{"data":{"effects":[{"id":"spacing-effect","name":"Spacing Effect","domain":"memory","mechanism":"Learning distributed across time produces far better long-term retention than the same total time massed together, because each spaced retrieval requires effortful reconstruction.","replication_status":"robust","canonical_source":{"authors":"Cepeda, N. J., Pashler, H., Vul, E., Wixted, J. T., & Rohrer, D.","year":2006,"title":"Distributed Practice in Verbal Recall Tasks: A Review and Quantitative Synthesis","venue":"Psychological Bulletin, 132(3), 354-380","doi":"10.1037/0033-2909.132.3.354"},"supporting_sources":[],"typical_effect_size":"Large; retention gains often exceed 100% over massed practice","confidence":0.92,"caveats":["One of the most robust findings in all of psychology, and among the most ignored in product design.","Optimal gap scales with the retention interval you care about — longer target, wider spacing."],"applications":[{"context":"Onboarding and education","tactic":"Distribute feature introduction across sessions instead of one setup tour.","example":"Three short spaced prompts beat one long walkthrough for actual adoption.","expected_direction":"increase"}],"manipulation_baseline":5,"countermeasures":[],"related_effects":["testing-effect","serial-position-effect"],"moderators":[],"moderator_coverage":{"moderators_known":0,"reversals_known":0,"not_yet_researched":false,"no_cross_cultural_evidence":false,"note":"No moderator evidence recorded for this effect."}},{"id":"variable-ratio-reinforcement","name":"Variable Ratio Reinforcement","domain":"motivation","mechanism":"Rewards delivered after an unpredictable number of responses produce the highest and most extinction-resistant response rates of any reinforcement schedule.","replication_status":"robust","canonical_source":{"authors":"Ferster, C. B., & Skinner, B. F.","year":1957,"title":"Schedules of Reinforcement","venue":"Appleton-Century-Crofts"},"supporting_sources":[],"typical_effect_size":"Very large and highly reliable across species","confidence":0.91,"caveats":["This is the mechanism underlying slot machines and loot boxes. It is the single highest-risk item in this corpus.","Regulated as gambling in several jurisdictions when tied to monetary spend.","Applying it to engagement metrics reliably produces compulsive use rather than value."],"applications":[{"context":"Product delight","tactic":"Use sparingly and without monetary coupling — occasional unexpected value, not engineered unpredictability.","example":"An occasional genuinely useful surfaced insight, not a randomised reward drop.","expected_direction":"increase"}],"manipulation_baseline":90,"countermeasures":["Track time spent against value received.","Impose external session limits."],"related_effects":[],"moderators":[],"moderator_coverage":{"moderators_known":0,"reversals_known":0,"not_yet_researched":true,"no_cross_cultural_evidence":false,"note":"No moderator evidence in this corpus yet. This is a coverage gap, not a finding that the effect is unmoderated."}},{"id":"cognitive-load","name":"Cognitive Load","domain":"cognition","mechanism":"Working memory is sharply limited. Load imposed by presentation (extraneous load) competes with load intrinsic to the task, and once total load exceeds capacity, performance and completion collapse rather than degrade gently.","replication_status":"robust","canonical_source":{"authors":"Sweller, J.","year":1988,"title":"Cognitive Load During Problem Solving: Effects on Learning","venue":"Cognitive Science, 12(2), 257-285","doi":"10.1207/s15516709cog1202_4"},"supporting_sources":[],"typical_effect_size":"Large for learning and task-completion outcomes","confidence":0.9,"caveats":["Reducing extraneous load is nearly always safe. Reducing intrinsic load means removing genuine information, which can mislead."],"applications":[{"context":"Checkout and forms","tactic":"Remove every element that does not serve the current decision.","example":"Strip global navigation from the checkout step.","expected_direction":"increase"}],"manipulation_baseline":10,"countermeasures":["Slow down and externalise the decision when a page feels overwhelming."],"related_effects":["millers-law","choice-overload","processing-fluency"],"moderators":[{"kind":"moderator","condition":"learner-paced vs system-paced presentation","condition_type":"medium","direction":"reverses","magnitude":"d=0.93 under system-paced presentation vs d=-0.14 under self-paced presentation, k=43 independent effects","evidence_type":"meta-analysis","k_studies":43,"n_total":null,"citation":"Ginns, P. (). Meta-analysis of the modality effect. Learning and Instruction, 15(4), 313-331","source_url":"https://doi.org/10.1016/j.learninstruc.2005.07.001","citable":true,"confidence":"medium","note":"Same caveat on confidence as the row above. The modality benefit of narration-plus-visuals over on-screen text depends on the system controlling the pace. Give the learner a pause or scrub control and the advantage does not just shrink, it goes slightly negative — narration becomes a worse choice than text once the user can linger.","is_reversal":true},{"kind":"moderator","condition":"post-2011 reformulation of germane load","condition_type":"measurement","direction":"eliminates","magnitude":"unquantified — a theoretical redefinition, not an effect size","evidence_type":"narrative-review","k_studies":null,"n_total":null,"citation":"Sweller, J., van Merriënboer, J. J. G., & Paas, F. (). Cognitive Architecture and Instructional Design: 20 Years Later. Educational Psychology Review, 31(2), 261-292","source_url":"https://doi.org/10.1007/s10648-019-09465-5","citable":true,"confidence":"high","note":"The theory's own originators no longer treat germane load as a third additive load type that stacks with intrinsic and extraneous. Since 2011 it denotes working-memory resources redistributed toward learning-relevant processing of intrinsic load, not a separate quantity you can raise independently. A vendor pitch to 'increase germane load' as a lever distinct from managing intrinsic and extraneous load is arguing from a superseded version of the theory.","is_reversal":false},{"kind":"moderator","condition":"element interactivity (material complexity)","condition_type":"context","direction":"attenuates","magnitude":"modality effect d=0.72 under high element interactivity vs d=0.10 (near null) under low element interactivity, k=43 independent effects","evidence_type":"meta-analysis","k_studies":43,"n_total":null,"citation":"Ginns, P. (). Meta-analysis of the modality effect. Learning and Instruction, 15(4), 313-331","source_url":"https://doi.org/10.1016/j.learninstruc.2005.07.001","citable":true,"confidence":"medium","note":"Bibliographic details and the k=43 figure were independently confirmed against the publisher record; the effect sizes were not cross-checked against the full text, so treat as medium rather than high confidence. Splitting instructions across visual and audio channels only helps when the material is complex enough to overload a single channel — on simple material the effect is close to zero, so the redesign cost buys almost nothing.","is_reversal":false}],"moderator_coverage":{"moderators_known":3,"reversals_known":1,"not_yet_researched":false,"no_cross_cultural_evidence":false,"note":"Conditions under which this effect changes size or direction. Reversals are listed first."}},{"id":"serial-position-effect","name":"Serial Position Effect","domain":"memory","mechanism":"Items at the beginning (primacy, via rehearsal into long-term memory) and end (recency, via short-term memory) of a sequence are recalled better than those in the middle.","replication_status":"robust","canonical_source":{"authors":"Murdock, B. B.","year":1962,"title":"The Serial Position Effect of Free Recall","venue":"Journal of Experimental Psychology, 64(5), 482-488","doi":"10.1037/h0045106"},"supporting_sources":[],"typical_effect_size":"Large and highly reliable","confidence":0.9,"caveats":["Recency is disrupted by any filled delay after presentation."],"applications":[{"context":"Feature lists and navigation","tactic":"Place the two most important items first and last.","example":"Primary CTA first in the nav, contact last.","expected_direction":"increase"}],"manipulation_baseline":10,"countermeasures":[],"related_effects":["von-restorff-effect","peak-end-rule"],"moderators":[],"moderator_coverage":{"moderators_known":0,"reversals_known":0,"not_yet_researched":true,"no_cross_cultural_evidence":false,"note":"No moderator evidence in this corpus yet. This is a coverage gap, not a finding that the effect is unmoderated."