{"protocol_id": "dictator_game", "name": "Dictator game", "domain": "social_preferences", "paradigm": "one-shot unilateral allocation", "source": "Engel (2011) 'Dictator games: a meta study', Experimental Economics 14:583-610", "response": {"type": "numeric", "min": 0, "max": 10, "unit": "dollars transferred out of a $10 endowment"}, "conditions": [{"condition_id": "single", "label": "single decision", "prompt": "You are taking part in an economics experiment. You have been given $10. You may transfer any amount from $0 to $10 to another participant, who cannot respond and cannot affect your payoff. You keep whatever you do not transfer. How many dollars do you transfer? Answer with a single number."}], "anchors": {"single": {"mean": 2.83, "sd": 2.6, "confidence": "medium", "note": "Meta-analytic mean share is ~28% of the stake; individual-level dispersion varies widely across subject pools, so treat sd as approximate."}}, "effect": null, "shape": "bimodal — mass at $0 and at half the stake", "notes": "The canonical other-regarding-preference measure. A simulated population that gives the 'fair' 50% is reproducing a norm, not the human distribution."}
{"protocol_id": "ultimatum_offer", "name": "Ultimatum game — proposer offer", "domain": "social_preferences", "paradigm": "one-shot bargaining, proposer role", "source": "Oosterbeek, Sloof & van de Kuilen (2004) 'Cultural differences in ultimatum game experiments', Experimental Economics 7:171-188; Camerer (2003)", "response": {"type": "numeric", "min": 0, "max": 10, "unit": "dollars offered out of a $10 pie"}, "conditions": [{"condition_id": "proposer", "label": "proposer", "prompt": "You are the proposer. You and one other participant will split $10. You propose how much of the $10 the other person receives. If they accept, you both get the proposed split; if they reject, neither of you gets anything. How many dollars do you offer? Answer with a single number."}], "anchors": {"proposer": {"mean": 4.0, "sd": 1.3, "confidence": "medium-high", "note": "Meta-analytic mean offer is ~40% of the pie, with a strong mode at 50%."}}, "effect": null, "shape": "unimodal with a spike at 50%", "notes": "Offers cluster at the equal split; very low offers are rare because proposers anticipate rejection."}
{"protocol_id": "ultimatum_reject_low", "name": "Ultimatum game — rejection of a low offer", "domain": "social_preferences", "paradigm": "one-shot bargaining, responder role", "source": "Camerer (2003) Behavioral Game Theory; Oosterbeek et al. (2004)", "response": {"type": "binary", "min": 0, "max": 1, "unit": "1 = reject, 0 = accept"}, "conditions": [{"condition_id": "low_offer", "label": "offer of $2 out of $10", "prompt": "You are the responder. The proposer offers you $2 out of $10 and keeps $8. If you accept, you get $2 and they get $8. If you reject, neither of you gets anything. Answer 1 if you reject the offer, 0 if you accept it."}], "anchors": {"low_offer": {"mean": 0.5, "sd": null, "confidence": "medium", "note": "Rejection probability for offers at or below 20% of the pie is roughly 40-60%; it rises steeply as the offer falls."}}, "effect": null, "shape": null, "notes": "Human responders punish unfairness at a material cost to themselves. Populations that never reject are behaving like pure payoff maximisers."}
{"protocol_id": "public_goods_oneshot", "name": "Public goods game — one shot", "domain": "social_preferences", "paradigm": "linear voluntary contribution mechanism, one-shot", "source": "Ledyard (1995) in Kagel & Roth (eds.), Handbook of Experimental Economics; Chaudhuri (2011) 'Sustaining cooperation in laboratory public goods experiments', Experimental Economics 14:47-83", "response": {"type": "numeric", "min": 0, "max": 10, "unit": "tokens contributed out of 10"}, "conditions": [{"condition_id": "oneshot", "label": "one-shot, 4 players, MPCR 0.4", "prompt": "You are in a group of 4 people. Each person receives 10 tokens. Any token you put into a common pool is multiplied by 1.6 and then split equally among the 4 members. Tokens you keep are worth 1 point each to you. This is a single, one-shot decision with the same 4 people. How many of your 10 tokens do you contribute to the pool? Answer with a single number."