Codebook

Definition
Whether a study was published or not.

Levels of the variable

• Published article: Study was reported in a published document
• Doctoral Dissertation: Study was reported in a doctoral dissertation
• Working paper: Study was reported in a working paper
• Master’s thesis: Study was reported in a master’s thesis
• Raw data: Study was coded from raw data

Inter-coder agreement

• Krippendorff’s α: 0.66
• Most frequent level (%): Published (97.1%)
• Agreement %: 98.5%

Definition
Participant’s field of study if the participant is a student.

Levels of the variable

• Other
• Economics
• Psychology
• Mixed
• Sociology

Inter-coder agreement

• Krippendorff’s α: 0.70
• Most frequent level (%): 999 (49.5%)
• Agreement %: 79,1%

Definition
Whether participants were interacting (and were aware of it) with people that they are not acquainted with or with people that they knew outside of the laboratory (e.g., friends, relatives, parents, romantic partners).

Levels of the variable

• Strangers: Their partner was a stranger
• Possible Acquaintance: Their partner was neither a complete stranger or an acquaintance. For example, participants were playing as part of a classroom exercise, so even if they were matched with strangers, there was a chance they know each other and they were not playing with total strangers in strict sense. Also, when participants were recruited from a small-scale society or were members of a small community, so again there was a chance they might know each other or make use of the reputational information of that community
• Acquaintance: Their partner has a relationship with the participant (e.g., acquaintance, spouse, friend, relative).

Inter-coder agreement
999

Definition
Country or region where the data collection took place (coded with the 3-letter country code following ISO 3166-1 alpha-3). This variable can overlap with participant’s nationality.

Levels of the variable
e.g., USA, NLD, CHN, JPN, etc.

Inter-coder agreement

• Krippendorff’s α: 0.92
• Most frequent level (%): USA (36.9%)
• Agreement %: 93.7%

Definition
Source of the country or region where the data collection took place.

Levels of the variable

• Specified country: The paper explicitly states that data have been collected at a certain location (e.g., University, lab, and/or participants pool)
• Most authors: The majority of the authors are affiliated to the same institution in a specific country or region
• All authors: The paper does not specify the country or region, but all authors are affiliated to the same institution in a specific country (indicate it for single-authored papers)
• Multiple countries: The paper specifies that most participants come from a specific country or region, but that a smaller percentage come from different countries (we indicate the percentages and nationalities in a note, if the information is available)

Inter-coder agreement

• Krippendorff’s α: 0.65
• Most frequent level (%): Specified country (84.1%)
• Agreement %: 90.3%

Definition
Maximum age of all sampled participants.

Inter-coder agreement

• Krippendorff’s α: 1.00

Definition
Minimum age of all sampled participants.

Inter-coder agreement

• Krippendorff’s α: 1.00

Definition
Participants’ mean age in years (after exclusion of participants, when that information was reported).

Inter-coder agreement

• Krippendorff’s α: 1.00

Definition
Proportion of male participants in the sample of a study (after exclusion of participants, when that information was reported).

Inter-coder agreement

• Krippendorff’s α: 0.92

Definition
The way participants were recruited to take part in the study.

Levels of the variable

• Participant pool: Participants were recruited from a participant pool which includes participants enrolled in a class or at the university, such as through ORSEE or SONA.
• MTurk: Mechanical Turk (MTurk)
• Other

Inter-coder agreement

• Krippendorff’s α: 0.76
• Most frequent level (%): Participant pool (49.6%)
• Agreement %: 85.5%

Definition
Whether participants were recruited from a student population.

Levels of the variable

• TRUE = Participants were recruited from a student population
• FALSE = Participants were not recruited from a student population

Inter-coder agreement

• Krippendorff’s α: 0.89
• Most frequent level (%): TRUE (85.3%)
• Agreement %: 97.3%

Definition
Total sample size in a single study after exclusion of participants.