}},{"id":"confirmation-bias","name":"Confirmation Bias","domain":"cognition","mechanism":"Information is sought, weighted, and remembered in ways that favour an existing belief, and disconfirming evidence is subjected to harsher scrutiny than confirming evidence.","replication_status":"robust","canonical_source":{"authors":"Wason, P. C.","year":1960,"title":"On the Failure to Eliminate Hypotheses in a Conceptual Task","venue":"Quarterly Journal of Experimental Psychology, 12(3), 129-140","doi":"10.1080/17470216008416717"},"supporting_sources":[{"authors":"Nickerson, R. S.","year":1998,"title":"Confirmation Bias: A Ubiquitous Phenomenon in Many Guises","venue":"Review of General Psychology, 2(2), 175-220","doi":"10.1037/1089-2680.2.2.175"}],"typical_effect_size":"Large and pervasive","confidence":0.9,"caveats":["Applies to your own A/B test interpretation as much as to your audience.","Not reduced by intelligence or expertise; expertise often strengthens it by supplying better counter-arguments."],"applications":[{"context":"Objection handling","tactic":"Surface disconfirming information yourself, before the reader finds it elsewhere.","example":"A candid \"who this is not for\" section.","expected_direction":"increase"}],"manipulation_baseline":35,"countermeasures":["Actively seek the strongest disconfirming evidence before concluding.","Pre-register what result would change your mind."],"related_effects":["availability-heuristic","overconfidence"],"moderators":[],"moderator_coverage":{"moderators_known":0,"reversals_known":0,"not_yet_researched":true,"no_cross_cultural_evidence":false,"note":"No moderator evidence in this corpus yet. This is a coverage gap, not a finding that the effect is unmoderated."}},{"id":"testing-effect","name":"Testing Effect (Retrieval Practice)","domain":"memory","mechanism":"Actively retrieving information strengthens retention more than restudying it for the same duration, because retrieval reconsolidates and strengthens the memory trace.","replication_status":"robust","canonical_source":{"authors":"Roediger, H. L., & Karpicke, J. D.","year":2006,"title":"Test-Enhanced Learning: Taking Memory Tests Improves Long-Term Retention","venue":"Psychological Science, 17(3), 249-255","doi":"10.1111/j.1467-9280.2006.01693.x"},"supporting_sources":[],"typical_effect_size":"Large; d≈0.5 to 0.8 at delayed test in general memory research","confidence":0.9,"caveats":["Learners consistently judge restudying as more effective, and are consistently wrong. Do not design from self-reported preference here.","Retrieval must be effortful but successful; failure without feedback does not help.","The global grade does not hold in mathematics learning — see domain_variants before applying this to a numeracy product."],"applications":[{"context":"Learning products and onboarding","tactic":"Replace passive review with low-stakes retrieval.","example":"A short recall prompt after a lesson rather than a summary to re-read.","expected_direction":"increase"}],"manipulation_baseline":5,"countermeasures":[],"related_effects":["spacing-effect","generation-of-effort"],"moderators":[],"moderator_coverage":{"moderators_known":0,"reversals_known":0,"not_yet_researched":false,"no_cross_cultural_evidence":false,"note":"No moderator evidence recorded for this effect."}},{"id":"survivorship-bias","name":"Survivorship Bias","domain":"cognition","mechanism":"Analysis restricted to surviving cases systematically distorts conclusions, because the failures that would contradict the pattern have been removed from the sample.","replication_status":"robust","canonical_source":{"authors":"Brown, S. J., Goetzmann, W., Ibbotson, R. G., & Ross, S. A.","year":1992,"title":"Survivorship Bias in Performance Studies","venue":"Review of Financial Studies, 5(4), 553-580","doi":"10.1093/rfs/5.4.553"},"supporting_sources":[],"typical_effect_size":"Can invert conclusions entirely rather than merely bias them","confidence":0.9,"caveats":["Directly undermines \"successful companies do X\" reasoning, including most growth advice.","Your own customer testimonials are a survivorship-biased sample by construction — churned users do not write them."],"applications":[{"context":"Evidence and case studies","tactic":"Report the denominator and say what happened to the cases that failed.","example":"\"Of 50 teams that started, 41 completed migration\" rather than 41 success stories.","expected_direction":"increase"}],"manipulation_baseline":30,"countermeasures":["Ask what happened to the cases that are not in the sample.","Seek the failure population explicitly."],"related_effects":["base-rate-neglect","availability-heuristic","confirmation-bias"],"moderators":[],"moderator_coverage":{"moderators_known":0,"reversals_known":0,"not_yet_researched":true,"no_cross_cultural_evidence":false,"note":"No moderator evidence in this corpus yet. This is a coverage gap, not a finding that the effect is unmoderated."}},{"id":"anchoring","name":"Anchoring","domain":"decision","mechanism":"An initially presented number becomes the reference point for subsequent numeric judgement. Adjustment away from it is insufficient, so the final estimate stays biased toward the anchor even when the anchor is known to be arbitrary or irrelevant.","replication_status":"robust","canonical_source":{"authors":"Tversky, A., & Kahneman, D.","year":1974,"title":"Judgment under Uncertainty: Heuristics and Biases","venue":"Science, 185(4157), 1124-1131","doi":"10.1126/science.185.4157.1124"},"supporting_sources":[{"authors":"Klein, R. A., et al.","year":2014,"title":"Investigating Variation in Replicability: A \"Many Labs\" Replication Project","venue":"Social Psychology, 45(3), 142-152","doi":"10.1027/1864-9335/a000178"}],"typical_effect_size":"d≈0.5 to 1.0 depending on anchor extremity","confidence":0.88,"caveats":["Attenuates but does not vanish with domain expertise.","Implausibly extreme anchors can trigger reactance and reverse the effect.","Requires the judgement to be genuinely uncertain; a buyer who knows the market price is barely moved."],"applications":[{"context":"Pricing page","tactic":"Present the highest tier first so lower tiers are evaluated against it.","example":"Enterprise $499 shown left-most, with Pro $99 read as modest by comparison.","expected_direction":"shift"},{"context":"Discount presentation","tactic":"Show the reference price adjacent to the sale price.","example":"was $240 / now $99","expected_direction":"increase"}],"manipulation_baseline":35,"countermeasures":["Consider the opposite: actively generate reasons the anchor is wrong.","Evaluate against an independently sourced market rate before viewing the offer."],"related_effects":["framing-effect","decoy-effect","left-digit-effect"],"moderators":[{"kind":"moderator","condition":"preferential/willingness-to-pay valuation task","condition_type":"context","direction":"attenuates","magnitude":"robust for numeric estimation but inconsistent/weaker in WTP valuation; effect shrinks when percentile gaps between anchors are small","evidence_type":"single-study","k_studies":null,"n_total":null,"citation":"Li, Y., et al. (). The robustness of anchoring effects on preferential judgments. Judgment and Decision Making","source_url":"https://www.cambridge.org/core/journals/judgment-and-decision-making/article/robustness-of-anchoring-effects-on-preferential-judgments/FDFCFCE27288DC24D0B157EDD4F6C80D","citable":true,"confidence":"medium","note":"Anchoring is far more reliable for numeric estimates than for how much someone will pay; do not assume lab WTP anchors