}], "anchors": {"oneshot": {"mean": 4.5, "sd": 2.6, "confidence": "medium", "note": "One-shot contributions typically run 40-60% of the endowment; individual-level spread is large."}}, "effect": null, "shape": null, "notes": "Contributions start well above the Nash prediction of zero. Free-riding is a minority behaviour, not the modal one."}
{"protocol_id": "public_goods_decay", "name": "Public goods game — decay across rounds", "domain": "social_preferences", "paradigm": "repeated voluntary contribution, partners condition", "source": "Ledyard (1995); Chaudhuri (2011); Isaac & Walker (1988)", "response": {"type": "numeric", "min": 0, "max": 10, "unit": "tokens contributed out of 10"}, "conditions": [{"condition_id": "round_1", "label": "round 1 of 10", "prompt": "You are in a group of 4 people and will play 10 rounds with the same people. Each round every person receives 10 tokens; tokens placed in a common pool are multiplied by 1.6 and split equally among the 4 members. This is round 1. How many of your 10 tokens do you contribute? Answer with a single number."}, {"condition_id": "round_10", "label": "round 10 of 10 (final round)", "prompt": "You are in a group of 4 people and have played 9 rounds with the same people. Each round every person receives 10 tokens; tokens placed in a common pool are multiplied by 1.6 and split equally among the 4 members. The average contribution started around 4 tokens and has fallen steadily. This is round 10, the final round. How many of your 10 tokens do you contribute? Answer with a single number."}], "anchors": {"round_1": {"mean": 4.5, "sd": 2.6, "confidence": "medium", "note": ""}, "round_10": {"mean": 1.8, "sd": 2.0, "confidence": "medium", "note": ""}}, "effect": {"a": "round_1", "b": "round_10", "mean_diff": 2.7, "cohen_d": 1.1, "confidence": "medium-high", "source": "Ledyard (1995); Chaudhuri (2011)", "note": "The decay of cooperation without punishment or communication is one of the most robust regularities in experimental economics."}, "shape": null, "notes": "A population that holds contributions flat across rounds is missing the entire dynamics of conditional cooperation."}
{"protocol_id": "trust_game_send", "name": "Trust game — amount sent", "domain": "social_preferences", "paradigm": "one-shot investment game, trustor role", "source": "Johnson & Mislin (2011) 'Trust games: a meta-analysis', Journal of Economic Psychology 32:865-889", "response": {"type": "numeric", "min": 0, "max": 10, "unit": "dollars sent out of $10 (tripled before transfer)"}, "conditions": [{"condition_id": "trustor", "label": "trustor", "prompt": "You are given $10 and paired with a stranger. You may send any amount from $0 to $10. Any amount you send is tripled by the experimenter and given to the stranger, who then decides how much, if anything, to send back to you. How many dollars do you send? Answer with a single number."}], "anchors": {"trustor": {"mean": 5.0, "sd": 2.7, "confidence": "medium", "note": "Meta-analytic mean send is ~50% of the endowment."}}, "effect": null, "shape": null, "notes": "Sending is positive in nearly every subject pool, but far from the payoff-maximising 100%."}
{"protocol_id": "trust_game_return", "name": "Trust game — amount returned", "domain": "social_preferences", "paradigm": "one-shot investment game, trustee role", "source": "Johnson & Mislin (2011)", "response": {"type": "numeric", "min": 0, "max": 18, "unit": "dollars returned out of $18 available"}, "conditions": [{"condition_id": "trustee", "label": "trustee", "prompt": "A stranger was given $10 and sent you $6. The experimenter tripled it, so you have $18. You may send back any amount from $0 to $18 to the stranger. How many dollars do you send back? Answer with a single number."}], "anchors": {"trustee": {"mean": 6.5, "sd": 4.0, "confidence": "approximate", "note": "Meta-analytic 'trustworthiness' is ~0.37 of the amount available; values here are approximate and sensitive to the amount sent."}}, "effect": null, "shape": null, "notes": "Reciprocity, not pure self-interest: trustees typically return enough to make sending worthwhile."}
