Prisoner's Dilemma Task with Reciprocity

Technical Manual

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

Created: January 04, 2018

Last Modified: January 29, 2023 by K. Borchert (katjab@millisecond.com), Millisecond

Script Copyright © Millisecond Software, LLC

Background

This script implements a Prisoner Dilemma Game within a reciprocity design. The Prisoner Dilemma Task is a decision making task about cooperation and competition in resource allocation in the absence of knowing what the other party's decision will be.

The Prisoner's Dilemma Task is a non-zero-sum game where the rewards are not a straightforward 'I win what you lose (or vice versa)' but rather a more varied pattern of resource allocation depending on the cooperative vs. competitive choices of both parties. The dilemma is to decide between cooperation and competition (aka self-interest) when competition might yield the highest (immediate) reward but loses to cooperation if both parties decide to compete.

The implemented procedure uses an 'investment framework' in which participants have the opportunity to play the game with the same 'partners' after a practice phase during which the participants are forced to always invest (aka cooperate).

The implemented procedure is based on:

Bell R, Mieth L, Buchner A (2017). Separating conditional and unconditional cooperation in a sequential Prisoner's Dilemma game. PLoS ONE 12(11): e0187952. https://doi. org/10.1371/journal.pone.0187952

Note: by default, multiple partners are used in the script

Duration

25 minutes

Description

Participants play an investment prisoner dilemma game. Each round they are paired with a different partner (introduced via images) and have to decide to invest money into a shared fund or not. For each investment/cooperation decision, the computer adds a certain amount of money to the fund (e.g. if participant decides to cooperate and invests money, the computer adds money. If the partner decides to invest money, the computer adds additional money.). At the end of each round the money in the shared fund gets equally divided between the two people. Thus: a) Partner cooperates-participant does not: participant makes net gain b) Mutual cooperation: participant makes a net gain (as high as the incentive amount provided by computer) c) Mutual non-cooperation: participant neither wins nor loses any money d) Participant cooperates-partner does not: participant makes a net loss

The game is divided into two phases: 1. Exposure phase: During the exposure phase, participants get to know their different partners via their images and 'practice' the game (though they will be forced to always cooperate, their 'partners' will not) 2. Test phase: During the test phase, participants are again paired with the same partners as well as a couple of new ones. This time participants are free to cooperate or compete.

Procedure

Stimuli-Assignment:
- the script automatically randomly selects 6 of each genderxemotion category to be used during the
exposure phase.
- computer balances the number of cooperation and non-cooperation 'decisions' across genders and facial expressions

1. Exposure Phase:
- 24 trials: participant plays the game with 12 female partners (6 look angry, 6 look happy) and 12 male partners
(6 look angry, 6 look happy)
- by default, the computer assigns to each partner image its investment decision (cooperation vs. non-cooperation)
(the number of cooperation and non-cooperation 'decisions' is balanced across genders and facial expressions;
given these constraints, the assignment is randomly done for each participant)
- the order of the images is selected at random for each participant
- by default, participant is forced to cooperate during the first round

2. Test Phase:
- 36 trials: participants play the game with the 24 exposure-phase partners and 12 new ones
(6 females: 3 happy, 3 angry; 6 males: 3 happy and 3 angry)
- the order of the images is selected at random for each participant
- by default, the investment decision of the partners is decided randomly by the computer
- by default, participant is forced to cooperate during the first round

Trial Sequence:
- presentation of current start balance (5000ms)
- presentation of current partner image and the cooperation question (until response)
- decision delay (1000ms)
- participant decision feedback (1000ms)
- partner decision delay (5000ms)
- partner decision feedback (5000ms)
- presentation of gain/loss outcome (5000ms)
- presentation of end balance (5000ms)

Round Duration (default settings): 27s + response time

Stimuli

Bell et al (2017) used the AR database: https://www2.ece.ohio-state.edu/~aleix/ARdatabase.html
This database is free to use for academic research

The stimuli used in this script are images collected from the internet and should be
treated as placeholders. They can be exchanged under section Editable Stimuli.

