Millisecond - aka Cambridge Gambling Task

Technical Manual

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

Created: January 06, 2014

Last Modified: January 09, 2025 by K. Borchert (katjab@millisecond.com), Millisecond

Script Copyright © Millisecond Software, LLC

Background

This script implements the Millisecond Gambling Task (MGT), a decision-making and risk-taking behavioral task similar to the Cambridge Gambling Task (CGT).

References

Rogers, R.D., Everitt, B.J., Baldacchino, A., Blackshaw, A.J, Swainson, R., Wynne, K., Baker, N.B., Hunter, J., Carthy, T., Booker, E., M. London, M., Deakin, J.F.W., Sahakian, B.J., & Robbins, T.W. (1999). Dissociable Deficits in the Decision-Making Cognition of Chronic Amphetamine Abusers, Opiate Abusers, Patients with Focal Damage to Prefrontal Cortex, and Tryptophan-Depleted Normal Volunteers: Evidence for Monoaminergic Mechanisms. NEUROPSYCHOPHARMACOLOGY, 20, 322–339.

Duration

20 minutes

Description

Participants are given 10 boxes. Some of these boxes are red, the others are blue. They are told that a yellow token is hidden under one of these boxes and they have to guess the color of the box under which the yellow token is hidden. Once they decide on the color, they are asked to bet points on this choice: the computer provides the bets in either ascending (bets get bigger) or descending order (bets get smaller) and participants are asked to click on a bet when they want to bet this number of points. If they win, the bet number is added to their total points. If they lose the bet number is taken away from their total points.

Procedure

4 ratios of blue/red (as well as red/blue) boxes are tested within participants: 1/9, 2/8, 3/7, 4/6

(a) Practice Block: selection of box color (the color of the box that hides token) only
• 5 trials with a fixed order of blue/red ratios (see LISTS -> list.rednumbers_practice) and fixed order
of token position (see LISTS -> list.tokenbox_practice)

counterbalanced order (by groupnumber) of (b) and (c)
=> odd groupnumbers run Ascending -> Descending
=> even groupnumbers run Descending -> Ascending

(b) Ascending Bet Block: after selection of box color, bets are offered to participants
• by default: 4 blocks with 9 trial sequences aka "displays" each (see Rogers et al, 1999);
participants start with 100 points at the beginning of each block (values.totalpoints- editable value)
( according to Rogers et al, 1999, the trial sequence for one block is prematurely aborted if the
total points go below 2, the total points reset to 100 and the next block is started)

• displays (ratios) are presented in fixed order: order was pseudo-randomly determined by Millisecond.
In this script all 4 ratios are sampled 9 times across all 36 trials; half the time there are more red, half
the time there are more blue boxes -> to change the fixed order go to LISTS -> list.rednumbers_ascending

• fixed order of token position: the order of the token position (box1-10) was pseudo-randomly determined by Millisecond.
In 1/6 of all trials the better choice (based on probability) leads to a loss (see LISTS -> list.tokenbox_ascending)

• 5 bets are provided in fixed order of proportions of current total points (see LISTS -> list.betpercentages_asc).

• bets are starting small and get bigger (each bet is presented for parameters.betduration, default 5s);
sounds accompany each bet (the higher the sound, the higher the bet)

• if no bet is selected the last bet is selected by default

(b) Descending Bet Block:
• bets are starting big and get smaller
• different fixed orders for ratios/token position
=> no other differences to Ascending Bet Blocks

Instructions

Instructions are not original to the task. They are provided by Millisecond
as htm/html pages and can be edited by changing the provided htm/html files.
To edit htm/html-files: open the respective documents in simple Text Editors such as TextEdit (Mac)
or Notepad (Windows).

Scoring

Risk Adjustment:
Participants typically gamble more if the odds are in their favor (for ratios 1:9 and 2:8)
than when they are not (ratios 3:7 and 4:6)

The Risk Adjustment measure is calculated by
(2 * meanRiskPercent_19 + 1 * meanRiskPercent_28) - (2 * meanRiskPercent_46 + 1 * meanRiskPercent37)
in 'bestChoice' selection trials only

Delay Aversion:
Unwillingness to wait is calculated by the difference between meanRiskPercent in descending and ascending order
The assumption is that people who don't like to wait bet at higher percentage in descending order (because
they come up faster) than in the ascending order (where they have to wait longer)

calculation: delayAversion = meanRiskPercentBestChoiceOnlyDesc - meanRiskPercentBestChoiceOnlyAsc