Inter-coder agreement

• Krippendorff’s α: 0.93

Definition
Year when the data was collected. If the information was not reported in the paper, the variable indicates the closest year to data collection.

Inter-coder agreement

• Krippendorff’s α: 0.98

Definition
Source of information about what year the data was collected.

Levels of the variable

• Received/Submitted: When the paper was received/Submitted to a journal
• Conducted: When the study was conducted
• Published: When the paper was published
• Accepted: When the paper was accepted for publication
• Presented: When the study was presented at a meeting
• Working paper published: When a working paper was published
• Available online: When the data was available online

Inter-coder agreement

• Krippendorff’s α: 0.78
• Most frequent level (%): Received/Submitted (55.8%)
• Agreement %: 85.9%

Definition
Whether there was a provision point for contributions to establish a public good (i.e., step-level public goods) versus each contribution provided an incremental benefit to the public good (i.e., continuous public good).

Levels of the variable

• Continuous: In the continuous public goods game, subjects are provided an endowment and decide any portion to allocate to the public good, and the remaining about is kept for themselves. The group contributions in the public pool are multiplied by a factor (greater than one and less than the number of players, N) and it is divided among players, regardless of the players contributions.
• Step-level: In the step level public goods game, subjects are provided an endowment and decide any portion of the endowment to allocate to the public good, and the remaining amount is kept for themselves. The group contributions to the public pool are multiplied by a factor and it is divided among players (regardless of their contributions) only if a certain threshold (provision point) is met.

Inter-coder agreement

• Krippendorff’s α: 0.78
• Most frequent level (%): N/A (53.4%)
• Agreement %: 87.9%

Definition
Whether deception was applied in the study.

Levels of the variable

• TRUE = deception was applied
• FALSE = deception was not applied

Inter-coder agreement

• Krippendorff’s α: 0.74
• Most frequent level (%): FALSE (71.2%)
• Agreement %: 89.1%

Definition
Whether communication was allowed between participants in the game.

Levels of the variable

• Uni-directional: Participants were allowed to send messages or communicate with other participants, but those participants were not allowed to respond to that communication (i.e., uni-directional communication).
• Absent: Communication was not allowed.
• Bi-directional: Participants were allowed to send messages or communicate with other participants, and those participants could respond to that communication (i.e., bi-directional communication).

Inter-coder agreement

• Krippendorff’s α: 0.68
• Most frequent level (%): Absent (93.2%)
• Agreement %: 96.6%

Definition
The setting in which the experiment was conducted.

Levels of the variable

• Lab: The experiment was conducted in a laboratory (Classrooms used as laboratory setting are considered as lab).
• Class: The experiment was conducted during a regular class (e.g., during some demonstration of a negotiation task for teaching purposes).
• Online: The experiment was conducted online on a web platform (e.g., MTurk).
• Lab in the field: The experiment had all the characteristics of a lab experiment but it is carried out at a field site.
• Field: The experiment involved a manipulation carried out in the field (Subjects may or may not be aware that they are part of an experiment).
• Natural experiment: A quasi-experiment in the field in which randomization is not controlled by the experimenter (e.g., a study using a TV game show).
• Other: Other experimental setting not included in the defined categories.

Inter-coder agreement

• Krippendorff’s α: 0.75
• Most frequent level (%): Lab (83.5%)
• Agreement %: 93.5%

Definition
Whether participants’ decisions in the game determine their payoffs in different forms.

Levels of the variable

• Hypothetical: Decisions in the game resulted in outcomes with hypothetical value for participants (e.g., points).
• Monetary: Decisions in the game resulted in monetary outcomes, often converted from game points.
• Non-monetary: Decisions in the game resulted in outcomes involving some amount of non-monetary material resource (e.g., candies, school supplies).
• Monetary lottery: Decisions in the game resulted in monetary outcomes gained through a lottery, often converted from game points.
• Non-monetary lottery: Decisions in the game resulted in outcomes involving some amount of non-monetary material resource (e.g., candies, school supplies) gained through a lottery.