There are 4 x 9 = 36 stimuli provided.

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).

Summary Data

File Name: prisonerdilemma_reciprocity_summary*.iqdat

Data Fields

NameDescription
inquisit.version Inquisit version number
computer.platform Device platform: win | mac |ios | android
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
computerStrategy The strategy the computer uses to make its choice.
If (parameter) predeterminedExposureDecisions is set to 'true', the selected strategy is overwritten
and the assigned exposureComputerDecision is used instead during the exposure phase.
1 = 'cooperation': selects the cooperative strategy X on every round
2 = 'competition': simply defects on every round of every game by choosing the non-cooperative strategy Y
3 = 'random': selects randomly
4 = 'tit-for-tat': cooperates on first round then mirrors the players choices from the previous rounds
5 = 'tit-for-two-tats': computer cooperates until 2 defects in a row and does not cooperate again until player does
6 = 'probabilistic': cooperates on first round, then selects the strategy that is most often used by player
( if player has selected the cooperative strategy exactly 50% of the time, computer selects cooperative strategy)
predeterminedExposureDecisions True (1) = the computer uses pre-assigned cooperation vs. non-cooperation decisions
that are carefully balanced by gender x facial expression during the exposure phase.
During the test phase, the computer generates the partner's decision based on (parameter) computerStrategy. (default)
false (0) = the computer generates the partner's decision based on (parameter) computerStrategy
during exposure and test phase.
propCoOperation Proportion cooperation selections during the test rounds
propReciprocity Proportion reciprocity during the test rounds
reciprocity = selecting cooperation during the test round for partners who
cooperated during the exposure phase.
propRetaliation Proportion retaliation during the test rounds
retaliation = selecting non-cooperation during the test round for partners who
did not cooperate during the exposure phase.
propReciprocityAngry Proportion reciprocity during the test rounds for 'famkliar' partners with angry expressions
reciprocity = selecting cooperation during the test round for partners who
cooperated during the exposure phase.
propReciprocityHappy Proportion reciprocity during the test rounds for 'old' partners with happy expressions
reciprocity = selecting cooperation during the test round for partners who
cooperated during the exposure phase.
propRetaliationAngry Proportion retaliation during the test rounds for 'old' partners with angry expressions
retaliation = selecting non-cooperation during the test round for partners who
did not cooperate during the exposure phase.
propRetaliationHappy Proportion retaliation during the test rounds for 'old' partners with happy expressions
retaliation = selecting non-cooperation during the test round for partners who
did not cooperate during the exposure phase.
mRT Mean decision latency (in ms) during the test phase
mRTCooperative Mean decision latency (in ms) of cooperative decisions during the test phase
mRTNoncooperative Mean decision latency (in ms) of cooperative decisions during the test phase
mRTReciprocity Mean decision latency (in ms) of reciprocity decisions (see propReciprocity)
mRTRetaliation Mean decision latency (in ms) of retaliation decisions (see propRetaliation)
mRTCooperativeAfterCooperative Mean decision latency (in ms) of cooperative decisions made after the last partner made a non-cooperative decision
mRTCooperativeAfterNoncooperative Mean decision latency (in ms) of cooperative decisions made after the last partner made a cooperative decision
mRTNoncooperativeAfterCooperative Mean decision latency (in ms) of non-cooperative decisions made after the last partner made a non-cooperative decision
mRTNoncooperativeAfterNoncooperative Mean decision latency (in ms) of non-cooperative decisions made after the last partner made a cooperative decision