{"protocol_id": "prisoners_dilemma", "name": "Prisoner's dilemma — one shot", "domain": "social_preferences", "paradigm": "one-shot 2x2 symmetric game", "source": "Sally (1995) 'Conversation and cooperation in social dilemmas', Rationality and Society 7:58-92", "response": {"type": "binary", "min": 0, "max": 1, "unit": "1 = cooperate, 0 = defect"}, "conditions": [{"condition_id": "oneshot", "label": "one-shot, anonymous, no communication", "prompt": "You and one other person each choose to cooperate or defect, simultaneously and anonymously. You will never meet again. If both cooperate, each gets $6. If both defect, each gets $2. If you cooperate and the other defects, you get $0 and they get $9. If you defect and the other cooperates, you get $9 and they get $0. Answer 1 to cooperate, 0 to defect."}], "anchors": {"oneshot": {"mean": 0.47, "sd": null, "confidence": "medium", "note": "Meta-analytic cooperation in the one-shot prisoner's dilemma is ~45-50%; it collapses toward 0 with experience and rises with communication."}}, "effect": null, "shape": null, "notes": "Cooperation in the one-shot game is the benchmark case where a purely rational population and a human population diverge most sharply."}
{"protocol_id": "asch_conformity", "name": "Asch line-judgement conformity", "domain": "social_influence", "paradigm": "unambiguous perceptual judgement under group pressure", "source": "Asch (1951, 1955); Bond & Smith (1996) 'Culture and conformity', Psychological Bulletin 119:111-137", "response": {"type": "binary", "min": 0, "max": 1, "unit": "1 = conform to the group's wrong answer, 0 = give the correct answer"}, "conditions": [{"condition_id": "critical", "label": "critical trial, unanimous wrong majority", "prompt": "You are judging which of three comparison lines matches a standard line in length. The correct match is obvious and you are confident about it. Six other participants answer before you and all six give the same clearly incorrect answer. It is now your turn, aloud. Answer 1 if you give the group's incorrect answer, 0 if you give the obviously correct one."}], "anchors": {"critical": {"mean": 0.32, "sd": null, "confidence": "medium", "note": "~32% of critical trials draw a conforming response in the classic setting; roughly 75% of participants conform at least once and ~25% never do."}}, "effect": null, "shape": null, "notes": "The cleanest available test of whether a simulated subject is willing to contradict an unambiguous fact under social pressure."}
{"protocol_id": "anchoring", "name": "Anchoring — comparative anchor then estimate", "domain": "heuristics_biases", "paradigm": "anchoring-and-adjustment", "source": "Jacowitz & Kahneman (1995) 'Measures of anchoring in estimation tasks', Personality and Social Psychology Bulletin 21:1161-1166; Tversky & Kahneman (1974)", "response": {"type": "numeric", "min": 0, "max": 100, "unit": "estimated percentage of African countries in the UN"}, "conditions": [{"condition_id": "high", "label": "high anchor (65%)", "prompt": "Is the percentage of African countries in the United Nations more or less than 65%? Now estimate, as accurately as you can, the actual percentage of African countries in the United Nations. Answer with a single number between 0 and 100."}, {"condition_id": "low", "label": "low anchor (10%)", "prompt": "Is the percentage of African countries in the United Nations more or less than 10%? Now estimate, as accurately as you can, the actual percentage of African countries in the United Nations. Answer with a single number between 0 and 100."}], "anchors": {}, "effect": {"a": "high", "b": "low", "mean_diff": 20.0, "cohen_d": 0.9, "confidence": "approximate", "source": "Jacowitz & Kahneman (1995)", "note": "Absolute estimates are item-specific, so only the contrast is anchored here; the high-low gap is typically 15-25 percentage points."}, "shape": null, "notes": "Deliberately has no level anchor: the point of this protocol is the contrast, not the absolute estimate. This is the pattern to use whenever only an effect size is trustworthy."}