transfer.","is_reversal":false},{"kind":"moderator","condition":"subliminal/incidental anchor presentation","condition_type":"measurement","direction":"eliminates","magnitude":"two highly powered preregistered replications found no subliminal anchoring effect","evidence_type":"multi-lab-replication","k_studies":2,"n_total":null,"citation":"Röseler, L., et al. (). Evidence against subliminal anchoring: Two close, highly powered, preregistered, and failed replication attempts. Journal of Experimental Social Psychology","source_url":"https://doi.org/10.1016/j.jesp.2020.104066","citable":true,"confidence":"medium","note":"Anchors must be consciously processed; incidental/subliminal number exposure does not reliably anchor.","is_reversal":false},{"kind":"moderator","condition":"extreme vs moderate anchor distance","condition_type":"dose","direction":"attenuates","magnitude":"curvilinear: moderate anchors produce stronger effects than extreme ones, but even implausible anchors (Gandhi=140) still shift judgments","evidence_type":"single-study","k_studies":null,"n_total":null,"citation":"Bahník, Š., et al. (). Individual Differences in Anchoring Effect: Evidence for the Role of Insufficient Adjustment. Psychological Reports / PMC6396698","source_url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC6396698/","citable":true,"confidence":"medium","note":"Extreme anchors give diminishing returns but do NOT reverse; the 'implausible anchor backfires' claim is unsupported.","is_reversal":false},{"kind":"cultural","condition":"10 countries (reanalysis/extension of Many Labs anchoring data)","condition_type":"individualism-collectivism","direction":"amplifies","magnitude":"Anchoring showed strong between-culture heterogeneity; Schwartz intellectual-autonomy and egalitarianism value orientations negatively predicted anchoring size. Contested against aggregate meta-analyses that find anchoring stable.","evidence_type":"multi-country-replication","k_studies":null,"n_total":6344,"citation":"Kakinohana, R.K., Pilati, R., Klein, R.A. (). Does Anchoring Vary Across Cultures? Expanding the Many Labs Analysis. European Journal of Social Psychology","source_url":"https://doi.org/10.1002/ejsp.2924","citable":true,"confidence":"low","note":"Contested: treat price/quantity anchors as broadly effective but expect meaningfully weaker anchoring in high intellectual-autonomy cultures.","is_reversal":false}],"moderator_coverage":{"moderators_known":4,"reversals_known":0,"not_yet_researched":false,"no_cross_cultural_evidence":false,"note":"Conditions under which this effect changes size or direction. Reversals are listed first."}},{"id":"negativity-bias","name":"Negativity Bias","domain":"emotion","mechanism":"Negative information is weighted more heavily than equivalently intense positive information in attention, memory, and judgement.","replication_status":"robust","canonical_source":{"authors":"Baumeister, R. F., Bratslavsky, E., Finkenauer, C., & Vohs, K. D.","year":2001,"title":"Bad Is Stronger Than Good","venue":"Review of General Psychology, 5(4), 323-370","doi":"10.1037/1089-2680.5.4.323"},"supporting_sources":[],"typical_effect_size":"Large and pervasive across domains","confidence":0.88,"caveats":["Explains why a single negative review outweighs several positive ones, and why unresolved complaints are disproportionately costly.","Fear-based marketing exploits this and reliably damages brand trust over time."],"applications":[{"context":"Reviews and objection handling","tactic":"Address the strongest objection directly and early rather than burying it.","example":"A visible, specific answer to the most common criticism.","expected_direction":"increase"}],"manipulation_baseline":40,"countermeasures":["Weight positive evidence deliberately to offset the asymmetry."],"related_effects":["loss-aversion","availability-heuristic"],"moderators":[{"kind":"moderator","condition":"age","condition_type":"population","direction":"reverses","magnitude":"meta-analysis of 100 studies, N=7,129: older adults show a significant positive-information processing bias, younger adults show the standard negativity-weighted pattern","evidence_type":"meta-analysis","k_studies":100,"n_total":7129,"citation":"Reed, A. E., Chan, L., & Mikels, J. A. (). Meta-analysis of the age-related positivity effect: age differences in preferences for positive over negative information. Psychology and Aging, 29(1), 1-15","source_url":"https://doi.org/10.1037/a0035194","citable":true,"confidence":"high","note":"This is a reversal, not a weakening. Fear and loss framing that works on a younger audience can underperform or backfire on an older one, which shows a measurable pull toward positive material instead. Age-target copy accordingly rather than assuming negativity bias is a human universal.","is_reversal":true},{"kind":"moderator","condition":"culture (East Asian vs Western)","condition_type":"culture","direction":"attenuates","magnitude":"Hong Kong participants showed significantly stronger positive attentional bias than UK participants on the Attentional Probe Task (t(73)=-2.38, p=.020, d=0.56); the difference did not appear on picture-based or interpretation-rating tasks","evidence_type":"cross-cultural-experiment","k_studies":null,"n_total":null,"citation":"Yiend, J., André, J., Smith, L., Chen, K., Toulopoulou, T., Chen, C., Sham, P., & Parkinson, B. (). Biased cognition in East Asian and Western cultures. PLOS ONE, 14(11), e0223358","source_url":"https://doi.org/10.1371/journal.pone.0223358","citable":true,"confidence":"high","note":"The moderator is itself moderated: it showed up on a word-based attention task but not on picture-based or interpretation tasks in the same study, so this is real evidence of cultural variation, not license to assume it generalises to every format. Worth flagging that this effect previously had no cross-cultural evidence at all in this corpus.","is_reversal":false}],"moderator_coverage":{"moderators_known":2,"reversals_known":1,"not_yet_researched":false,"no_cross_cultural_evidence":false,"note":"Conditions under which this effect changes size or direction. Reversals are listed first."}},{"id":"hyperbolic-discounting","name":"Hyperbolic Discounting","domain":"decision","mechanism":"Future rewards are discounted steeply in the near term and shallowly later, so preferences reverse as a deadline approaches. A choice made calmly in advance is reversed in the moment.","replication_status":"robust","canonical_source":{"authors":"Ainslie, G.","year":1975,"title":"Specious Reward: A Behavioral Theory of Impulsiveness and Impulse Control","venue":"Psychological Bulletin, 82(4), 463-496","doi":"10.1037/h0076860"},"supporting_sources":[{"authors":"Laibson, D.","year":1997,"title":"Golden Eggs and Hyperbolic Discounting","venue":"Quarterly Journal of Economics, 112(2), 443-477","doi":"10.1162/003355397555253"}],"typical_effect_size":"Large and consistent across domains","confidence":0.88,"caveats":["Explains why annual plans convert worse than monthly despite better value — the cost is immediate and the saving is distant.","Also explains why free trials work: the cost is deferred past the discounting horizon.","Exploiting it to obscure long-term cost is precisely what negative-option rules target."],"applications":[{"context":"Pricing structure","tactic":"Move benefit earlier and cost later, without concealing the total.","example":"Immediate onboarding value, first charge after value is demonstrated.","expected_direction":"increase"}],"manipulation_baseline":45,"countermeasures":["Compute total cost over the full commitment before deciding.","Pre-commit while calm rather than deciding at the deadline."],"related_effects":["loss-aversion","default-effect"],"moderators":[],"moderator_coverage":{"moderators_known":0,"reversals_known":0,"not_yet_researched":true,"no_cross_cultural_evidence":false,"note":"No moderator evidence in this corpus yet. This is a coverage gap, not a finding that the effect is unmoderated."