Raw Data

File Name: prisonerdilemma_reciprocity_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 Name of the current block
blockNum Number of the current block
trialCode Name of the current trial
trialNum Number of the current trial
computerStrategy The strategy the computer uses to make its choice.
If (parameter) predeterminedExposureDecisions is set to 'true', the selected strategy is overwritten
and the assigned exposureComputerDecision is used instead during the exposure phase.
1 = 'cooperation': selects the cooperative strategy X on every round
2 = 'competition': simply defects on every round of every game by choosing the non-cooperative strategy Y
3 = 'random': selects randomly
4 = 'tit-for-tat': cooperates on first round then mirrors the players choices from the previous rounds
5 = 'tit-for-two-tats': computer cooperates until 2 defects in a row and does not cooperate again until player does
6 = 'probabilistic': cooperates on first round, then selects the strategy that is most often used by player
( if player has selected the cooperative strategy exactly 50% of the time, computer selects cooperative strategy)
phase "exposure" vs. "test"
countRounds Tracks the number of round played per phase (resets after practice)
partnerGender "female" vs. "male"
facialExpression "happy" vs. "angry" (facial expression of current partner image)
familiarity "old" (images used during exposure) vs. "new" (images used only during the test phase)
faceStim Stores the image file of the current partner
exposurePartnerDecision 1 = cooperative decision during exposure of this partner
2 = non-cooperative decision during exposure of this partner
"N/A" (for new items)
previousPartnerDecision The partner decision from the previous round: 1 = cooperative; 2 = noncooperative
partnerDecision 1 = cooperative partner decision
2 = non-cooperative partner decision
response Default variable to capture the response recorded during the current trial
decision 1 = cooperative participant decision
2 = non-cooperative participant decision
latency Latency (in ms) of currently recorded trial
rtDecision Latency of decision (in ms)
countConsecutiveNonCooperative Counts the number of consecutive non-cooperative decisions made by participant
startBalance Stores the balance at the start of each round
endBalance Stores the balance at the end of each round
fundMoney Stores the total money in the shared fund
investmentAmount The amount (in cents) participants are asked to invest per round (default: 30cents)

Parameters

The procedure can be adjusted by setting the following parameters.

NameDescriptionDefault
startCapital The amount (in cents) participants receive at the start of the practice and test session
(default: 100 cents)
investmentAmount The amount (in cents) participants are asked to invest per round 30cents
bonusAmount The bonus amount (in cents) added to the shared fund for each cooperation decision
(either from participant or partner) (default: 10cents)
predeterminedExposureDecisions True (1) = the computer uses pre-assigned cooperation vs. non-cooperation decisions
that are carefully balanced by gender x facial expression during the exposure phase.
During the test phase, the computer generates the partner's decision based on parameters.computerStrategy. (default)
false (0) = the computer generates the partner's decision based on parameters.computerStrategy
during exposure and test phase.
computerStrategy The strategy the computer uses to make its choice.
If parameters.predeterminedExposureDecisions is set to 'true', the selected strategy is overwritten
and the assigned values.exposureComputerDecision is used instead during the exposure phase.
Furthermore, some of these strategies may not make sense given the experimental design used.
(e.g. 'tit-for-tat' only makes sense if the same partner is used repeatedly)
1 = 'cooperation': selects the cooperative strategy X on every round
2 = 'competition': simply defects on every round of every game by choosing the non-cooperative strategy Y
3 = 'random': selects randomly
4 = 'tit-for-tat': cooperates on first round then mirrors the players choices from the previous rounds
5 = 'tit-for-two-tats': computer cooperates until 2 defects in a row and does not cooperate again until player does
6 = 'probabilistic': cooperates on first round, then selects the strategy that is most often used by player
( if player has selected the cooperative strategy exactly 50% of the time, computer selects cooperative strategy)
presentBalanceTest True (1) = the balance is presented at start and end of each test trial (default)
false (0) = the balance is not presented at start and end of each test trial
if the default is changed, you might have to change instructions accordingly
presentPartnerDecisionTest True (1) = the participant receives feedback in regard to the partner's decision (same as in practice)
false (0) = no feedback in regard to partner decision is given during test (default)
if the default value is changed, change the instructions accordingly