Do people judge risk by long-run payoff or by how often they lose?
Scoring the same Iowa Gambling Task choices by long-term payoff versus by how often decks punish gave different pictures of risk-taking, and the gap between them depended on how long people waited before retesting and, after a short wait, on sleep loss.
Source
Dual conception of risk in the Iowa Gambling Task: effects of sleep deprivation and test-retest gap
Study at a glance
- Design
- Human experiment — 2 x 2 x 2 mixed design: participants randomly assigned to sleep deprivation or normal sleep and to a 24-hour or 12-week retest gap; IGT choices scored by both the intertemporal and the frequency attribute (within-subject)
- N
- N=40 · 40 male PhD students randomly split across four sleep-by-gap groups (group sizes not stated); each completed 100 IGT trials at baseline and 100 at retest
- Population
- Healthy, non-smoking, right-handed male PhD students in India (mean age 24.9)
- Outcome
- Change from baseline to retest in net IGT score computed by long-term payoff (C'+D' minus A'+B') and by frequency of reinforcement (B'+D' minus A'+C')
Structured fields used in claim comparison tables when every cited study has a complete layer.
What they did
Forty male PhD students played the computerized Iowa Gambling Task, in which four card decks differ both in long-run net payoff and in how often they deliver wins and losses. They were randomly assigned to be retested either 24 hours or 12 weeks later, and either after a night of total sleep deprivation in a supervised dormitory or after a normal night's sleep. The author scored each person's change in choices two ways: preference for decks that pay off over time (the intertemporal attribute) and preference for decks with infrequent losses (the frequency attribute).
What they found
The two scoring methods differed significantly, and the difference depended on the retest gap and on a three-way interaction with sleep deprivation (each explaining roughly 12% of variance). Sleep deprivation had no main effect on its own. After a 12-week gap the two attributes diverged regardless of sleep, whereas after a 24-hour gap the difference between them depended on whether participants had been sleep deprived. Overall, choices based on long-term payoff improved from baseline to retest (net score rising from about 12.5 to 24.4), while frequency-based choices dropped slightly but not significantly.
The limits
What it doesn't show
With only 40 men split across four conditions, the effects are fragile and may not generalize to women or non-student populations. Sleep loss was not verified physiologically, and mood, fatigue and personality were not measured, so the mechanism behind the interaction is speculative. The text describes the short-gap sleep effect inconsistently (once as attenuating, once as enhancing the difference between attributes), so its direction is unclear from the paper alone. No incentives were offered, and the long-gap and short-gap groups also differed in motivation and memory across sessions, which the author acknowledges cannot be equated.
Key terms
- Iowa Gambling Task
- A card game with four decks; two give big wins but net losses over time, two give small wins but net gains, used to study decision making under uncertainty and risk.
- Intertemporal attribute
- Judging decks by their long-run net outcome, so the decks that lose money over many draws count as risky.
- Frequency attribute
- Judging decks by how often they deliver losses, so decks with frequent small punishments feel risky even if they pay off overall.
- Somatic marker hypothesis
- Damasio's idea that bodily emotional signals built up from past outcomes guide people away from bad options, the theory the IGT was built to test.
- Dual-process theory
- The view that decisions come from a fast, automatic, emotion-based system and a slower, reflective, cognition-based system.
- Three-way interaction
- When the effect of one factor depends on the combination of two other factors, here attribute type, sleep condition and retest gap.
Flashcards
0 of 11 answers reviewed
Research intelligence for this paper
See its role on concept claims, tensions it is part of, placement history, and related discoveries.
Quiz yourself
Which pair of IGT decks gives net gains over many draws?
Common questions
Why score the same choices in two different ways?
The IGT decks differ in both long-run payoff and frequency of losses, but the standard score only counts payoff. Many people prefer decks with rare losses even when they lose money overall, so a second score captures a different, possibly more automatic, notion of risk.
Did sleep deprivation make people take more risks?
Not on its own. It had no main effect; it only mattered in combination with the short retest gap and the scoring method.
Why were only men tested?
The author cites sex differences in sleep-deprivation risk behaviour and IGT performance, and female students were reluctant to stay overnight in the testing dormitory.
More on Decision making