Florida-And-Georgia Gambling Task (FLAG)
Background
The Florida-and-Georgia gambling task (FLAG) measures the influence of three cognitive biases (sensitivity to outliers, loss aversion, and primacy-recency effects) on decision-making under uncertainty. Shensheng Wang and colleagues from the University of Florida and Georgia Institute of Technology developed the task (initially introduced in 2024 and peer-published in 2025) specifically to provide an experimental decision-making task that allows the computational modeling of integrating influences of cognitive biases and Prospect Theory's predictions of utility evaluations on people's decision-making under uncertainty.
Building on the design of the popular Iowa Gambling Task, the FLAG cuts down people's cognitive load by having to follow four decks of cards and their various gains and losses to two simple choices: a choice between a deck with uncertain outcomes and one with a sure outcome. Before the choice, participants must uncover several cards (5) from the 'uncertain' deck to get a sense of what might happen on a potential 6th draw. and compare that potential to the offered sure bet.
Wang et al findings from a study of 170 young adults revealed that choices were heavily biased by recent outcomes (recency bias) and that individuals systematically underweighted extreme outlier values for both gains and losses. Interestingly, contrary to standard Prospect Theory assumptions, the data showed no evidence of traditional loss aversion among participants.
Task Procedure
After a short practice phase that runs four trials in fixed order to familiarize participants with the FLAG procedure, participants work on 100 test trials. Each trial starts with the 'sampling phase' during which participants uncover the five cards of the 'uncertain deck' one-by-one. Each card is presented for 1000ms before the next card can be uncovered.
The order in which the cards of the respective decks are uncovered is fixed. Once all five cards have been viewed, the participant gets the choice to uncover the final 6th card for a potential win or take the provided sure offer by pressing a response button that corresponds to the location of the choice. The placement of the sure and gamble decks is randomly decided with half the gamble decks being presented on the left. If participants decide to gamble, the 6th card is uncovered; if the participant decides to take the sure bet, the 6th card is not revealed. Totals are updated according to current wins or losses before the next trial is announced with the cue "New Deck". The 100 test decks are presented in random order for each participant.
The deck value cards are centered around an average score, which differs for each deck and can either be positive or negative. Across the 100 trials, the median of the uncovered cards ranges from -240 to +210. Nineteen decks present a large win sample card (e.g. "+1350") and 19 of the test decks presents a large loss sample card (e.g. "-1250"). Furthermore, across the 100 test decks, 36 present all gain sample cards, 39 present all loss sample cards and 25 present a mixture of gain and loss cards.
What it Measures
The Millisecond Gambling Task (MGT) measures decision-making, risk-seeking, and impulsivity under explicit risk conditions
Psychological domains
- Decision-making: Response to potential rewards and losses over time
- Risk-taking: Preference for high-reward/high risk or low-reward/low-risk
- Executive Control: The ability of our prefrontal cortex to override automatic, reward-seeking behavior to execute an exploratory choice
Main Performance Metrics
- Proportion Gamble Selection: Proportion of times that participant chose the gamble; measure of risk aversion
- Proportion Gamble Selection with Superior and Inferior Sure Offers: Additional measures for loss insensitivity or extreme, rigid risk aversion
- Deliberation Time: Time spent deciding which deck to choose; measure of cognitive processing speed and information gathering
- Switch Rate: Proportion of times participant switched from gamble to sure bet or vice versa => measure of choice volatility
- Proportion Repeat Gambling after Success: Repeat Gambling Choices after Successes; measure of reinforcement learning
Psychiatric Conditions
The FLAG is a novel experimental and computational modeling paradigm that has not yet been formally used in clinical settings.
A decision making framework suggested by Wang et al to study the impact of value integration on gambling under uncertainty.
References
Search Google Scholar for peer-reviewed, published research using the Inquisit Florida-And-Georgia Gambling Task (FLAG).
Wang,S., Wilson,R.C., Ebner,N.C. ,4, and Lighthall,N.R. (preprint). Modeling value integration during decision making under uncertainty: The Florida-and-Georgia (FLAG) gambling task. preprint of article: https://osf.io/preprints/psyarxiv/6gp9n_v1