Framing Effect Audit¶
Essence¶
Framing Effect Audit is the intervention pattern for situations where a conclusion may depend on how information is presented. It asks whether people are responding to the underlying situation or to wording, ordering, visual emphasis, baseline choice, default structure, or reference frame. The archetype does not assume that one perfectly neutral frame exists. Instead, it makes frame sensitivity visible so the presentation can be redesigned, disclosed, pluralized, or selected with a defensible rationale.
The core move is to treat presentation as an audit object. The team identifies which frame variables may matter, constructs plausible alternate presentations, compares the resulting interpretations or decisions, and then decides what downstream users need to know.
Compression statement¶
When the same underlying information can produce different conclusions because of wording, order, visuals, baselines, defaults, or reference frames, identify the frame variables, compare alternate presentations, measure or interpret shifts, and choose, redesign, or disclose the frame with materiality in view.
Canonical formula: underlying_information + presentation_frame -> interpretation_shift; audit(frame_variables, alternate_presentations, response_shift, materiality_threshold) -> selection | redesign | disclosure | multi_frame_output
When to Use This Archetype¶
Use this archetype when a survey, dashboard, decision aid, evaluation, report, interface, or message will be treated as evidence, yet the response may change if the same information is presented differently. It is especially useful when wording, option order, chart scale, denominator, baseline, comparison class, label, default, or gain/loss language could change perceived risk, fairness, urgency, priority, or value.
It is less appropriate when the goal is simply to persuade. It also does not replace Frame Shift Intervention: if the task is to find a more useful interpretive frame for action, use a frame-shift pattern. Use Framing Effect Audit when the question is whether conclusions are stable across plausible presentations.
Structural Problem¶
The structural problem is that presentation choices can become hidden causal inputs into judgment. A decision maker may believe they are seeing a stable preference, risk estimate, ranking, or evaluation when they are actually seeing the output of a specific frame. Because presentation feels like packaging rather than substance, its effects are often under-documented.
This creates several risks: a survey can appear to measure public preference when it is partly measuring wording; a dashboard can appear to show priority when it is partly showing baseline choice; an evaluation can appear objective when evidence order or labels shape interpretation; a risk message can appear informative while steering action through gain/loss framing.
Intervention Logic¶
The intervention begins by naming the decision or inference that could become frame-sensitive. It then inventories possible frame variables and constructs alternate presentations. Those alternates are compared through user testing, expert review, split samples, sensitivity tables, or other mechanisms appropriate to the stakes.
The audit does not stop at detection. It applies a materiality threshold: did the frame change interpretation enough to matter for the decision? If not, the result may be treated as reasonably robust. If yes, the team must decide whether to redesign the presentation, disclose frame sensitivity, present multiple frames, or avoid relying on a single framed response.
Key Components¶
Framing Effect Audit treats presentation as an audit object: it asks whether a conclusion reflects the underlying situation or is partly the output of a particular frame. The work begins with a Framing Variable Inventory that names the presentation choices that might steer judgment — wording, order, labels, units, baselines, denominators, defaults, comparison classes, visual emphasis, and reference points — so the audit does not collapse into vague message testing. From that inventory the team constructs an Alternate Presentation Set: plausible versions of the same information designed to isolate meaningful frame variables rather than vary arbitrarily. A Response Comparison Design specifies how interpretation will be compared across frames — qualitative review, split sample, expert review, or simulation — observing the actual response or decision the frame may change.
The remaining components prevent the audit from producing false reassurance or unaccountable selection. The Frame Effect Threshold defines when an observed shift is large or consequential enough to require disclosure, redesign, or multi-frame presentation, with stricter thresholds at higher stakes — not every difference matters, but a hidden material difference does. The Equivalence or Difference Record states what information is held constant across presentations and what differs, so reviewers can distinguish a true framing effect from a substantive content change. Finally, the Frame Disclosure or Selection Rule closes the audit with an explicit decision about what to do with frame sensitivity — present one frame, multiple frames, add a disclosure note, redesign, or refuse to rely on a frame-sensitive result for the downstream use. The archetype's discipline is that detection alone is not enough: material frame sensitivity must drive a defensible output choice, not be quietly passed through.