}},{"id":"ambiguity-aversion","name":"Ambiguity Aversion","domain":"decision","mechanism":"Known risks are preferred to unknown ones, even when the unknown option has equal or better expected value. Missing information is treated as adverse.","replication_status":"robust","canonical_source":{"authors":"Ellsberg, D.","year":1961,"title":"Risk, Ambiguity, and the Savage Axioms","venue":"Quarterly Journal of Economics, 75(4), 643-669","doi":"10.2307/1884324"},"supporting_sources":[],"typical_effect_size":"Large and highly reliable","confidence":0.88,"caveats":["The best argument for publishing pricing. A hidden price is read as a bad price.","Also explains why vague trust badges underperform specific, verifiable claims."],"applications":[{"context":"Pricing and trust","tactic":"Remove unknowns rather than arguing benefits harder.","example":"Published prices, a named SLA, a real support response time.","expected_direction":"increase"}],"manipulation_baseline":15,"countermeasures":["Distinguish \"unknown\" from \"bad\" — ask rather than assume."],"related_effects":["loss-aversion","status-quo-bias"],"moderators":[],"moderator_coverage":{"moderators_known":0,"reversals_known":0,"not_yet_researched":true,"no_cross_cultural_evidence":false,"note":"No moderator evidence in this corpus yet. This is a coverage gap, not a finding that the effect is unmoderated."}},{"id":"planning-fallacy","name":"Planning Fallacy","domain":"decision","mechanism":"Task completion time is systematically underestimated because planning focuses on the specific successful path and neglects the distribution of similar past outcomes.","replication_status":"robust","canonical_source":{"authors":"Buehler, R., Griffin, D., & Ross, M.","year":1994,"title":"Exploring the \"Planning Fallacy\": Why People Underestimate Their Task Completion Times","venue":"Journal of Personality and Social Psychology, 67(3), 366-381","doi":"10.1037/0022-3514.67.3.366"},"supporting_sources":[],"typical_effect_size":"Large; underestimates of 30-50% are routine","confidence":0.88,"caveats":["Persists even among people who know about it and who have personally missed similar estimates before.","Relevant to your own roadmap as much as to how you present setup time to customers."],"applications":[{"context":"Setup and migration claims","tactic":"Quote times from measured distributions rather than the happy path.","example":"\"Most teams finish in 20-40 minutes\" rather than \"takes 15 minutes\".","expected_direction":"increase"}],"manipulation_baseline":20,"countermeasures":["Take the outside view: how long did comparable past tasks actually take?","Estimate from a reference class, not from the plan."],"related_effects":["overconfidence","base-rate-neglect"],"moderators":[],"moderator_coverage":{"moderators_known":0,"reversals_known":0,"not_yet_researched":true,"no_cross_cultural_evidence":true,"note":"No moderator evidence in this corpus yet. This is a coverage gap, not a finding that the effect is unmoderated."}},{"id":"social-proof","name":"Social Proof","domain":"social","mechanism":"Under uncertainty, the behaviour of similar others is used as evidence about the correct action, because imitation is a cheap and usually reliable inference.","replication_status":"robust","canonical_source":{"authors":"Cialdini, R. B.","year":1984,"title":"Influence: The Psychology of Persuasion","venue":"Harper Business"},"supporting_sources":[{"authors":"Goldstein, N. J., Cialdini, R. B., & Griskevicius, V.","year":2008,"title":"A Room with a Viewpoint: Using Social Norms to Motivate Environmental Conservation in Hotels","venue":"Journal of Consumer Research, 35(3), 472-482","doi":"10.1086/586910"}],"typical_effect_size":"Field effects commonly 5-30% behaviour change","confidence":0.87,"caveats":["Strength scales with perceived similarity of the referent group, not with raw numbers.","Advertising low adoption backfires: it communicates the opposite norm.","Fabricated counts, fake testimonials, and invented \"12 people viewing\" widgets are unlawful commercial practices in the EU and actionable under FTC rules."],"applications":[{"context":"Landing page","tactic":"Show specific adoption by a group the visitor identifies with.","example":"\"Used by 340 Series-A fintech teams\" beats \"Trusted by thousands\".","expected_direction":"increase"}],"manipulation_baseline":40,"countermeasures":["Ask whether the referent group actually shares your constraints."],"related_effects":["authority","unity"],"moderators":[{"kind":"moderator","condition":"low/counter-normative descriptive norm (boomerang)","condition_type":"context","direction":"reverses","magnitude":"households below the neighborhood average increased consumption after descriptive-norm feedback; adding an injunctive (approval) cue removed the boomerang","evidence_type":"field-experiment","k_studies":null,"n_total":null,"citation":"Schultz, P. W., et al. (). The Constructive, Destructive, and Reconstructive Power of Social Norms. Psychological Science","source_url":"https://assets.csom.umn.edu/assets/118375.pdf","citable":true,"confidence":"high","note":"Never advertise a low or 'better than average' descriptive norm; it drags good performers back toward the mean unless you add an injunctive signal.","is_reversal":true},{"kind":"moderator","condition":"dissimilar/out-group reference in hyper-polarized context","condition_type":"culture","direction":"reverses","magnitude":"Biden supporters increased mask-wearing intentions when told most Trump supporters never wore masks (backfire); standard effect held in a non-polarized context","evidence_type":"single-study","k_studies":4,"n_total":null,"citation":"Rand, D. G., & Yoeli, E. (). Descriptive norms can 'backfire' in hyper-polarized contexts. PNAS Nexus","source_url":"https://doi.org/10.1093/pnasnexus/pgae303","citable":true,"confidence":"medium","note":"Citing an out-group's behavior can push a polarized audience the opposite way; match the reference group to the audience.","is_reversal":true},{"kind":"moderator","condition":"field descriptive-norm energy intervention effect size","condition_type":"context","direction":"attenuates","magnitude":"Opower home-energy descriptive-norm effect d≈0.32 (avg 2.0% usage reduction); highly heterogeneous by baseline usage","evidence_type":"field-experiment","k_studies":null,"n_total":600000,"citation":"Allcott, H. (). Social norms and energy conservation. Journal of Public Economics","source_url":"https://www.sciencedirect.com/science/article/abs/pii/S0047272711000478","citable":true,"confidence":"high","note":"At field scale social-proof effects are small (≈2% behavior change), far below lab persuasion demos.","is_reversal":false},{"kind":"cultural","condition":"17 countries; higher conformity in collectivist (Fiji, Hong Kong, Brazil, Japan) vs individualist (USA, UK, France, Netherlands)","condition_type":"individualism-collectivism","direction":"amplifies","magnitude":"Asch-paradigm conformity effect sizes: Fiji 2.48, Hong Kong 1.93, Japan 1.42 vs France 0.56, Netherlands 0.74, USA 0.90.","evidence_type":"meta-analysis","k_studies":null,"n_total":133,"citation":"Bond, R., Smith, P.B. (). Culture and Conformity: A Meta-Analysis of Studies Using Asch's (1952b, 1956) Line Judgment Task. Psychological Bulletin","source_url":"https://doi.org/10.1037/0033-2909.119.1.111","citable":true,"confidence":"high","note":"Conformity/social-proof pressure is substantially larger in collectivist cultures; n_total is number of studies, not participants.","is_reversal":false},{"kind":"cultural","condition":"United States vs Poland (university students)","condition_type":"individualism-collectivism","direction":"amplifies","magnitude":"Social-proof appeal (listing how often peers had complied with similar requests) produced greater compliance among Poles than Americans.","evidence_type":"cross-cultural-experiment","k_studies":null,"n_total":200,"citation":"Cialdini, R.B., Wosinska, W., Barrett, D.W., Butner, J., Gornik-Durose, M. (). Compliance with a Request in Two Cultures: The Differential Influence of Social Proof and Commitment/Consistency on Collectivists and Individualists. Personality and Social Psychology Bulletin","source_url":"https://doi.org/10.1177/0146167299258006","citable":true,"confidence":"medium","note":"Peer/crowd proof carries more weight in collectivist audiences; weight social-proof widgets higher there.","is_reversal":false}],"moderator_coverage":{"moderators_known":5,"reversals_known":2,"not_yet_researched":false,"no_cross_cultural_evidence":false,"note":"Conditions under which this effect changes size or direction. Reversals are listed first."