Summary Data

File Name: millisecondgamblingtask_summary*.iqdat

Data Fields

NameDescription
inquisit.version Inquisit version number
computer.platform Device platform: win | mac |ios | android
computer.touch 0 = device has no touchscreen capabilities; 1 = device has touchscreen capabilities
computer.hasKeyboard 0 = no external keyboard detected; 1 = external keyboard detected
startDate Date the session was run
startTime Time the session was run
subjectId Participant ID
groupId Group number
sessionId Session number
elapsedTime Session duration in ms
completed 0 = Test was not completed
1 = Test was completed
conditionOrder 1 = ascending-descending
2 = descending-ascending
countTrialsAscending Number of trials run in ascending condition
countTrialsDescending Number of trials run in descending conditions
totalPointsAscending Final totalpoints in ascending condition
totalPointsDescending Final totalpoints in descending condition
Best Choices Made: Quality Of Decision Making By Betorder And Ratio
propBestChoices19Asc Proportion of bestChoice (higher probability color) selection for trials with "1:9" ratio in ascending bet trials
propBestChoices28Asc Proportion of bestChoice (higher probability color) selection for trials with "2:8" ratio in ascending bet trials
propBestChoices37Asc Proportion of bestChoice (higher probability color) selection for trials with "3:7" ratio in ascending bet trials
propBestChoices46Asc Proportion of bestChoice (higher probability color) selection for trials with "4:6" ratio in ascending bet trials
propBestChoicesAsc Proportion of bestChoice (higher probability color) selection in ascending bet trials (across ratios)
propBestChoices19Desc Proportion of bestChoice (higher probability color) selection for trials with "1:9" ratio in descending bet trials
propBestChoices28Desc Proportion of bestChoice (higher probability color) selection for trials with "2:8" ratio in descending bet trials
propBestChoices37Desc Proportion of bestChoice (higher probability color) selection for trials with "3:7" ratio in descending bet trials
propBestChoices46Desc Proportion of bestChoice (higher probability color) selection for trials with "4:6" ratio in descending bet trials
propBestChoicesDesc Proportion of bestChoice (higher probability color) selection in descending bet trials (across ratios)
propBestChoices19 Proportion of bestChoice (higher probability color) selection for trials with "1:9" ratio (across ascending and descending bet order)
propBestChoices28 Proportion of bestChoice (higher probability color) selection for trials with "2:8" ratio (across ascending and descending bet order)
propBestChoices37 Proportion of bestChoice (higher probability color) selection for trials with "3:7" ratio (across ascending and descending bet order)
propBestChoices46 Proportion of bestChoice (higher probability color) selection for trials with "4:6" ratio (across ascending and descending bet order)
propBestChoices Proportion of bestChoice (higher probability color) selection in bet trials (across ratios and across ascending and descending bet order)
Deliberationrt By Betorder And Ratio
meanDeliberationRT19Asc Mean DeliberationRT (in ms) of bestChoice (higher probability color) selection for trials with "1:9" ratio in ascending bet trials
meanDeliberationRT28Asc Mean DeliberationRT (in ms) of bestChoice (higher probability color) selection for trials with "2:8" ratio in ascending bet trials
meanDeliberationRT37Asc Mean DeliberationRT (in ms) of bestChoice (higher probability color) selection for trials with "3:7" ratio in ascending bet trials
meanDeliberationRT46Asc Mean DeliberationRT (in ms) of bestChoice (higher probability color) selection for trials with "4:6" ratio in ascending bet trials
meanDeliberationRTAsc Mean DeliberationRT (in ms) of bestChoice (higher probability color) selection in ascending bet trials (across ratios)
meanDeliberationRT19Desc Mean DeliberationRT (in ms) of bestChoice (higher probability color) selection for trials with "1:9" ratio in descending bet trials
meanDeliberationRT28Desc Mean DeliberationRT (in ms) of bestChoice (higher probability color) selection for trials with "2:8" ratio in descending bet trials
meanDeliberationRT37Desc Mean DeliberationRT (in ms) of bestChoice (higher probability color) selection for trials with "3:7" ratio in descending bet trials
meanDeliberationRT46Desc Mean DeliberationRT (in ms) of bestChoice (higher probability color) selection for trials with "4:6" ratio in descending bet trials
meanDeliberationRTDesc Mean DeliberationRT (in ms) of bestChoice (higher probability color) selection in descending bet trials (across ratios)
meanDeliberationRT19 Mean DeliberationRT (in ms) of bestChoice (higher probability color) selection for trials with "1:9" ratio (across ascending and descending bet order)
meanDeliberationRT28 Mean DeliberationRT (in ms) of bestChoice (higher probability color) selection for trials with "2:8" ratio (across ascending and descending bet order)
meanDeliberationRT37 Mean DeliberationRT (in ms) of bestChoice (higher probability color) selection for trials with "3:7" ratio (across ascending and descending bet order)
meanDeliberationRT46 Mean DeliberationRT (in ms) of bestChoice (higher probability color) selection for trials with "4:6" ratio (across ascending and descending bet order)