| Component | Description |
|---|---|
| Framing Variable Inventory ↗ | This component names the presentation choices that might steer judgment. It includes wording, order, labels, units, baselines, denominators, defaults, comparison classes, visual emphasis, and reference points. Without this inventory, the work tends to become vague message testing. |
| Alternate Presentation Set ↗ | The alternate presentation set creates plausible versions of the same underlying information. These versions are not arbitrary variants; they are designed to isolate meaningful frame variables. When substantive content differs, that difference must be documented. |
| Response Comparison Design ↗ | A framing audit needs a way to compare interpretation across frames. The comparison may be qualitative, quantitative, split-sample, expert-review, or simulation-based. The key is that the design observes the response or decision that the frame may change. |
| Frame Effect Threshold ↗ | Not every difference matters. This component defines when an observed shift is large or consequential enough to require disclosure, redesign, or multi-frame presentation. The threshold should become stricter when stakes are higher. |
| Equivalence or Difference Record ↗ | This record protects the audit from false comparison. It states what information is held constant across presentations and what differs. It helps reviewers distinguish a true framing effect from a substantive information change. |
| Frame Disclosure or Selection Rule ↗ | The audit must end with a decision about what to do with frame sensitivity. The rule may call for one presentation, multiple presentations, a disclosure note, redesign, or rejection of a frame-sensitive result for a particular downstream use. |
Common Mechanisms¶
A wording test compares alternative labels, phrases, question stems, or descriptions. It is a mechanism, not the archetype, because the archetype also requires an explicit frame inventory, response comparison, materiality threshold, and disclosure or selection rule.
An order-effect check varies the sequence of questions, options, evidence, or visual elements. It implements the archetype when order is treated as a frame variable that may change interpretation.
A gain/loss frame comparison presents equivalent outcomes as gains, losses, avoided losses, costs, savings, survival, or failure. It is useful when valence or reference point changes perceived desirability or risk.
A dashboard framing review audits display choices such as units, scales, color, baselines, sorting, default filters, and missing denominators. The dashboard itself is not the archetype; it is the artifact whose presentation choices are being audited.
A survey frame split sample assigns people to alternate presentations so response differences can be estimated. It supports the audit when the split is designed around frame variables and downstream interpretation.
A visual framing audit reviews charts, images, icons, scale, color, cropping, and layout for salience or implied comparison effects. A presentation sensitivity table records the tested variables, alternate frames, observed shifts, thresholds, and final decision.
Parameter / Tuning Dimensions¶
The first tuning dimension is audit rigor. Low-stakes contexts may use expert review or a small comparison; high-stakes contexts may need split samples, blinded review, segmentation, and formal documentation.
The second dimension is frame breadth. A narrow audit may test one wording change; a broad audit may test wording, order, visual scale, denominator, baseline, and default structure together.
The third dimension is equivalence strictness. Some audits require mathematically equivalent presentations. Others compare plausible but not identical presentations, provided the substantive differences are documented.
The fourth dimension is audience segmentation. If different groups interpret frames differently, the audit may need segment-level comparison rather than average effects.
The fifth dimension is the output rule. The audit can select one frame, redesign a frame, disclose sensitivity, present multiple frames, or treat the result as too frame-sensitive for unsupported use.
Invariants to Preserve¶
The audit must preserve traceability from frame variable to observed response. It must also preserve honesty about equivalence: if the underlying information changes, the audit should say so.
It should preserve interpretive validity rather than optimizing for the designer's preferred answer. It should also preserve downstream transparency when frame effects are material. Finally, it should preserve the humility that all presentations frame reality to some degree; the goal is not frame-free communication but accountable framing.
Target Outcomes¶
A successful Framing Effect Audit tells decision makers whether a conclusion is robust across plausible presentations. It reveals when a survey response, dashboard interpretation, evaluation score, or risk judgment is partly presentation-dependent.
The practical outcomes are better disclosure, fairer presentation design, more cautious interpretation of frame-sensitive results, and reduced risk that hidden baselines, defaults, or labels will steer decisions without accountability.
Tradeoffs¶
The main tradeoff is rigor versus speed. More alternatives, more reviewers, and stronger comparison designs improve confidence but take time.