}},{"id":"von-restorff-effect","name":"Von Restorff (Isolation) Effect","domain":"memory","mechanism":"An item that differs perceptually from its neighbours attracts disproportionate attention and is recalled better, because distinctiveness aids encoding.","replication_status":"robust","canonical_source":{"authors":"von Restorff, H.","year":1933,"title":"Über die Wirkung von Bereichsbildungen im Spurenfeld","venue":"Psychologische Forschung, 18, 299-342","doi":"10.1007/BF02409636"},"supporting_sources":[],"typical_effect_size":"Large for the isolated item","confidence":0.87,"caveats":["Isolating more than one element destroys the effect for all of them.","Colour alone is insufficient for accessibility; pair with shape, weight, or position."],"applications":[{"context":"Pricing table","tactic":"Give exactly one tier a distinct treatment.","example":"One highlighted plan; every other tier visually identical.","expected_direction":"increase"}],"manipulation_baseline":15,"countermeasures":[],"related_effects":["serial-position-effect"],"moderators":[],"moderator_coverage":{"moderators_known":0,"reversals_known":0,"not_yet_researched":true,"no_cross_cultural_evidence":false,"note":"No moderator evidence in this corpus yet. This is a coverage gap, not a finding that the effect is unmoderated."}},{"id":"illusory-truth","name":"Illusory Truth Effect","domain":"cognition","mechanism":"Repeated statements are judged more likely to be true than novel ones, because familiarity is misread as evidence of prior verification. The effect operates even when the person knows the correct answer.","replication_status":"robust","canonical_source":{"authors":"Hasher, L., Goldstein, D., & Toppino, T.","year":1977,"title":"Frequency and the Conference of Referential Validity","venue":"Journal of Verbal Learning and Verbal Behavior, 16(1), 107-112","doi":"10.1016/S0022-5371(77)80012-1"},"supporting_sources":[{"authors":"Dechêne, A., Stahl, C., Hansen, J., & Wänke, M.","year":2010,"title":"The Truth About the Truth: A Meta-Analytic Review of the Truth Effect","venue":"Personality and Social Psychology Review, 14(2), 238-257","doi":"10.1177/1088868309352251"}],"typical_effect_size":"d≈0.4 meta-analytically; persists across delays","confidence":0.87,"caveats":["Works on statements the person could evaluate correctly if prompted — knowing better is not protective.","This is the mechanism behind misinformation persistence. Using it to make a weak claim feel true is deception, not marketing.","Repeating a *correction* can paradoxically entrench the original claim if the falsehood is restated in the process."],"applications":[{"context":"Consistent messaging","tactic":"Repeat a small number of true, verifiable claims verbatim across touchpoints.","example":"The same specific benefit sentence on the site, in email, and in the product.","expected_direction":"increase"}],"manipulation_baseline":55,"countermeasures":["Check the source of a familiar-feeling claim rather than trusting the feeling.","Track whether you believe something because of evidence or because of exposure."],"related_effects":["processing-fluency","mere-exposure","availability-heuristic"],"moderators":[{"kind":"moderator","condition":"prior truth judgment / item type / veracity cues at exposure","condition_type":"measurement","direction":"attenuates","magnitude":"corrected overall g=0.37 [0.30,0.44]; heterogeneity explained by item type, first-exposure instructions, veracity cues, and presentation duration","evidence_type":"meta-analysis","k_studies":182,"n_total":31184,"citation":"Ye, S., et al. (). Systematic review and meta-analysis of the evidence for an illusory truth effect and its determinants. Nature Communications","source_url":"https://doi.org/10.1038/s41467-026-70041-x","citable":true,"confidence":"high","note":"Prompting people to evaluate a claim's truth on first contact reduces later repetition-driven belief; effect is small-to-moderate, not large.","is_reversal":false},{"kind":"moderator","condition":"WEIRD convenience samples only","condition_type":"culture","direction":"attenuates","magnitude":"all 181 eligible studies conducted at Western universities with mostly convenience samples; generality untested","evidence_type":"narrative-review","k_studies":181,"n_total":null,"citation":"Henderson, E. L., Westwood, S. J., & Simons, D. J. (). A reproducible systematic map of research on the illusory truth effect. Psychonomic Bulletin & Review","source_url":"https://doi.org/10.3758/s13423-021-01995-w","citable":true,"confidence":"medium","note":"The illusory-truth evidence base is entirely WEIRD; treat non-Western generalization as unproven.","is_reversal":false}],"moderator_coverage":{"moderators_known":2,"reversals_known":0,"not_yet_researched":false,"no_cross_cultural_evidence":false,"note":"Conditions under which this effect changes size or direction. Reversals are listed first."}},{"id":"processing-fluency","name":"Processing Fluency","domain":"cognition","mechanism":"Ease of mental processing is misattributed to properties of the stimulus itself, so material that is easier to read is judged more true, more familiar, and more likeable.","replication_status":"robust","canonical_source":{"authors":"Reber, R., Winkielman, P., & Schwarz, N.","year":1998,"title":"Effects of Perceptual Fluency on Affective Judgments","venue":"Psychological Science, 9(1), 45-48","doi":"10.1111/1467-9280.00008"},"supporting_sources":[],"typical_effect_size":"Small to moderate, but extremely broad in scope","confidence":0.86,"caveats":["Fluency raises believability regardless of accuracy, which is why clear writing carries an ethical obligation to be correct.","Excessive simplification can reduce perceived expertise for technical audiences."],"applications":[{"context":"All copy","tactic":"Raise contrast, shorten sentences, cut clause depth, use concrete nouns.","example":"Rewriting a 34-word sentence into two 12-word ones measurably lifts comprehension.","expected_direction":"increase"}],"manipulation_baseline":15,"countermeasures":["Judge claims on evidence, not on how effortless they were to read."],"related_effects":["mere-exposure","cognitive-load"],"moderators":[{"kind":"moderator","condition":"general/main effect across conditions","condition_type":"measurement","direction":"eliminates","magnitude":"no significant main effect of disfluency on retention or transfer across two experiments, N=76 and N=74; corroborated by an independent meta-analysis (recall d=-0.01, transfer d=0.03)","evidence_type":"meta-analysis","k_studies":null,"n_total":150,"citation":"Beege, M., Mengelkamp, C., & Nebel, S. (). Challenging the notion of a general disfluency effect: the moderating effect of element-interactivity on perceptual disfluency. Frontiers in Education, 10, art. 1731080","source_url":"https://doi.org/10.3389/feduc.2025.1731080","citable":true,"confidence":"high","note":"There is no general 'make it harder to read so people remember it' effect. Disfluency reliably increases extraneous load and degrades people's confidence in what they learned, with no compensating benefit on average. The popular disfluency-aids-learning claim does not hold as a blanket rule.","is_reversal":false},{"kind":"moderator","condition":"low element interactivity (simple material)","condition_type":"context","direction":"amplifies","magnitude":"disfluency significantly improved retention only under low element interactivity (p=0.043, p=0.045 across two experiments); no benefit under high element interactivity","evidence_type":"single-study","k_studies":null,"n_total":150,"citation":"Beege, M., Mengelkamp, C., & Nebel, S. (). Challenging the notion of a general disfluency effect: the moderating effect of element-interactivity on perceptual disfluency. Frontiers in Education, 10, art. 1731080","source_url":"https://doi.org/10.3389/feduc.2025.1731080","citable":true,"confidence":"medium","note":"The exception to the row above, and the reason the general null does not mean 'disfluency never helps'. On simple material — low element interactivity — a harder-to-read presentation raised retention, plausibly by forcing exactly the extra processing simple material would otherwise skip. This is a single study, not yet independently replicated: treat as a lead, not a settled finding.","is_reversal":false}],"moderator_coverage":{"moderators_known":2,"reversals_known":0,"not_yet_researched":false,"no_cross_cultural_evidence":true,"note":"Conditions under which this effect changes size or direction. Reversals are listed first."