meanDeliberationRT Mean DeliberationRT (in ms) of bestChoice (higher probability color) selection in bet trials (across ratios and across ascending and descending bet order)
Risk Taking: By Betorder And Ratio - Riskpercentage (Only For Trials In Which The Best Choice Was Selected)
meanRiskPercentBestChoiceOnly19Asc Mean risk percentage* in bestChoice (higher probability color) selection trials for trials with "1:9" ratio in ascending bet trials
mean risk percentage = risk percentage in bet box at time of bet placing
meanRiskPercentBestChoiceOnly28Asc Mean risk percentage in bestChoice (higher probability color) selection trials for trials with "2:8" ratio in ascending bet trials
meanRiskPercentBestChoiceOnly37Asc Mean risk percentage in bestChoice (higher probability color) selection trials for trials with "3:7" ratio in ascending bet trials
meanRiskPercentBestChoiceOnly46Asc Mean risk percentage in bestChoice (higher probability color) selection trials for trials with "4:6" ratio in ascending bet trials
meanRiskPercentBestChoiceOnlyAsc Mean risk percentage in bestChoice (higher probability color) selection trials in ascending bet trials (across ratios)
meanRiskPercentBestChoiceOnly19Desc Mean risk percentage in bestChoice (higher probability color) selection trials for trials with "1:9" ratio in descending bet trials
meanRiskPercentBestChoiceOnly28Desc Mean risk percentage in bestChoice (higher probability color) selection trials for trials with "2:8" ratio in descending bet trials
meanRiskPercentBestChoiceOnly37Desc Mean risk percentage in bestChoice (higher probability color) selection trials for trials with "3:7" ratio in descending bet trials
meanRiskPercentBestChoiceOnly46Desc Mean risk percentage in bestChoice (higher probability color) selection trials for trials with "4:6" ratio in descending bet trials
meanRiskPercentBestChoiceOnlyDesc Mean risk percentage in bestChoice (higher probability color) selection trials in descending bet trials (across ratios)
meanRiskPercentBestChoiceOnly19 Mean risk percentage in bestChoice (higher probability color) selection trials for trials with "1:9" ratio (across ascending and descending bet order)
meanRiskPercentBestChoiceOnly28 Mean risk percentage in bestChoice (higher probability color) selection trials for trials with "2:8" ratio (across ascending and descending bet order)
meanRiskPercentBestChoiceOnly37 Mean risk percentage in bestChoice (higher probability color) selection trials for trials with "3:7" ratio (across ascending and descending bet order)
meanRiskPercentBestChoiceOnly46 Mean risk percentage in bestChoice (higher probability color) selection trials for trials with "4:6" ratio (across ascending and descending bet order)
meanRiskPercentBestChoiceOnly Mean risk percentage in bestChoice (higher probability color) selection trials in bet trials (across ratios and across ascending and descending bet order)
Riskpercentages By Betorder And Ratio (All Selections = Across Bestchoice And Worstchoice Selections)
meanRiskPercentAll19Asc Mean risk percentage* all selection trials for trials with "1:9" ratio in ascending bet trials
mean risk percentage = risk percentage in bet box at time of bet placing
meanRiskPercentAll28Asc Mean risk percentage all selection trials for trials with "2:8" ratio in ascending bet trials
meanRiskPercentAll37Asc Mean risk percentage all selection trials for trials with "3:7" ratio in ascending bet trials
meanRiskPercentAll46Asc Mean risk percentage all selection trials for trials with "4:6" ratio in ascending bet trials
meanRiskPercentAllAsc Mean risk percentage all selection trials in ascending bet trials (across ratios)
meanRiskPercentAll19Desc Mean risk percentage all selection trials for trials with "1:9" ratio in descending bet trials
meanRiskPercentAll28Desc Mean risk percentage all selection trials for trials with "2:8" ratio in descending bet trials
meanRiskPercentAll37Desc Mean risk percentage all selection trials for trials with "3:7" ratio in descending bet trials
meanRiskPercentAll46Desc Mean risk percentage all selection trials for trials with "4:6" ratio in descending bet trials
meanRiskPercentAllDesc Mean risk percentage all selection trials in descending bet trials (across ratios)
meanRiskPercentAll19 Mean risk percentage all selection trials for trials with "1:9" ratio (across ascending and descending bet order)
meanRiskPercentAll28 Mean risk percentage all selection trials for trials with "2:8" ratio (across ascending and descending bet order)
meanRiskPercentAll37 Mean risk percentage all selection trials for trials with "3:7" ratio (across ascending and descending bet order)
meanRiskPercentAll46 Mean risk percentage all selection trials for trials with "4:6" ratio (across ascending and descending bet order)
meanRiskPercentAll Mean risk percentage all selection trials in bet trials (across ratios and across ascending and descending bet order)
Risk Adjustment
riskAdjustAsc Risk Adjustment for ascending trials
riskAdjustDesc Risk Adjustment for descending trials
riskAdjust Risk Adjustment across ascending and descending trials
Delay Aversion
delayAversion19 Delay Aversion in "1:9" ratio trials
delayAversion28 Delay Aversion in "2:8" ratio trials
delayAversion37 Delay Aversion in "3:7" ratio trials
delayAversion46 Delay Aversion in "4:6" ratio trials
delayAversion Delay Aversion across the four ratio trials

