A card-choice decision task using the original four-deck payoff schedule of Bechara et al. (1994), the paradigm widely known as the Iowa Gambling Task. Independent implementation; not affiliated with or endorsed by PAR, Inc. or the original authors, and not the commercial Iowa Gambling Task™.
Configure parameters and run an interactive preview exactly as participants will experience it. No data is recorded.
Adjust parameters below, then start the preview on the right.
Task parameters
Euro by default, with the published amounts unchanged, as in the euro version of Mas-Cuesta et al. (2022); the original studies used US dollars.
100 self-paced choices with no time limit: about 12 min. The gain and the loss are each shown for 2 s in published versions (Mata et al., 2023; Wetzels et al., 2010); shorter durations are lab choices.
Fixed, as in Bechara et al. (1994): four decks A–D; every card wins 100 (A, B) or 50 (C, D); losses follow the original 40-card sequence of each deck, which starts again after 40 cards from one deck; start total 2000; 100 choices, which participants are not told (Bechara et al., 1999). Before the next choice: 1–3 s after a card with a loss, 3–5 s after a gain-only card (Mata et al., 2023). Amounts are play money in the chosen currency; euro with the published amounts follows published euro versions (Mas-Cuesta et al., 2022).
No practice: every choice counts. Use it once per participant, e.g. at baseline: it is a learning task, so a second run measures something else. Outputs: net score (C + D) − (A + B) for each block of 20 choices and in total, choices and choice rate per deck, final total and winnings (final total − 2000), deck switches, win-stay and lose-shift rates, median choice time; per choice: deck (A–D and 1–4), card position (1–40) and cycle, gain, loss, net, running total and response time.
This is a researcher preview. No participant data is recorded.
Simulated participant view
A card-choice decision task using the original four-deck payoff schedule of Bechara et al. (1994), the paradigm widely known as the Iowa Gambling Task. Independent implementation; not affiliated with or endorsed by PAR, Inc. or the original authors, and not the commercial Iowa Gambling Task™.
No data is recorded
Four face-down decks, A to D, in one row with the running total above them. The participant taps a deck to take a card; the screen shows the win ("You win +100 €"), on loss cards also the loss in red ("but you also lose −250 €"), then the updated total. Choices are self-paced. There is no trial counter or progress bar, because participants are not told how many cards they will take (100). Amounts are play money, shown in euros, dollars, pounds, Swiss francs or points.
Useful in addiction, impulsivity, gambling, neuropsychological and clinical research where choices that trade immediate reward against long-term loss are a target outcome, and where results should be comparable with the large literature and the many-labs reference data that use the same payoff schedule. Run it once per participant, e.g. at baseline.
Not for repeated or EMA measurement: it is a learning task, and participants who know the decks choose differently on a second run. Not a diagnostic instrument and not the commercial Iowa Gambling Task™ with its norms. Healthy participants vary widely and many prefer deck B (frequent wins, rare large losses), so a low net score alone does not indicate impairment.
Keep the original schedule, the 100 choices, the start total of 2000 and the built-in instructions, which say that some decks are worse than others but not how many choices there are. Gains and losses are shown for 2 s each and the next choice follows after 1–3 s (after a loss) or 3–5 s (after a gain only), as in Mata et al. (2023); shorter feedback is a lab choice. Report the four decks separately as well as the net scores, and use the trial-level data for learning models.
Decks A and B pay 100 per card and C and D 50, but every 10 cards A and B lose 1250 and C and D 250, so A and B lose 250 per 10 cards (disadvantageous) and C and D win 250 (advantageous). A and C give frequent small losses, B and D rare large ones. A net score that rises across the blocks shows learning to prefer the advantageous decks. Losses follow the fixed 40-card sequence of the original task, the same for every participant; after 40 cards from one deck the sequence starts again, as in PEBL and other computerised versions, and the cycle is recorded. The sequence follows PEBL's Bechara Gambling Task (2012 revision) and matches the raw data of Maia & McClelland (2004) and Fridberg et al. (2010). The total may fall below zero. Response times run from the drawn decks to the touch-down.
Add Four-Deck Gambling Task to your ESM study on SMAAT — free to get started.