Accessibility Research

What this book is

A case outside every published framework

The author's brother drowned at roughly two and a half years old and survived with hypoxic-anoxic brain injury, losing speech before he had established a voice or laid down the motor memories that voice-banking and reconstruction research depend on. He now communicates through an AAC pad using sustained pressure rather than discrete taps — an input modality the hardware wasn't designed around. ALS reconstruction work can model from a voice that once existed. BCI decoding for spinal cord injury can draw on years of speaking history. This case has no equivalent baseline, and the literature offers nothing to reconstruct from.

This book documents what happens when the map runs out before it reaches your case and has to be built before the work begins, with Claude used as a research navigator across four real domains: voice personalization for someone who never had a voice, AAC input signal design for an atypical motor profile, BCI speech decoding with no speech history to decode from, and the integration architecture connecting decoded neural output to a personalized voice model. Claude did not supply expertise the author lacked in any of the four domains — it supplied structured scaffolding for building the map, locating the gap, and drafting the instruments that gave unstructured research a form.

What you’ll learn

Eleven chapters, from the hardest case to the methodology document

• the-hardest-case — Characterizing a case that sits outside every published framework, and why that characterization is the first real piece of research work.
• research-map — Building a credible orientation to a domain from outside your trained background, before you know enough to know what you don't know.
• locating-your-gap — A four-layer structure — consensus, active research, speculation, open problem — for finding exactly where your case creates a gap worth occupying.
• designing-without-a-template — Designing a study when no published protocol fits your subject.
• minimal-sample-problem — Working the constraints n-of-one research imposes on design, analysis, and claims.
• building-tools-at-the-frontier — Building the coding schemes and observation protocols a lab would normally have built for you.
• reading-the-literature-with-claude — Using Claude as a domain-mapping partner for literature that assumes background you don't have.
• n-of-one-research-design — Research design specific to single-subject studies, and what changes when there is no cohort to average across.
• the-pipeline-vision — The integration architecture connecting decoded neural output to a personalized voice model.
• the-methodology-document — Writing a research statement that stands on its own without institutional backing behind it.
• conclusion — The asymmetry between observer access and subject consent that creates the obligation running through the whole record.

From the book

A preview

Abe has no equivalent baseline. The published literature offers nothing to reconstruct from: he never had a voice to record, his motor memory of speech production never formed, and the sustained-pressure input finding has no AAC profile built around it. These frameworks are sound; they simply do not reach this case.

Who it’s for

Researchers working without a lab or a template

This is for researchers working without a lab or institutional affiliation, outside the population their training covered, who need to build the methodology before the study can start — and for anyone using Claude as a thinking partner for research design questions no existing protocol resolves. It assumes ELAN and BORIS mechanics are already in place from Behavioral Annotation Pipelines (Book 41) and that the full methodology sequence — pre-registration, causal modeling, design cycle — is already understood from Building a Methodology (Book 20); it picks up specifically at the research-navigator problem.

A longer excerpt is available to newsletter subscribers.

Buy Ebook on Amazon