Inter-coder agreement

• Krippendorff’s α: 0.77
• Most frequent level (%): Monetary (76.1%)
• Agreement %: 92.1%

Definition
Economic game that participants play during the experimental session.

Levels of the variable

• Public Goods Game: In the continuous public goods game, subjects are provided an endowment and decide any portion to allocate to the public good, and the remaining amonut is kept for themselves. The group contributions in the public pool are multiplied by a factor (greater than one and less than the number of players, N) and it is divided among players, regardless of the players contributions. In the step level public goods game, subjects are provided an endowment and decide any portion of the endowment to allocate to the public good, and the remaining amonut is kept for themselves. The group contributions to the public pool are multiplied by a factor and it is divided among players (regardless of their contributions) only if a certain threshold (provision point) is met.
• Other: Other game types not included in the defined categories (e.g., Intergroup games).
• Resource Dilemma: In the standard resource dilemma game, players harvest resources from a common resource pool of known size, and after each trial the pool is replenished at a predetermined rate and is exhausted when the withdrawals exceed the resource in the pool.
• Prisoner’s Dilemma Game: In the standard prisoner’s dilemma game, the relative value of the four outcomes reflects the following relationships: DC > CC > DD > CD.

Inter-coder agreement

• Krippendorff’s α: 0.75
• Most frequent level (%): Public Goods Game (43.7%)
• Agreement %: 82.9%

Definition
If none of the game type categories apply, then we report the name of the game included in the study.

Definition
Overall number of people affected by the choices in the game.

Inter-coder agreement

• Krippendorff’s α: 0.97

Definition
Highest choice option allowed to participants. The value of the highest choice option. Is ‘1’ in a binary choice between non-numeric options (such as ‘cooperate’ vs. ‘defect’; ‘C’ vs. ‘D’).

Inter-coder agreement

• Krippendorff’s α: 0.92

Definition

For 2-persons PDG, it is calculated as (R-P)/(T-S), and 0 < K < 1. For N-persons PDG, it is calculated as (A(n)-B(0))/(0(max)-0(min)), and 0 < K < 1.

Inter-coder agreement

• Krippendorff’s α: 0.91

Definition

Whether participants know the exact number of trials at the beginning of the experiment. Endgame is considered known also when participants played a one-shot game.

Levels of the variable

• TRUE = Known endgame
• FALSE = Endgame not known

Inter-coder agreement

• Krippendorff’s α: 0.78
• Most frequent level (%): TRUE (73.4%)
• Agreement %: 90.7%

Definition
Lowest choice option allowed to participants. The value of the lowest choice option. Is ‘0’ in a binary choice between non-numeric options (such as ‘cooperate’ vs. ‘defect’; ‘C’ vs. ‘D’).

Inter-coder agreement

• Krippendorff’s α: 0.71

Definition
How participants are paired with others during interactions.

Levels of the variable

• Stranger: Participants interact with one person for one trial (i.e., one-shot), or switch partners after each trial across many trials (i.e., one-shot repeated).
• Partner: Participants interact with the same partner for multiple trials.

Inter-coder agreement

• Krippendorff’s α: 0.84
• Most frequent level (%): Stranger (49.8%)
• Agreement %: 91.6%

Definition
The ratio of benefits to costs for a member to contribute one monetary unit to the group account. Calculated as multiplier divided by group size.

Inter-coder agreement

• Krippendorff’s α: 0.95

Definition
The number of choice options players have when making the contribution. Is ‘2’ when making a binary choice (such as cooperate vs. defect); (n+1) for continuous measure of cooperation; 3 for 3-choice Prisoner’s Dilemma, etc.

Inter-coder agreement

• Krippendorff’s α: 0.96

Definition
Whether participants played the game with the same person only once (this also applies if participants switch partners after each trial) versus played the game repeatedly with the same person.