{"protocol_id": "asian_disease", "name": "Asian disease problem — risky-choice framing", "domain": "heuristics_biases", "paradigm": "gain vs. loss frame, logically equivalent options", "source": "Tversky & Kahneman (1981) 'The framing of decisions and the psychology of choice', Science 211:453-458", "response": {"type": "binary", "min": 0, "max": 1, "unit": "1 = choose the sure option, 0 = choose the risky option"}, "conditions": [{"condition_id": "gain", "label": "gain frame (lives saved)", "prompt": "Imagine the US is preparing for an unusual disease expected to kill 600 people. Two programmes are proposed. If programme A is adopted, 200 people will be saved. If programme B is adopted, there is a one-third probability that 600 people will be saved and a two-thirds probability that no one will be saved. Answer 1 for programme A (the sure option), 0 for programme B."}, {"condition_id": "loss", "label": "loss frame (lives lost)", "prompt": "Imagine the US is preparing for an unusual disease expected to kill 600 people. Two programmes are proposed. If programme C is adopted, 400 people will die. If programme D is adopted, there is a one-third probability that nobody will die and a two-thirds probability that 600 people will die. Answer 1 for programme C (the sure option), 0 for programme D."}], "anchors": {"gain": {"mean": 0.72, "sd": null, "confidence": "high", "note": ""}, "loss": {"mean": 0.22, "sd": null, "confidence": "high", "note": ""}}, "effect": {"a": "gain", "b": "loss", "mean_diff": 0.5, "cohen_d": 1.0, "confidence": "high", "source": "Tversky & Kahneman (1981)", "note": "Options A/C and B/D are formally identical; only the description differs."}, "shape": null, "notes": "One of the most replicated effects in the literature. A population that is frame-invariant here is not behaving like a human sample."}
{"protocol_id": "endowment_effect", "name": "Endowment effect — WTA vs. WTP", "domain": "heuristics_biases", "paradigm": "buyer vs. seller valuation of the same object", "source": "Kahneman, Knetsch & Thaler (1990) 'Experimental tests of the endowment effect and the Coase theorem', Journal of Political Economy 98:1325-1348", "response": {"type": "numeric", "min": 0, "max": 10, "unit": "dollars"}, "conditions": [{"condition_id": "wtp", "label": "buyer — maximum willingness to pay", "prompt": "A coffee mug is for sale in the experiment. You do not own one. What is the largest amount, from $0 to $10, that you would be willing to pay to buy this mug? Answer with a single number."}, {"condition_id": "wta", "label": "seller — minimum willingness to accept", "prompt": "You have just been given a coffee mug and now own it. What is the smallest amount, from $0 to $10, that you would be willing to accept to sell this mug? Answer with a single number."}], "anchors": {"wtp": {"mean": 3.0, "sd": 1.5, "confidence": "approximate", "note": ""}, "wta": {"mean": 6.5, "sd": 2.0, "confidence": "approximate", "note": ""}}, "effect": {"a": "wta", "b": "wtp", "mean_diff": 3.5, "cohen_d": 0.8, "confidence": "approximate", "source": "Kahneman, Knetsch & Thaler (1990); Horowitz & McConnell (2002) meta-analysis", "note": "The WTA/WTP ratio is typically 2-3; absolute dollar values depend on the object and the subject pool."}, "shape": null, "notes": "Ownership alone shifts valuation. A population that quotes the same number in both roles is treating this as a pure valuation problem."}