}},{"id":"base-rate-neglect","name":"Base Rate Neglect","domain":"decision","mechanism":"Specific, vivid case information dominates judgement while background frequency is underweighted, producing badly calibrated probability estimates.","replication_status":"robust","canonical_source":{"authors":"Kahneman, D., & Tversky, A.","year":1973,"title":"On the Psychology of Prediction","venue":"Psychological Review, 80(4), 237-251","doi":"10.1037/h0034747"},"supporting_sources":[],"typical_effect_size":"Large; attenuates when frequencies are presented as natural counts","confidence":0.86,"caveats":["Substantially reduced by expressing probabilities as natural frequencies (\"8 in 1,000\") instead of percentages.","Directly relevant to how you present conversion statistics and testimonial evidence."],"applications":[{"context":"Evidence presentation","tactic":"State outcomes as natural frequencies with the denominator visible.","example":"\"312 of 400 teams completed setup in under an hour\" beats \"78% success rate\".","expected_direction":"increase"}],"manipulation_baseline":30,"countermeasures":["Ask for the denominator before accepting any rate."],"related_effects":["availability-heuristic","confirmation-bias"],"moderators":[],"moderator_coverage":{"moderators_known":0,"reversals_known":0,"not_yet_researched":true,"no_cross_cultural_evidence":false,"note":"No moderator evidence in this corpus yet. This is a coverage gap, not a finding that the effect is unmoderated."}},{"id":"framing-effect","name":"Framing Effect","domain":"decision","mechanism":"Logically equivalent descriptions of the same outcome produce different preferences, because gains and losses are evaluated against a reference point rather than in absolute terms. Gain frames encourage risk aversion; loss frames encourage risk seeking.","replication_status":"robust","canonical_source":{"authors":"Tversky, A., & Kahneman, D.","year":1981,"title":"The Framing of Decisions and the Psychology of Choice","venue":"Science, 211(4481), 453-458","doi":"10.1126/science.7455683"},"supporting_sources":[],"typical_effect_size":"d≈0.5, large in the classic Asian-disease paradigm","confidence":0.85,"caveats":["Effect size shrinks substantially with numerate audiences and with joint (side-by-side) evaluation.","Loss framing raises attention but also raises anxiety and can suppress conversion for considered purchases."],"applications":[{"context":"Benefit copy","tactic":"Test gain-framed against loss-framed statements of the identical fact.","example":"\"Keep 98% of revenue\" vs \"Lose only 2% to fees\"","expected_direction":"shift"}],"manipulation_baseline":30,"countermeasures":["Restate the offer in the opposite frame before deciding."],"related_effects":["loss-aversion","anchoring"],"moderators":[{"kind":"moderator","condition":"publication-bias correction / analysis method","condition_type":"measurement","direction":"amplifies","magnitude":"Kühberger 1998 raw d=0.31; p-curve corrected d=0.52; Many Labs gain-loss framing d=0.60","evidence_type":"meta-analysis","k_studies":136,"n_total":30000,"citation":"Steiger, A., & Kühberger, A. (). A Meta-Analytic Re-Appraisal of the Framing Effect. Zeitschrift für Psychologie","source_url":"https://doi.org/10.1027/2151-2604/a000321","citable":true,"confidence":"high","note":"Risky-choice framing is real and robust (d≈0.5-0.6), not a p-hacking artifact.","is_reversal":false},{"kind":"moderator","condition":"framing task manipulation (risky vs attribute vs goal)","condition_type":"context","direction":"attenuates","magnitude":"no single 'framing effect'; size varies by manipulation type; goal framing d≈0.44","evidence_type":"meta-analysis","k_studies":136,"n_total":30000,"citation":"Kühberger, A. (). The Influence of Framing on Risky Decisions: A Meta-analysis. Organizational Behavior and Human Decision Processes","source_url":"https://doi.org/10.1006/obhd.1998.2781","citable":true,"confidence":"high","note":"Specify which framing type you mean; attribute framing is the most reliable, goal framing weaker.","is_reversal":false},{"kind":"cultural","condition":"WEIRD vs less-WEIRD samples across 28 classic effects, 36 countries (moderation test)","condition_type":"individualism-collectivism","direction":"unchanged","magnitude":"Very little heterogeneity attributable to WEIRDness; only 1 of 28 effects had tau>.20; variability driven more by which effect than by sample/setting.","evidence_type":"multi-country-replication","k_studies":null,"n_total":15305,"citation":"Klein, R.A., Vianello, M., Hasselman, F., et al. (). Many Labs 2: Investigating Variation in Replicability Across Samples and Settings. Advances in Methods and Practices in Psychological Science","source_url":"https://doi.org/10.1177/2515245918810225","citable":true,"confidence":"high","note":"For most classic effects, WEIRD vs non-WEIRD moderation is small; default to 'probably generalizes' unless a specific effect (self, choice, fairness, endowment) is on the reversal list.","is_reversal":false},{"kind":"cultural","condition":"North American vs South Korean samples (attribute framing)","condition_type":"individualism-collectivism","direction":"amplifies","magnitude":"\"Korean participants were markedly more influenced by the valence of the frame than North American participants,\" attributed to greater prevention focus (no single effect size in abstract).","evidence_type":"cross-cultural-experiment","k_studies":null,"n_total":200,"citation":"Cheon, B.K., Nam, G., Kim, J., Lee, S., Park, H.G., Kim, Y.H. (Front. Psychol. attribute-framing team) (). Cultural Variability in the Attribute Framing Effect. Frontiers in Psychology","source_url":"https://doi.org/10.3389/fpsyg.2021.754265","citable":true,"confidence":"low","note":"Positive/negative attribute framing (e.g. '90% fat-free') may bite harder for prevention-focused Korean/East Asian audiences; test both valences. Verify exact author order at the DOI before citing.","is_reversal":false}],"moderator_coverage":{"moderators_known":4,"reversals_known":0,"not_yet_researched":false,"no_cross_cultural_evidence":false,"note":"Conditions under which this effect changes size or direction. Reversals are listed first."}},{"id":"authority","name":"Authority","domain":"social","mechanism":"Signals of legitimate expertise or position shift compliance, because deferring to credible expertise is normally an efficient shortcut around costly evaluation.","replication_status":"robust","canonical_source":{"authors":"Milgram, S.","year":1963,"title":"Behavioral Study of Obedience","venue":"Journal of Abnormal and Social Psychology, 67(4), 371-378","doi":"10.1037/h0040525"},"supporting_sources":[{"authors":"Burger, J. M.","year":2009,"title":"Replicating Milgram: Would People Still Obey Today?","venue":"American Psychologist, 64(1), 1-11","doi":"10.1037/a0010932"}],"typical_effect_size":"Large for compliance with legitimate authority","confidence":0.85,"caveats":["Depends on perceived legitimacy in the relevant domain; borrowed authority from an unrelated field is easily detected and backfires.","Displaying certifications or affiliations you do not hold is fraud, not persuasion."],"applications":[{"context":"Credibility block","tactic":"Surface domain-relevant credentials, audits, and named expertise.","example":"SOC 2 Type II report link, not a generic security badge graphic.","expected_direction":"increase"}],"manipulation_baseline":35,"countermeasures":["Verify the credential is real and relevant to this specific claim."],"related_effects":["social-proof"],"moderators":[],"moderator_coverage":{"moderators_known":0,"reversals_known":0,"not_yet_researched":true,"no_cross_cultural_evidence":true,"note":"No moderator evidence in this corpus yet. This is a coverage gap, not a finding that the effect is unmoderated."