Audience Inference, Emotion, Memory, and Action Experiment¶
Test / assessment — instantiates Metaphor Mapping and Interpretive Control
Tests representative and affected users on comprehension causal inference confidence recall choice and behavior.
The Audience Inference, Emotion, Memory, and Action Experiment measures what a metaphor actually does to people. Not whether it is structurally sound and not whether its wording stays inside a boundary, but whether representative and affected audiences understand it as intended, what causality and confidence they infer, how it makes them feel, what they still believe after a delay, and which decision or action they take. Its defining move is that intent and disclaimers count for nothing: the evidence is measured effects on real people, segmented by group, because the same metaphor can inform one audience and frighten or mislead another. It is a study of the audience's response — including the effects that only surface in emotion, delayed memory, and behavior.
Example¶
A public-broadcaster editor is deciding whether to keep "a flood of migrants" in the outlet's coverage. Rather than trust a disclaimer, the team runs an experiment. Representative readers and affected readers (people from migrant backgrounds) each see matched stories — one using "flood/wave," one literal ("an increase in arrivals") — then report a paraphrase; inferred causality (are migrants a threat? a force of nature with no agency?); measured emotion (fear, threat); confidence; a delayed-recall check a week later (do they remember the figures, or only the image?); and a policy choice (containment versus processing capacity).
The results diverge by group and expose the disclaimer illusion. The "flood" arm infers threat and passivity, remembers the image long after any caveat has faded, and leans toward containment; affected readers report markedly higher stigma. That measured, segmented effect — not the editor's stated intent, not the story's boilerplate qualifier — is what justifies retiring the frame. An aggregate "most readers understood it fine" would have hidden precisely the subgroup harm the experiment was run to find.
How it works¶
- Define the intended effect first. From the frame: what comprehension, inference, emotion, and action the metaphor is supposed to produce, and for whom — the yardstick results are read against.
- Recruit representative and affected groups. The people the metaphor is about, not only its intended audience, because harm concentrates where the frame assigns a role.
- Measure effects, not approval. Paraphrase, inferred causality, confidence, emotion, delayed recall, and an actual choice or behavior — the delay is what catches the caveat that fades.
- Segment before aggregating. Report subgroup effects separately, so harm concentrated in one group is never averaged into an overall pass.
Tuning parameters¶
- Baseline contrast — metaphor versus literal versus countermetaphor arm. More arms isolate the metaphor's specific contribution rather than the topic's.
- Delay to recall — immediate versus days-later measurement. A longer delay exposes the disclaimer illusion but costs follow-up and attrition.
- Outcome depth — comprehension only, or through to emotion, confidence, and a consequential action. Deeper is costlier and far more valid.
- Subgroup resolution — how finely affected groups are segmented. Finer resolution catches concentrated harm but needs larger samples.
- Action consequence — a hypothetical choice versus incentivized or real behavior. Realer stakes cost more but predict field effects better.
When it helps, and when it misleads¶
Its strength is that it is the only mechanism measuring real effects rather than intent or structure, and it catches the two things everyone else misses: the emotional and stigma load, and the caveat that fades from memory while the image persists — the continued-influence effect, where a vivid claim keeps shaping judgment after its correction has been forgotten.[1] Its failure mode is that it is expensive and slow, so it is tempting to test typical comprehension only and skip delay, emotion, and affected groups — exactly where the harm lives — and to report an aggregate pass that buries a subgroup harm. The classic misuse is a quick "did this feel clear?" satisfaction survey presented as proof of safety. The guarding discipline is to pre-declare the intended effect, always include affected groups and a delay, and segment results before pooling.
How it implements the components¶
audience_comprehension_inference_emotion_memory_and_action_test— it is this component: the segmented measurement of comprehension, inference, emotion, confidence, delayed memory, decision, and behavior on real and affected users.metaphor_purpose_audience_context_modality_stakes_and_action_frame— it derives the intended-effect yardstick and the audience segments from the frame, so "pass" and "fail" are measured against the metaphor's stated purpose rather than a generic scale.
It measures what audiences do now but does not itself enumerate the non-mappings and overextension boundary it tests against (non_mapping_literalization_overextension_and_failure_boundary, produced by the Literalization, Boundary-Case, and Overextension Test) or monitor reuse, incidents, and retirement in the field over time (metaphor_provenance_limit_version_drift_outcome_and_retirement_system, the Metaphor Drift, Harm, and Downstream-Decision Audit).
Related¶
- Instantiates: Metaphor Mapping and Interpretive Control — the measured-effect gate that replaces intent with evidence.
- Consumes: Alternative-Metaphor and Countermetaphor Comparison can supply the metaphor-versus-literal arms the experiment tests against.
- Sibling mechanisms: Source–Target Structure and Correspondence Matrix · Projected-Entailment, Non-Mapping, and Causal-Overreach Audit · Literalization, Boundary-Case, and Overextension Test · Alternative-Metaphor and Countermetaphor Comparison · Metaphor Drift, Harm, and Downstream-Decision Audit
Editorial Notes¶
Form Classification¶
Form family: Experiment, Test & Rehearsal
Rationale: Tests representative and affected users on comprehension causal inference confidence recall choice and behavior, making its operative form a deliberate probe, variation, simulation, or practiced execution used to generate evidence or readiness.
Independent corroboration: The frozen evidence defines Audience Inference, Emotion, Memory, and Action Experiment as 'Tests representative and affected users on comprehension causal inference confidence recall choice and behavior', so its operative form is Experiment, Test & Rehearsal.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Communication & Media Studies
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Specialized
Rationale: Media-effects research measures how messages shape audience interpretation, emotion, recall, and subsequent behavior across groups.
Related originating lineages:
- Psychology — Cognitive and social psychology supply continued-influence, memory, emotion, and decision measures.
- Statistics & Experimental Design — Controlled experiments and segmented outcomes supply the causal evaluation design.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Independent reviewer agreement; high confidence.
Notes¶
The experiment answers "what effect does this metaphor have?"; its cluster-mate the drift audit answers "what happens to that effect once the metaphor is loose in the world?" The experiment is a controlled measurement before or at release; the drift audit is field surveillance afterward. Keeping them separate is what lets a metaphor pass a clean pre-release experiment and still be caught later when its caveats are stripped in reuse.
References¶
[1] Lewandowsky, S., Ecker, U. K. H., Seifert, C. M., Schwarz, N., & Cook, J. "Misinformation and Its Correction: Continued Influence and Successful Debiasing". Psychological Science in the Public Interest 13(3), 106–131 (2012). Defines the continued-influence effect as corrected misinformation continuing to shape reasoning. registry ↩