Neural frame audit

Description

Elicits free associations to a product name or key term, codes them into the conceptual frames they activate, and checks whether those frames support or undermine the intended meaning.

Frame semantics holds that every word evokes a structured scene, with roles, expectations and values, and that people reason within the frame a word activates rather than from its dictionary definition. A neural frame audit applies this to product language. Participants give rapid free associations to a product name, feature label or category term; associations are coded into frames (for example 'wallet' evoking loss, theft, pickpocketing, cash or carrying) and the roles and inferences each frame licenses are made explicit. The audit compares the frame users actually activate with the frame the product team intended. Rigour comes from timed elicitation that limits deliberate rationalisation, structured coding with a second coder, and sampling across the segments whose frames are likely to differ, such as rural and urban or first-time and experienced users.

Attributes

Loss type
Unmigrated
Effort
Days

Steps

  1. List the 3 to 10 load-bearing terms to audit: product name, category noun, primary actions and key metaphors (for example 'wallet', 'top up', 'credit').
  2. Recruit 12 to 30 participants per segment whose frames may differ, and run a timed free-association task: show each term and collect the first 3 to 5 words or images within about 20 seconds, in the participant's own language.
  3. Follow with a short laddering probe on the first association: 'What happens in that situation? Who is involved? What could go wrong?' to surface the frame's roles and expected events.
  4. Code associations into candidate frames using a codebook developed on the first 20% of data; have a second coder code at least 25% and aim for Cohen's kappa of 0.70 or above before proceeding.
  5. For each term, tabulate frame frequencies by segment and record the inferences each dominant frame licenses (for example 'wallet' implies money can be stolen or lost).
  6. Compare dominant frames against the intended frame and flag terms where the intended frame is activated by fewer than about half of participants in any priority segment, or where a threat frame dominates.
  7. Generate alternative terms, re-run the elicitation on them with fresh participants, and select terms whose dominant frame matches intent.

Sources

  • Women Fire & Dangerous Things (Lakoff)
  • Metaphors We Live By (Lakoff)
  • Methods Atlas — frame audit (Gopinath)
  • The Inquiry Series, Ch 19 — Neural Framing & the Architecture of Meaning

Also known as

  • Frame audit
  • Free-association frame elicitation
  • Product name frame audit

Categories

  • Content
  • Linguistics, Semiotics & Discourse