Levels of the variable

• Repeated: Participants played the game repeatedly with the same person.
• One-shot: Participants played the game with the same person only once.

Inter-coder agreement

• Krippendorff’s α: 0.81
• Most frequent level (%): FALSE (52.2%)
• Agreement %: 90.2%

Definition
Whether participants are paired with different partner(s) after each trial across many trials.

Levels of the variable

• TRUE = repeated one-shot
• FALSE = one-shot

Inter-coder agreement

• Krippendorff’s α: 0.69
• Most frequent level (%): N/A (50.5%)
• Agreement %: 80.5%

Definition
Whether participants interact with real or imagined participants in the game.

Levels of the variable

• Real: Participants involved in real-time interactions with real participant(s).
• Hypothetical: Participants imagined interactions with others or played with a computer.
• Deception: Participants believe that they are interacting with real participants, but experimenter strategy was used.

Inter-coder agreement

• Krippendorff’s α: 0.77
• Most frequent level (%): Real (70.2%)
• Agreement %: 89.2%

Definition
The replenishment rate in the resource dilemma. The remaining resource in the common pool is multiplied by the replenishment rate after each trial.

Inter-coder agreement

• Krippendorff’s α: 1

Definition
Whether a sanction (punishment or reward) mechanism was in place in the game.

Levels of the variable

• TRUE = sanction was present
• FALSE = sanction was absent

Definition
Whether group members make their decisions simultaneously or sequentially, i.e., whether participants take decisions one after another while receiving some form of feedback about preceding decisions.

Levels of the variable

• Simultaneous: Participants make decisions simultaneously.
• Sequential turn-taking: Participants make decisions one after another following some (endogenous or exogenous) order.

Inter-coder agreement

• Krippendorff’s α: 0.66
• Most frequent level (%): Simultaneous (89.3%)
• Agreement %: 93.6%

Definition
Whether participants received a show-up fee for the study (i.e., a certain payment for participation).

Levels of the variable

• Paid: Participants received a monetary payment as show-up fee.
• Course credit: Participants received course credit as show-up fee.
• Non-monetary: Participants received a non-monetary payment as show-up fee.
• Absent: Participants did not receive a show-up fee.

Inter-coder agreement

• Krippendorff’s α: 0.73
• Most frequent level (%): Paid (63.2%)
• Agreement %: 84.9%

Definition
Whether specific aspects of the game that were different (i.e., asymmetric) for participants.

Levels of the variable

• TRUE = Symmetric
• FALSE = Asymmetric

Inter-coder agreement

• Krippendorff’s α: 0.59
• Most frequent level (%): TRUE (93.1%)
• Agreement %: 94.6%

Defenition
The minimum threshold of total contribution by all group members when the public good will be provided. This also defines the criticality of contributions, as contributions are more critical if other players cannot reach the threshold on their own (e.g. in sequential games where others’ prior contributions are known).

Inter-coder agreement

• Krippendorff’s α: 0.82

Definition
Cooperative behavior for games (e.g. prisoner’s dilemma and public goods game) with two choice options. Higher numbers equal higher cooperation.

Inter-coder agreement

• Krippendorff’s α: 0.95

Definition
Withdrawal behavior when the game type is a resource dilemma and the maximum choice range is about two. Higher numbers equate to lower cooperation.

Inter-coder agreement

• Krippendorff’s α: 0.564

Definition
Cooperative behavior when the game (e.g., prisoner’s dilemma and public goods dilemma) had a maximum number of choices greater than two. Higher numbers equate to higher cooperation.

For resource dilemmas, it should be calculated as: upper choice range – overall M withdrawal + lower choice range

Inter-coder agreement

• Krippendorff’s α: 0.97

Definition
Standard deviation of contributions or withdrawals, only applicable when maximum number of choices is greater than two.