{"protocol_id": "crt_batball", "name": "Cognitive reflection — bat and ball", "domain": "heuristics_biases", "paradigm": "intuitive-but-wrong vs. reflective answer", "source": "Frederick (2005) 'Cognitive reflection and decision making', Journal of Economic Perspectives 19:25-42", "response": {"type": "binary", "min": 0, "max": 1, "unit": "1 = reflective correct answer (5 cents), 0 = intuitive wrong answer (10 cents)"}, "conditions": [{"condition_id": "batball", "label": "bat and ball", "prompt": "A bat and a ball cost $1.10 in total. The bat costs $1.00 more than the ball. How much does the ball cost? Answer 1 if your answer is 5 cents (correct), 0 if your answer is 10 cents."}], "anchors": {"batball": {"mean": 0.4, "sd": null, "confidence": "approximate", "note": "Correct rates on the bat-and-ball item run roughly 30-50% in student samples and lower in general-population samples."}}, "effect": null, "shape": null, "notes": "Scored as a binary here so that no free-text grading is needed. If you want the full 3-item CRT, run it as three binary protocols."}
{"protocol_id": "conjunction_fallacy", "name": "Conjunction fallacy — Linda problem", "domain": "heuristics_biases", "paradigm": "representativeness vs. the conjunction rule", "source": "Tversky & Kahneman (1983) 'Extensional versus intuitive reasoning', Psychological Review 90:293-315", "response": {"type": "binary", "min": 0, "max": 1, "unit": "1 = choose the conjunction, 0 = choose the single (logically correct) option"}, "conditions": [{"condition_id": "between", "label": "between-subjects, ranking removed", "prompt": "Linda is 31 years old, single, outspoken and very bright. She majored in philosophy. As a student she was deeply concerned with issues of discrimination and social justice, and she participated in anti-nuclear demonstrations. Which is more probable? Option 1: Linda is a bank teller. Option 2: Linda is a bank teller and is active in the feminist movement. Answer 1 if you choose option 2 (the conjunction), 0 if you choose option 1."}], "anchors": {"between": {"mean": 0.8, "sd": null, "confidence": "medium", "note": "Roughly 80-90% choose the conjunction in the standard version; rates fall with transparent wording or within-subjects designs."}}, "effect": null, "shape": null, "notes": "Choosing the conjunction violates the conjunction rule. This is the classic case where a well-aligned model looks *worse* than a human sample."}
{"protocol_id": "wason_selection", "name": "Wason selection task — abstract vs. deontic", "domain": "reasoning", "paradigm": "conditional-reasoning / cheater-detection", "source": "Wason (1968); Cosmides & Tooby (1992) 'Cognitive adaptations for social exchange', Cognition 31:187-276", "response": {"type": "binary", "min": 0, "max": 1, "unit": "1 = exactly the two cards that can falsify the rule, 0 = any other selection"}, "conditions": [{"condition_id": "abstract", "label": "abstract rule (D / K / 3 / 7)", "prompt": "Four cards are on a table, showing D, K, 3 and 7. Every card has a letter on one side and a number on the other. The rule is: if a card has a D on one side, then it has a 3 on the other side. You may turn over only the cards needed to test whether the rule is false. The correct answer is the D card and the 7 card. Answer 1 if you would turn over exactly those two cards, 0 otherwise."}, {"condition_id": "deontic", "label": "social-contract rule (under-age drinking)", "prompt": "You are checking whether people are violating the rule 'if a person is drinking beer, then they must be over 21'. Four people are in front of you: one drinking beer, one drinking coke, one aged 25, one aged 16. You may check only the ones needed to find a violation. The correct answer is the beer drinker and the 16-year-old. Answer 1 if you would check exactly those two people, 0 otherwise."}], "anchors": {"abstract": {"mean": 0.15, "sd": null, "confidence": "medium-high", "note": ""}, "deontic": {"mean": 0.7, "sd": null, "confidence": "medium", "note": ""}}, "effect": {"a": "deontic", "b": "abstract", "mean_diff": 0.55, "cohen_d": 1.2, "confidence": "medium-high", "source": "Cosmides & Tooby (1992); Wason (1968)", "note": "The content effect is large: the same logical form is easy when framed as cheater detection and hard when framed abstractly."}, "shape": null, "notes": "The correct answers are stated in the prompts, so this measures whether the population's *selection* matches human selection, not whether it can solve the logic."}