Another tradeoff is neutrality versus plural framing. Sometimes no single presentation is neutral enough, and the more honest output is to show multiple frames. That improves transparency but can increase cognitive load.
There is also a standardization tradeoff. A standard frame improves comparability, while tailored frames may improve comprehension for particular audiences. The audit should state which value matters for the downstream use.
Failure Modes¶
One failure mode is generic testing drift: the audit becomes ordinary A/B testing or usability testing. The mitigation is to require explicit frame variables, equivalence records, response comparison, and materiality thresholds.
Another failure mode is answer-seeking frame choice. The team selects the frame that produces the desired result and calls it validated. Blinded review, documentation, and separate selection criteria help reduce this risk.
A third failure mode is hidden frame sensitivity. The team detects material frame effects but passes a single result downstream without disclosure. The output rule should make disclosure, redesign, or multi-frame presentation mandatory when thresholds are crossed.
A fourth failure mode is false equivalence. Alternate presentations may differ in content, denominator, or scope. The equivalence or difference record is the mitigation.
Neighbor Distinctions¶
Framing Effect Audit is distinct from Frame Shift Intervention because it tests frame sensitivity rather than simply moving to a more useful frame. It is distinct from Reframing for Action because the goal is not to unlock action through a new frame, but to determine whether conclusions depend on the frame.
It is distinct from Cognitive Representation Externalization because externalization makes implicit mental structure inspectable, while this archetype audits how presentation alternatives change judgment. It is distinct from Meta-Symbolic Rule Reflection because it does not primarily revise a symbol or rule system; it compares how presentation choices affect interpretation.
It is distinct from generic A/B testing because optimization is not enough. A/B testing can implement the audit only when the test is explicitly about framing variables, interpretation validity, materiality, and downstream disclosure.
It is adjacent to Observer Effect Accounting and Representation Invariance Check. Observer Effect Accounting focuses on observation changing the observed system; Framing Effect Audit focuses on presentation changing the observer's response. Representation Invariance Check tests stability under equivalent representations; Framing Effect Audit includes that concern but adds disclosure and redesign choices.
Variants and Near Names¶
Survey Framing Audit focuses on question wording, order, labels, response options, and context effects. Dashboard Framing Audit focuses on displays, baselines, color, units, sorting, defaults, and visual emphasis. Gain/Loss Framing Audit focuses on equivalent outcomes presented as gains, losses, costs, savings, survival, or failure.
Reference-Point Framing Audit focuses on denominators, baselines, comparison classes, and defaults. Risk Communication Frame Audit focuses on risk formats such as absolute risk, relative risk, frequency, severity, and uncertainty language.
Near names include Presentation Sensitivity Test, Framing Effect Check, Frame Bias Check, Wording Effect Audit, and Visual Framing Review. Concrete names such as Wording Test, Order-Effect Check, Survey Wording Test, and Generic A/B Testing should usually be treated as mechanisms unless they preserve the full audit logic.
Cross-Domain Examples¶
In survey design, a team compares whether respondents answer differently when a program is described as a tax, investment, benefit, or service. The audit prevents a single phrasing from masquerading as stable public preference.
In operations, a dashboard team compares raw counts, incident rates, severity-weighted scores, and trend displays. The audit reveals whether priority judgments are driven by underlying risk or by display choices.
In risk communication, an organization compares absolute-risk, relative-risk, natural-frequency, and gain/loss presentations before publishing a decision aid. The audit makes sure the chosen frame supports fair understanding rather than hidden steering.
In evaluation, reviewers test whether seeing narrative examples before scores changes ratings. The audit surfaces whether evidence order frames the evaluation.
In product design, a team checks whether labels and default order make one option appear recommended when no recommendation is intended. The audit detects implied endorsement created by presentation.
Non-Examples¶
A campaign choosing the most persuasive wording and hiding alternatives is not Framing Effect Audit. It is persuasion optimization.
A facilitator reframing a conflict from blame to shared problem solving is not this archetype unless the intervention audits how alternate frames change interpretation. It is closer to a frame-shift pattern.
A designer improving spacing, typography, or visual polish is not this archetype unless the work tests whether those choices change judgment or decision.
A generic A/B test that optimizes conversion is not this archetype unless it explicitly treats presentation as a frame variable and addresses validity, materiality, and disclosure.