Inter-coder agreement

• Krippendorff’s α: 0.94

Definition
Percentage of endowment contributed, calculated as (M – Lower choice option) / (Upper choice option – Lower choice option).

Inter-coder agreement

• Krippendorff’s α: 0.79

Definition
Trial on which cooperation was assessed.

Levels of the variable

• All trials: Cooperation was assessed across all trials.
• First trial: Cooperation was assessed in the first trial.
• Other trials: Cooperation was assessed in other trials.
• First and last trials: Cooperation was assessed in the first and the last trials.
• Last trial: Cooperation was assessed in the last trial.

Inter-coder agreement

• Krippendorff’s α: 0.69
• Most frequent level (%): All trials (92.3%)
• Agreement %: 94.9%

Definition
This variable is used as sample size to compute variance for meta-regressions predicting cooperation. It corresponds to the total sample size in a single study after exclusion of participants (N), but it uses the number of observations for studies that uses both individual and groups as unit of analysis (e.g., interindividual-intergroup discontinuity studies).

Definition
Whether the treatment or effect size quantitative information was published.

Levels of the variable

• TRUE = The treatment or effect size came from published information.
• FALSE = The treatment or effect size was calculated from raw data or requested to authors.

Definition
Whether the variable was manipulated using a between-participants versus within-participants experimental design.

Levels of the variable

• Between-participants experimental design
• Within-participants experimental design

Definition
Whether the independent variable was experimentally manipulated or measured.

Levels of the variable

• Measured: The variable was measured (e.g., self-report scale).
• Manipulated: The variable was manipulated and participants were exposed to different levels of the variable (e.g., experimental designs).

Definition
The type of cooperative behavior used as a dependent variable.

Levels of the variable

• Cooperative behavior: when game (e.g., prisoner’s dilemma or public goods dilemma) had a maximum of two choice options. Higher numbers equal higher cooperation.
• Cooperative behavior: when game (e.g. prisoner’s dilemma or public goods dilemma) has a maximum number of choices great than two. Higher numbers equal higher cooperation.
• Withdrawal behavior: when game type is a resource dilemma and with a maximum choice range greater than two. Higher numbers equal lower cooperation.

Definition
The trial on which cooperative behavior was assessed.

Levels of the variable

• All trials
• Other trials
• First trial
• First and last trials
• Last trial

Definition
Whether cooperation was coded as a percentage of endowment.

Levels of the variable

• TRUE = Percentage
• FALSE = Mean

Definition
Whether the reported results were main effects of specific variables or interactions between two or more variables.

Levels of the variable

• Main effect
• Interaction: Interaction effect
• Main effect, independent subsamples: Main effect computed on independent subsamples
• Main effect, dependent subsamples: Main effect computed on dependent subsamples

Definition
The number of levels or conditions that were coded for an independent variable (coded as N/A if a correlation was reported with a measured variable).

Definition
The level at which the data were analyzed (e.g., individual or group-based analyses).

Levels of the variable

• Trial: Data were analyzed at the trial level, so that cooperation for each trial is used as observation in the analysis.
• Block: Data were analyzed at the block level, so that cooperation for each block is used as observation in the analysis.
• Subject: Data were analyzed at the subject level, so that cooperation displayed by each participant is used as observation in the analysis.
• Group: Data were analyzed at the group level, so that cooperation displayed by each group as a whole is used as observation in the analysis.
• Session: Data were analyzed at the session level, so that cooperation displayed in each experimental session is used as observation in the analysis.
• Other: Data were analyzed at a level of analysis not included among the level of analysis categories.

Definition
A statistic is a measurement datum to describe a dataset or a variable. It is generated by a calculation on set of observed data. Definition extracted from STATO (the statistical methods ontology).