Raw Data

File Name: millisecondgamblingtask_raw*.iqdat

Data Fields

NameDescription
build Inquisit version number
computer.platform Device platform: win | mac |ios | android
computer.touch 0 = device has no touchscreen capabilities; 1 = device has touchscreen capabilities
computer.hasKeyboard 0 = no external keyboard detected; 1 = external keyboard detected
date Date the session was run
time Time the session was run
subject Participant ID
group Group number
session Session number
blockcode The name the current block (built-in Inquisit variable)
blocknum The number of the current block (built-in Inquisit variable)
trialcode The name of the currently recorded trial (built-in Inquisit variable)
trialnum The 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.
conditionOrder 1 = ascending-descending
2 = descending-ascending
countSequence Counts sequences per condition
countTrialsSequence Counts trials in a sequence
countTrialsCondition Counts trials per conditions
countTrialsTotal Counts trial across conditions
ratio Stores the current ratio
redNumber Stores the number of the current red boxes
blueNumber Stores the number of the current blue boxes
bestChoice Stores the best choice (either 'red' or 'blue' depending on which color has more boxes)
tokenBox Stores the number of the token box (1-10)
betOrder 1 = ascending; 2 = descending
rtSelectColor Latency in ms to select the color (measured from time the interface appears on screen until the color box is clicked)
selectedColor Stores the selected color (red vs. blue)
selectBestChoice 1 = participant selected best choice (= more probable to win); 2 = participant did not select best choice
pReward Probability of reward for picking this color given the number of the boxes with that color
(Example: there are 3/10 red boxes and participant picked red -> p_reward = 0.3)
rtSelectbet The time it cumulatively took to pick a bet (measured from start of displaying first bet until bet was selected or time for betting was up)
percentBet Percentage of totalpoint that is picked to display current bet
bet Currently displayed bet (calculated by values.percent_bet x values.totalpoints)
selectbet 1 = participant did select a bet; 0 = participant did NOT select a bet (and therefore last bet was taken as default)
result 1 = participant wins; 2 = participant loses
totalPoints The number of totalpoints at the end of the current trialsequence

Parameters

The procedure can be adjusted by setting the following parameters.

NameDescriptionDefault
Design
startAmount Amount of money allocated at the start of each sequence100
minPoints Participant needs to have more than parameters.min_points for a sequence to continue1
roundsPerCondition Number of rounds per condition4
displaysperBlock Number of displays per round9
Timing Parameters
betDuration Duration of each bet in ms5000
resultDuration Duration of result display in ms2000
waitforresult Duration that participant needs to wait until result is revealed in ms1000