Codebook frameworks
The UX and user-research frameworks you can tag with, and what each one brings.
Bristlenose ships several codebooks built from established user-research and design frameworks. You can apply their tags by hand, and AutoCode proposes tags from them (see Use codebooks and AutoCode). Each brings a different lens to the same set of quotes, and you can layer more than one — tagging the same quotes through several frameworks at once.
Start here
Emotional & Cognitive Signals — seven sentiment categories the pipeline applies automatically, so every quote arrives pre-tagged. The negative signals are more granular than the positive ones because frustration, confusion, and doubt each point to a different design problem — performance, information architecture, and credibility respectively. Use this layer to orient: scan the lay of the land, find the interesting quotes, then hide it with the eye toggle as you develop tags of your own.
Bristlenose UXR Codebook — a general-purpose, domain-agnostic codebook: eight groups covering behaviour, pain points, needs and desires, mental models, motivation, emotion and trust, context, and learning. It reflects the practical vocabulary working researchers already use rather than any one theory, and works equally well for fishkeeping, fintech, or healthcare. The default starting point when no specific framework fits.
Design and usability frameworks
The Design of Everyday Things — Don Norman. Seven principles that explain why interactions succeed or fail: discoverability, feedback, conceptual models, signifiers, mapping, constraints, and the difference between slips and mistakes. A bottom-up lens for diagnosing why something is hard to use.
10 Usability Heuristics for User Interface Design — Jakob Nielsen. Ten rules of thumb — from "visibility of system status" to "help and documentation" — refined from a factor analysis of 249 real usability problems and adapted here for coding participant quotes. The most widely used checklist in usability evaluation, and especially strong at surfacing status confusion, control anxiety, and error-message quality.
The Elements of User Experience — Jesse James Garrett. Five mutually exclusive planes from abstract to concrete: strategy, scope, structure, skeleton, and surface. A top-down way to place a finding in the UX decision hierarchy — useful when the question is at what level a problem lives.
The User Experience Honeycomb — Peter Morville. Seven facets of a worthwhile experience: useful, usable, desirable, findable, accessible, credible, and valuable. Broader than ease-of-use alone, it is particularly good at surfacing credibility and findability findings that interaction-focused frameworks miss.
Laws of UX — Jon Yablonski. Psychological principles translated into design guidance — Hick's Law (choice paralysis), Fitts's Law (motor difficulty), Miller's Law (working-memory limits), Jakob's Law (expectation transfer), the peak-end rule, and the Gestalt principles. Each law names the specific mechanism behind a participant's reaction. Of the shipped frameworks, this is the one grounded most directly in the primary research described in Research foundations.
Specialist codebooks
Command-Line UX. For research on terminal-based tools — CLIs, developer tools, and sysadmin workflows. Six groups (discoverability, mental model, errors and recovery, workflow, learning and docs, trust and risk) tell you why the friction happened: flag ergonomics, an illegible error message, exit-code confusion, or a broken mental model of piping. Pair it with the sentiment layer — sentiment tells you the participant was frustrated, these tags tell you the cause. Written for API designers, DX engineers, and OSS maintainers.
Platonic Ontology & Epistemology. A scholarly codebook for tagging Plato's dialogues — Theory of Forms, knowledge versus opinion, dialectic method, soul and psychology, ethics and virtue, argumentation and irony — grounded in the readings of Plato scholars including Vlastos, Fine, Kraut, and Sedley. It expects philosophical arguments rather than buttons and navigation, and it demonstrates that a Bristlenose codebook can code any transcribed corpus, not only product interviews.
Choosing one
There is no single correct framework. Norman and Nielsen suit usability and interaction problems; Garrett and Morville suit broader experience and information-architecture questions; Yablonski's laws suit questions about attention, memory, and decision-making; Command-Line UX suits developer tooling. The sentiment layer is always present to orient you, and you can install as many of the others as you like and tag the same quotes through several lenses.
For the steps to install, apply, and AutoCode against a codebook, see Use codebooks and AutoCode.