Levels of the variable

• F: F statistic
• t: t statistic
• X: Chi-square statistic
• r: Pearson’s correlation
• B: Standardized regression coefficient (beta)
• b: Unstandardized regression coefficient
• z: Z statistic
• H: Kruskal-Wallis H statistic
• U: Mann-Whitney U statistic
• W: Wilcoxon signed-rank test
• E: Fisher’s exact test
• rho: Spearman’s rank correlation coefficient
• W-X: Wald Chi-square statistic
• KS: Kolmogorov-Smirnov test
• G2: G-square or delta G-square

Definition
From STATO: The degree of freedom numerator is the number of degrees of freedom that the estimate of variance used in the numerator is based on.

Definition
From STATO: The degree of freedom denominator is the number of degrees of freedom that the estimate of variance used in the denominator is based on.

Definition
The numeric value of the test statistic indicated by ‘Test’.

Definition
From STATO: A quantitative confidence value that represents the probability of obtaining a result at least as extreme as that actually obtained, assuming that the actual value was the result of chance alone.

Definition
The direction of a comparison between two treatments.

Levels of the variable

• Negative: Level 1 > level 2
• Positive: Level 2 > level 1

Definition
For effect sizes reported in the original paper, the effect size measure.

Definition
The standardized effect size estimate reported in the manuscript.

Definition
Sample size in the experimental condition. If this is not reported, we assume equal sample size across conditions. Also, for within-subject comparison we code the number of participants in each condition, and not number of observations.

Definition
Proportion of cooperative choices in the experimental condition. Proportions are reported when the game (e.g., prisoner’s dilemma or public goods dilemma) had a maximum number of two choice options. Higher numbers equal higher cooperation.

Definition
Mean of cooperative or withdrawal behavior in the experimental condition. Means are reported when there was a maximum number of choices greater than two.

Definition
Standard deviation of contributions or withdrawals in the experimental condition. Only applicable when maximum number of choices is greater than two.

Definition

Percentage of endowment contributed, calculated as (M – Lower choice option) / (Upper choice option – Lower choice option).

Definition

The log-transformed proportion of the endowment contributed / rate of cooperation.

Definition

Coefficient of variation.

Definition
Annotators used comments sometimes to communicate some specific aspects of how a study was coded. This was originally intended to be for other coders, but we have now decided to make this information public in case it will be useful to users of the Databank.

Definition
Annotators were requested to provide a written description of the independent variables in the study. This was originally intended to be for other coders, but we have now decided to make this description public information in case it will be useful to users of the Databank.

Definition
Annotators were requested to list other variables that were measured in the study. This was originally intended to be for other coders, but we have now decided to make this information public in case it will be useful to users of the Databank.

Definition
Participant’s gender as reported in the paper. Positive correlations effect size always mean that men cooperated more than women.

Levels of the variable

• Male
• Female

Definition
Whether groups had a mixed gender composition (both males and women).

Levels of the variable

• TRUE: Groups had a mixed gender composition (both males and women)
• FALSE: Groups had a homogeneous gender composition (only males or only women)

Definition
Whether participants were aware of the gender of their partner(s).

Levels of the variable

• TRUE: Participants were aware of the gender of their partner(s)
• FALSE: Participants were not aware of the gender of their partner(s)

Definition
Punishment is the (possible) imposition of negative payoffs on a specific participant after the regular round of play.

Levels of the variable

• -1: The baseline treatment. This is the treatment compared against any treatment coded as punishment treatment (= 1 identifies the punishment effect). This means that the treatment is equal to the punishment treatment, except for the availability of punishment (which includes other concepts that are part of punishment, such as punishment effectiveness, etc.)
• 1: The punishment treatment
• 0: Is not a punishment treatment. This is any treatment that is neither a punishment treatment nor a baseline relative to the punishment treatment (In practice, these are ignored in the platform)

Definition
Who can enact punishment on other participants.

