifically look for resistance as positive signal rather than compliance failure?

4. How do you do participatory design with ADHD adults who are not academics? The paper’s solutions assume institutional research contexts. What about commercial tool development, open source projects, or individual tool building?

Key Quotes Worth Keeping

“if people with ADHD are not systematically engaged in designing and making meaning of the technologies built to operate in their lives, the resulting artefacts will very likely embody and materialise neurotypical expectations more than self-determined support.”

“the focus of research in this corpus is not to support people with ADHD in their daily lives, but rather appears to be lying on bringing ‘relief’ to their parasocial environment through the upholding of neurotypical standards of behaviour and expression.”

“Technologies have a space in all our lives; but if they are largely relegated to categorising and altering the behaviours of isolated groups of people, we – as technology researchers – have a responsibility to provide alternatives.”

“We strongly recommend moving away from behaviourist interventions that solely aim at disciplining people with ADHD into acting more neurotypically. This has fundamentally negative impacts on their lives largely due to associated stigmas as seen in increased depression as well as suicide rates.”

Connections to Related Research

Paper	Connection
ICSE 2025 (Newman)	Curb-cut data supports universality, but Spiel warns against subordinating ND needs to NT benefit
Sankesara 2025	Empirical data on variability – Spiel would say: don’t use this to “fix” variability, use it to design for variability
Deshmukh 2025	Framework paper proposing nudges/sensing – Spiel would critique this as deficit-framed intervention on individuals
Mew 2025	AI-assisted ADHD programming – Spiel would ask: who chose the tasks? whose productivity metric?
Olinic 2025	Wearables review – many devices in this review are exactly the surveillance tools Spiel critiques
Parnin 2011	Interruption recovery research – Spiel would note this treats all programmers the same, missing ND-specific patterns
Leroy 2009/2018	Attention residue – useful cognitive science, but the “Ready-to-Resume Plan” is an externally imposed structure
Mark 2008	Interruption research – Spiel would note this is one of the papers that uses ADHD as “interesting argument” without studying ADHD people

Rating

Theoretical importance: 10/10. This is the paper that names the problem that tool sovereignty responds to.

Practical actionability: 4/10. No prototypes, no specific design patterns, no empirical data. It tells you what’s wrong and what direction to go, but not how to get there.

Methodological rigor: 8/10. The positioned methodology is novel, well-justified, and transparently executed. The emotional register may not satisfy positivist reviewers, but it produces insights that neutral analysis misses.

Relevance to tool sovereignty: 10/10. This is the theoretical foundation. Every design decision should be checkable against Spiel’s critique: does this tool enforce neurotypical compliance, or does it support self-determined engagement?