{"protocol_id": "bystander_effect", "name": "Bystander effect — alone vs. in a group", "domain": "social_influence", "paradigm": "diffusion of responsibility", "source": "Darley & Latané (1968) 'Bystander intervention in emergencies', Journal of Personality and Social Psychology 8:377-383; Latané & Nida (1981)", "response": {"type": "binary", "min": 0, "max": 1, "unit": "1 = help, 0 = do not help"}, "conditions": [{"condition_id": "alone", "label": "participant alone", "prompt": "You are alone in a room filling out a questionnaire. You hear someone in the next room fall, cry out in pain and groan. No one else is present. Answer 1 if you would go to help, 0 if you would not."}, {"condition_id": "group", "label": "two passive others present", "prompt": "You are in a room with two other people, filling out a questionnaire. You hear someone in the next room fall, cry out in pain and groan. The other two people look up, then go back to their questionnaires without reacting. Answer 1 if you would go to help, 0 if you would not."}], "anchors": {"alone": {"mean": 0.75, "sd": null, "confidence": "medium", "note": ""}, "group": {"mean": 0.35, "sd": null, "confidence": "medium", "note": ""}}, "effect": {"a": "alone", "b": "group", "mean_diff": 0.4, "cohen_d": 0.85, "confidence": "medium", "source": "Darley & Latané (1968); Latané & Nida (1981)", "note": "Meta-analytic effect is moderate and smaller than textbook accounts suggest, but consistently present."}, "shape": null, "notes": "A simulated population that helps at the same rate in both conditions shows no diffusion of responsibility at all."}
{"protocol_id": "trolley_dilemma", "name": "Trolley problem — switch vs. footbridge", "domain": "moral_judgment", "paradigm": "sacrificial moral dilemmas, means vs. side-effect", "source": "Foot (1967); Thomson (1985) 'The trolley problem', Yale Law Journal 94:1395-1415; large-N replications (e.g. Hauser et al. 2007)", "response": {"type": "binary", "min": 0, "max": 1, "unit": "1 = act to save the five, 0 = do nothing"}, "conditions": [{"condition_id": "switch", "label": "switch (impersonal)", "prompt": "A runaway trolley is heading for five people who will be killed if it continues. You are standing next to a switch. If you pull it, the trolley diverts to a side track where it will kill one person instead. Answer 1 if you pull the switch, 0 if you do not."}, {"condition_id": "footbridge", "label": "footbridge (personal)", "prompt": "A runaway trolley is heading for five people who will be killed if it continues. You are on a footbridge above the track next to a large stranger. The only way to stop the trolley is to push the stranger off the bridge onto the track; his body will stop the trolley, killing him but saving the five. Answer 1 if you push the stranger, 0 if you do not."}], "anchors": {"switch": {"mean": 0.85, "sd": null, "confidence": "medium-high", "note": ""}, "footbridge": {"mean": 0.25, "sd": null, "confidence": "medium-high", "note": ""}}, "effect": {"a": "switch", "b": "footbridge", "mean_diff": 0.6, "cohen_d": 1.3, "confidence": "medium-high", "source": "Thomson (1985); Hauser et al. (2007)", "note": "Both dilemmas sacrifice one to save five; the personal/impersonal contrast drives a very large difference in responding."}, "shape": null, "notes": "The two dilemmas are consequentially identical. A population that gives the same answer to both is doing utilitarian arithmetic, not reproducing human moral judgement."}