}},{"id":"reactance","name":"Psychological Reactance","domain":"motivation","mechanism":"A perceived threat to behavioural freedom produces motivational arousal to restore it, often by doing the opposite of what is urged. This is the ceiling on persuasive pressure.","replication_status":"robust","canonical_source":{"authors":"Brehm, J. W.","year":1966,"title":"A Theory of Psychological Reactance","venue":"Academic Press"},"supporting_sources":[{"authors":"Steindl, C., Jonas, E., Sittenthaler, S., Traut-Mattausch, E., & Greenberg, J.","year":2015,"title":"Understanding Psychological Reactance: New Developments and Findings","venue":"Zeitschrift für Psychologie, 223(4), 205-214","doi":"10.1027/2151-2604/a000222"}],"typical_effect_size":"Moderate to large where the freedom threat is explicit","confidence":0.85,"caveats":["The single most important brake on every other effect in this corpus. Aggressive application of scarcity, urgency, or social proof triggers it and reverses the intended direction.","Stronger in individualist cultures and among people with high trait autonomy.","Phrases that restore choice (\"but it is up to you\") measurably reduce it."],"applications":[{"context":"Any persuasive copy","tactic":"Explicitly preserve the reader’s freedom to decline.","example":"\"No pressure — you can keep the free plan as long as you like.\"","expected_direction":"increase"}],"manipulation_baseline":5,"countermeasures":["Notice whether you are refusing on the merits or refusing to be pushed."],"related_effects":["scarcity","social-proof","commitment-consistency"],"moderators":[{"kind":"moderator","condition":"forceful/controlling ('you must') language","condition_type":"medium","direction":"reverses","magnitude":"threatening language increases perceived freedom threat and reactance; reactance negatively associated with persuasion (r≈-.18, k=39)","evidence_type":"meta-analysis","k_studies":35,"n_total":10658,"citation":"Ma, R., Ma, Z., Kalny, C. S., & Walter, N. (). Words that trigger: a meta-analysis of threatening language, reactance, and persuasion in health. Journal of Communication","source_url":"https://doi.org/10.1093/joc/jqaf004","citable":true,"confidence":"high","note":"Commanding language ('you must act now') backfires; autonomy-supportive phrasing persuades better.","is_reversal":true},{"kind":"cultural","condition":"Individualist (US/German) vs collectivist / interdependent-primed samples","condition_type":"individualism-collectivism","direction":"attenuates","magnitude":"Reactance to threats to personal freedom depends on self-construal; interdependent selves show reactance for threats to collective rather than personal freedom (unquantified).","evidence_type":"cross-cultural-experiment","k_studies":null,"n_total":200,"citation":"Jonas, E., Graupmann, V., Kayser, D.N., Zanna, M., Traut-Mattausch, E., Frey, D. (). Culture, Self, and the Emergence of Reactance: Is There a 'Universal' Freedom?. Journal of Experimental Social Psychology","source_url":"https://doi.org/10.1016/j.jesp.2009.05.006","citable":true,"confidence":"low","note":"Hard-sell/pressure that threatens personal autonomy triggers reactance in individualist audiences; collective-freedom threats matter more for interdependent selves. n_total approximate.","is_reversal":false}],"moderator_coverage":{"moderators_known":2,"reversals_known":1,"not_yet_researched":false,"no_cross_cultural_evidence":false,"note":"Conditions under which this effect changes size or direction. Reversals are listed first."}},{"id":"halo-effect","name":"Halo Effect","domain":"social","mechanism":"A salient positive attribute biases evaluation of unrelated attributes, without the evaluator being aware the transfer occurred.","replication_status":"robust","canonical_source":{"authors":"Nisbett, R. E., & Wilson, T. D.","year":1977,"title":"The Halo Effect: Evidence for Unconscious Alteration of Judgments","venue":"Journal of Personality and Social Psychology, 35(4), 250-256","doi":"10.1037/0022-3514.35.4.250"},"supporting_sources":[],"typical_effect_size":"Moderate to large","confidence":0.85,"caveats":["Why visual polish raises perceived reliability and security — attributes the design cannot actually speak to.","Cuts both ways: one visible defect drags unrelated judgements down with it."],"applications":[{"context":"Design quality","tactic":"Invest in craft on the first surface a visitor sees.","example":"A fast, typographically careful landing page raises trust in the infrastructure behind it.","expected_direction":"increase"}],"manipulation_baseline":35,"countermeasures":["Evaluate each attribute against evidence specific to it."],"related_effects":["processing-fluency","authority"],"moderators":[],"moderator_coverage":{"moderators_known":0,"reversals_known":0,"not_yet_researched":true,"no_cross_cultural_evidence":true,"note":"No moderator evidence in this corpus yet. This is a coverage gap, not a finding that the effect is unmoderated."}},{"id":"self-reference-effect","name":"Self-Reference Effect","domain":"memory","mechanism":"Information encoded in relation to the self is remembered better, because the self-schema is a richly elaborated structure that supports deeper encoding.","replication_status":"robust","canonical_source":{"authors":"Rogers, T. B., Kuiper, N. A., & Kirker, W. S.","year":1977,"title":"Self-Reference and the Encoding of Personal Information","venue":"Journal of Personality and Social Psychology, 35(9), 677-688","doi":"10.1037/0022-3514.35.9.677"},"supporting_sources":[{"authors":"Symons, C. S., & Johnson, B. T.","year":1997,"title":"The Self-Reference Effect in Memory: A Meta-Analysis","venue":"Psychological Bulletin, 121(3), 371-394","doi":"10.1037/0033-2909.121.3.371"}],"typical_effect_size":"d≈0.5 meta-analytically","confidence":0.85,"caveats":["The real basis for personalisation — relevance to the reader’s situation, not inserting their first name.","Token personalisation (\"Hi {{name}}\") does not engage the self-schema and is widely recognised as automated."],"applications":[{"context":"Copy and onboarding","tactic":"Frame benefits in terms of the reader’s specific situation and role.","example":"Segment-specific landing copy beats a name merge field by a wide margin.","expected_direction":"increase"}],"manipulation_baseline":20,"countermeasures":[],"related_effects":["processing-fluency","unity"],"moderators":[{"kind":"cultural","condition":"Mainland Chinese (immigrants to US) vs American norms","condition_type":"individualism-collectivism","direction":"attenuates","magnitude":"For Chinese participants, memory for self-referenced and mother-referenced items is comparable (no self>close-other advantage); Westerners show a self advantage even over the mother.","evidence_type":"cross-cultural-experiment","k_studies":null,"n_total":60,"citation":"Gutchess lab (Sparks et al.) (). Influence of Acculturation and Cultural Values on the Self-Reference Effect. Scientific Reports","source_url":"https://doi.org/10.1038/s41598-023-46210-z","citable":true,"confidence":"medium","note":"'Make it about YOU' personalization boosts recall in Western audiences; for East Asian audiences, referencing close family can be as memorable as self-reference.","is_reversal":false}],"moderator_coverage":{"moderators_known":1,"reversals_known":0,"not_yet_researched":false,"no_cross_cultural_evidence":false,"note":"Conditions under which this effect changes size or direction. Reversals are listed first."