Levels of the variable

• Peer: Other participants of the same role
• Leader: A participant with some form of higher, relative social power; see Power
• Network: A peer in a game with network structure
• Experimenter: The person in charge of running the experiment (e.g., principle investigator, research assistant)
• Third Party: A person who does not interact with the participants in the main game e.g., a third-party punisher
• Institution: Institutional mechanism inside (e.g., impersonal sanctioning system) or outside (e.g., public institutions like government)
• Computer: Computerized opponent, and players were informed to be interacting with a computer
• Other: Other agent types not included in the current categories

Definition
The form in which participants’ punishment decisions affect their partner’s payoffs (see game incentive).

Levels of the variable

• Hypothetical: Hypothetical value for participants (e.g., points)
• Monetary: An amount of money often converted from game points
• Non-monetary Material: Amount of other resources (e.g., candies, school supplies)
• Non-monetary Social: Participants receive social incentives, such as blame or praise, from others based on the outcomes of the game

Definition

Whether there was a chance of lottery for punishment to be enforced (see lottery incentive).

Levels of the variable

• TRUE: A lottery was implemented for punishment to be enforced
• FALSE: A lottery was not implemented for punishment to be enforced

Definition

Whether group members make their punishment decisions simultaneously or sequentially, i.e., whether players take decisions one after another while receiving some form of feedback on preceding decisions (see Sequentiality).

Levels of the variable

• Simultaneous: Players make the punishment decision at the same time (i.e., without receiving information about others’ punishment decisions on that trial)
• Sequential turn-taking: Each player makes a punishment decision in turn
• Sequential leadership-by-example: In a game with leadership by example, there is a single privileged player that makes a punishment decision before all others

Definition

The ratio of the cost of punishment for the participant who decides to punish, to the cost of punishment for the participant who receives the punishment (e.g., if punishment costs 1 point to inflict a cost of 3 points, = .33).

Definition

The probability that a punishment decision is meted out (and thus becomes costly to the punished agent).

Definition

A rule according to which some number of players are punished.

Levels of the variable

• None: No punishment rule
• Contribution-based: Players who made a contribution less (or more) than a specified amount are punished
• Rank-based: The n lowest- (or highest-) ranked players are punished
• Outcome-based: Some set of players is punished if a collective threshold is (not) reached
• Random: Random players are punished

Definition

The number of iterations of punishment per punishment phase.

Definition
Whether communication was allowed between participants in the game.

Levels of the variable

• TRUE: Participants could communicate in the game
• FALSE: Participants could not communicate in the game

Definition
The mode of communication between participants in the game.

Levels of the variable

• Verbal: Participate could speak to each other, either in person or via phone/voice chat
• Written: Participants could communicate via written messages (either using free text entry or pre-written messages)
• Nonverbal: Other modes of communication, including nonverbal eye gaze, touch, and sound

Definition
Whether, in an iterated game, agents could communicate only once or throughout the experiment, such as before every trial.

Levels of the variable

• One-shot: Participants could only communicate once, before the first trial of the game
• Ongoing: Participants could communicate before multiple (all, or at least several) trials (if the game was one-shot, then communication was coded as ongoing)

Definition

Whether communication is restricted to pre-specified messages, or participants are free to formulate their own messages.

Levels of the variable

• TRUE: Communication restricted to pre-specified messages
• FALSE: Free communication

Definition

What was the content of communication, either given by pre-specified messages or coded from free communication.

Levels of the variable

• Irrelevant: Messages irrelevant to the game
• Threat: Threats (e.g., of non-cooperation or punishment)
• Promise: Promises (e.g., of own cooperation or reward)
• Request: Requests and suggestions (e.g., to cooperate or contribute)
• Other: Other content of messages
• N/A: If communication was not restricted, communication content is not coded.

Definition

Whether participants had a choice to communicate (or not) with their partner.

Levels of the variable

• TRUE: Optional communication
• FALSE: Mandatory communication

Definition

Whether the communication received by participants was real or fake.

Levels of the variable

• TRUE: Real communication
• FALSE: Fake communication