Florida-And-Georgia Gambling Task - FLAG

Technical Manual

Script Author: Katja Borchert, Ph.D. (katjab@millisecond.com), Millisecond

Created: January 14, 2025

Last Modified: January 16, 2025 by K. Borchert (katjab@millisecond.com), Millisecond

Script Copyright © Millisecond Software, LLC

Background

This script implements the Florida-And-Georgia Gambling Task (FLAG) proposed by Wang et al (preprint). The FLAG provides a framework to study the impact of value integration on gambling under uncertainty.

References

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

Original Game Code (Javascript) Available Here:
https://github.com/swan232/FLAG.

Duration

25 minutes

Description

Participants play a card game for points. They are given decks of cards with variable point distributions (positive and negative). Five cards are revealed to them ('sampling phase') before they have to make a choice ('choice phase'): uncover a sixth card from the deck and win (or lose) the points of this uncovered card or take a sure offer instead.

Procedure

(1) Intro

(2) Practice: 4 practice decks with additional instructional pointers
- practice trials are all self-paced; timeouts are suspended
- the four practice decks are presented in fixed order (see list.trialInfoPractice; contains deck information from original javaScript program)

(3) Test: 100 test decks
- decks are presented in randomized order (see list.trialInfoTest; contains deck information from original javaScript program)

Trial Sequence:
- 'new deck' announcement: 1000ms

Sampling Phase: Repeats For 5 Cards ( Sampling Of Cards Is Done In Fixed Order For Each Deck)
- sampling start card (until response)
- card revelation (1000ms)

Choice Phase:
- choice: gamble deck vs. sure offer (until response) ( half the time the gamble deck is located on the left)
- feedback and score update: 1000ms

Deck Information
• information for the 100 decks was taken from the original javaScript program (https://github.com/swan232/FLAG)
(see list.trialInfoTest)
• each deck is assigned:
- a fixed sequence of cards for the sampling phase ( the last value is the amount that can be won from turning over the 6th card)
- the location of the gamble deck during the choice phase (0 = right; 1 = left); balanced
- the sure amount that is offered during the choice phase

from Wang et al (preprint)

- "Deck outcomes centered around an average score, which differed across games and
could either be positive or negative (e.g., +250 to -250 points)."
- "FLAG task symmetrically varies both the magnitude and frequency of gains in addition to losses,
resulting in an equal number (19) of trials with either large gains (e.g., 1310 points) or losses (e.g., -1300 points).
Of the one hundred games, thirty-six included all gain trials, thirty-nine all loss trials,
and twenty-five a mixture of both gains and losses."

Stimuli

deck cards provided by Millisecond - can be edited under section 'Editable Stimuli'

Instructions

All instructions are modelled after the ones provided in the original javaScript program
(https://github.com/swan232/FLAG)

They can be edited in script flag_instructions_inc.iqjs

Summary Data

File Name: xxx_summary*.iqdat

Data Fields

NameDescription
inquisit.version Inquisit version number
computer.platform Device platform: win | mac |ios | android
startdate Date script was run
starttime Time script was started
subjectid Assigned subject id number
groupid Assigned group id number
sessionid Assigned session id number
elapsedtime Time it took to run script (in ms); measured from onset to offset of script
completed 0 = Test was not completed
1 = Test was completed
decksPlayed Number of decks played (during test)
propGamble Proportion of gamble chosen (during test) (as opposed to sure amount)

Raw Data

File Name: XXX_raw*.iqdat

Data Fields

NameDescription
build Inquisit version number
computer.platform Device platform: win | mac |ios | android
date Date the session was run
time Time the session was run
subject Participant ID
group Group number
session Session number
blockcode, blocknum The name and number of the current block (built-in Inquisit variable)
trialcode, trialnum The name and number of the currently recorded trial (built-in Inquisit variable)
trialnum is a built-in Inquisit variable; it counts all trials run; even those
that do not store data to the data file.
phase "Practice" vs. "Test"
deckCounter Tracks the number of decks played (resets after practice)
Deck Information
index The current index of the deck (Range: 0-99) (see list.trialInfoTest/list.trialInfoPractice)
outcomes The current 6 deck cards (presented in order)
side 0 = gamble deck is presented on the right
1 = gamble deck is presented on the left
offer The current sure offer
outcome The last revealed deck card ( for trial.choice 'outcome' contains the value of the 6th card; not known to participant)
Dvs
respCategory "gamble" (participant chose deck during trial.choiceTest)
"offer" (participant chose sure offer during )
response The response of participant (scancode of response button)
responseText Scancode translated into actual key
latency Response latency (in ms); measured from: onset of trial
Score Information
score The current score
win Stores the 'winning' points of the current choice trial
Card Locations
picture.gambleCard.x The horizontal screen position (in %canvas width) of the gamble deck card during trial.choiceTest
picture.offerCard.x The horizontal screen position (in %canvas width) of the sure offer 'card' during trial.choiceTest
25% = left; 50% = center; 75% = right

Parameters

The procedure can be adjusted by setting the following parameters.

NameDescriptionDefault
Design
numberPracticeTrials The number of practice trials to run
Max: 4
the original javaScript program plays with 4
4
numberTestTrials The number of test trials to run
Max: 100
the original javaScript program plays with 100
100
startCapital Number of points at start of test1000
Sizing Parameters
cardHeightPct The height of the card image in %canvas height50%
offerFontSizePct The fontsize of the reported outcomes in %canvas height10%
Timing Parameters
deckStartMS Duration (in ms) of the 'new Deck' trial during the test phase1000
samplingMS Duration (in ms) that each drawn card is presented during test phase1000
feedbackDurationMS Duration (in ms) of the feedback trial during the test phase1000
Responsekeys
leftResponseKey "F"
rightResponseKey "J"