}},{"id":"generation-of-effort","name":"Generation Effect","domain":"memory","mechanism":"Information generated by the learner is remembered better than the same information read passively, because generation requires deeper semantic processing.","replication_status":"robust","canonical_source":{"authors":"Slamecka, N. J., & Graf, P.","year":1978,"title":"The Generation Effect: Delineation of a Phenomenon","venue":"Journal of Experimental Psychology: Human Learning and Memory, 4(6), 592-604","doi":"10.1037/0278-7393.4.6.592"},"supporting_sources":[],"typical_effect_size":"Moderate to large, highly reliable","confidence":0.85,"caveats":["Requires the learner to succeed at generating. Too-hard generation without support produces frustration and abandonment."],"applications":[{"context":"Onboarding","tactic":"Have the user produce their own first artefact rather than watch a demo.","example":"Build a real query against their own data during setup.","expected_direction":"increase"}],"manipulation_baseline":10,"countermeasures":[],"related_effects":["testing-effect","ikea-effect"],"moderators":[],"moderator_coverage":{"moderators_known":0,"reversals_known":0,"not_yet_researched":false,"no_cross_cultural_evidence":false,"note":"No moderator evidence recorded for this effect."}},{"id":"overconfidence","name":"Overconfidence","domain":"decision","mechanism":"Confidence in judgements exceeds their accuracy, appearing as overestimation of performance, overplacement relative to others, and excessive precision in interval estimates.","replication_status":"robust","canonical_source":{"authors":"Moore, D. A., & Healy, P. J.","year":2008,"title":"The Trouble with Overconfidence","venue":"Psychological Review, 115(2), 502-517","doi":"10.1037/0033-295X.115.2.502"},"supporting_sources":[],"typical_effect_size":"Large for precision; direction of overplacement varies with task difficulty","confidence":0.85,"caveats":["Moore & Healy show the three forms dissociate — on hard tasks people often *under*place themselves while remaining overprecise.","Do not treat \"people are overconfident\" as a single uniform effect."],"applications":[{"context":"Estimates and forecasts","tactic":"Ask for ranges wide enough to be right 90% of the time, then widen them.","example":"Confidence intervals on projected results rather than point estimates.","expected_direction":"shift"}],"manipulation_baseline":20,"countermeasures":["Widen every interval estimate; calibration training measurably helps.","Track prediction accuracy over time."],"related_effects":["planning-fallacy","confirmation-bias","dunning-kruger"],"moderators":[{"kind":"cultural","condition":"Western vs East Asian (self-enhancement / better-than-average)","condition_type":"individualism-collectivism","direction":"reverses","magnitude":"Meta-analysis: Westerners self-enhanced significantly more than East Asians on 30 of 31 methods; average cultural difference d=0.84 (large); East Asians showed near-zero self-enhancement.","evidence_type":"meta-analysis","k_studies":null,"n_total":91,"citation":"Heine, S.J., Hamamura, T. (). In Search of East Asian Self-Enhancement. Personality and Social Psychology Review","source_url":"https://doi.org/10.1177/1088868306294587","citable":true,"confidence":"high","note":"'You're above average' and aspirational self-enhancement copy is a Western lever; East Asian audiences trend self-critical, so modesty/self-improvement framing fits better. n_total = number of comparisons, not participants.","is_reversal":true},{"kind":"cultural","condition":"China vs Japan vs United States (probability calibration)","condition_type":"individualism-collectivism","direction":"amplifies","magnitude":"Chinese subjects markedly more overconfident/worse-calibrated on general-knowledge items; Japanese and American calibration were essentially the same (the 'Japanese exception').","evidence_type":"cross-cultural-experiment","k_studies":null,"n_total":300,"citation":"Yates, J.F., Zhu, Y., Ronis, D.L., Wang, D.F., Shinotsuka, H., Toda, M. (). Probability Judgment Accuracy: China, Japan, and the United States. Organizational Behavior and Human Decision Processes","source_url":"https://doi.org/10.1016/0749-5978(89)90048-4","citable":true,"confidence":"medium","note":"Do not assume East Asians are uniformly less overconfident; Chinese samples are MORE overconfident while Japanese match US calibration. n_total approximate.","is_reversal":false}],"moderator_coverage":{"moderators_known":2,"reversals_known":1,"not_yet_researched":false,"no_cross_cultural_evidence":false,"note":"Conditions under which this effect changes size or direction. Reversals are listed first."}},{"id":"active-learning","name":"Active Learning","domain":"cognition","mechanism":"Instruction requiring students to do something and reflect on it — rather than listen — improves performance, plausibly by forcing retrieval and exposing gaps while an instructor is present to address them.","replication_status":"robust","canonical_source":{"authors":"Freeman, S., Eddy, S. L., McDonough, M., Smith, M. K., Okoroafor, N., Jordt, H., & Wenderoth, M. P.","year":2014,"title":"Active Learning Increases Student Performance in Science, Engineering, and Mathematics","venue":"Proceedings of the National Academy of Sciences, 111(23), 8410-8415","doi":"10.1073/pnas.1319030111"},"supporting_sources":[{"authors":"Crouch, C. H., & Mazur, E.","year":2001,"title":"Peer Instruction: Ten Years of Experience and Results","venue":"American Journal of Physics, 69(9), 970-977","doi":"10.1119/1.1374249"}],"typical_effect_size":"d=0.47 on exam performance across 225 studies; failure rates fall from 34% to 22%","confidence":0.85,"caveats":["Measured against lecture in university STEM classrooms. It says nothing about solo, self-paced software, where there is no lecture to be better than.","\"Active\" covers a wide range of interventions, from clicker questions to full studio courses, and the meta-analysis pools them."],"applications":[],"manipulation_baseline":5,"countermeasures":[],"related_effects":["testing-effect","generation-of-effort"],"moderators":[],"moderator_coverage":{"moderators_known":0,"reversals_known":0,"not_yet_researched":false,"no_cross_cultural_evidence":false,"note":"No moderator evidence recorded for this effect."}}],"total":61,"excluded_failed":0,"corpus":{"total":61,"by_replication_status":{"robust":35,"mixed":14,"contested":6,"failed":6},"version":"2026-07"},"field_vs_lab":{"finding":"Nudge and default interventions produce roughly one sixth of the published effect when run at scale: 1.4 percentage points of take-up in two government nudge units against 8.7 points in academic journals.","k_studies":126,"n_total":23000000,"citation":"DellaVigna, S., & Linos, E. (2022). RCTs to Scale: Comprehensive Evidence From Two Nudge Units. Econometrica, 90(1), 81-116","source_url":"https://doi.org/10.3982/ECTA18709","cause":"The paper does not attribute the gap mainly to publication bias. Its one stated fraction: controlling for nudge characteristics — topic, delivery channel, institutional constraints — explains about two thirds of the gap, explicitly not via statistical power. A separate bias-correction model shrinks the academic estimate from 8.7pp to 3.16pp, still above the 1.4pp field figure, so genuine differences in what was implemented account for real ground the first figure alone would not show. The authors state directly that they cannot separate the two forces cleanly.","implication":"Treat any effect size in this corpus as an upper bound on what a live deployment will produce, unless the row itself is a field estimate. Do not assume the gap closes just by fixing publication incentives — real differences in how an intervention is delivered explain at least as much."}},"_meta":{"request_id":"req_c05b4253caa44cabb3b539e9e66303af","product":"psych/effects","version":"2.0.0","receipt_id":null,"cost":{"quoted":0,"charged":0,"currency":"USD","display":"0.000000"},"cache":{"hit":true,"age_s":41222},"provenance":{"sourcing_class":"B","upstream":["internal:psych-corpus@2026-07"],"fetched_at":"2026-10-13T11:52:02.850Z","license":"proprietary"},"freshness":{"as_of":"2026-07","staleness_s":41222},"confidence":1}}