Between-group competition elicits within-group cooperation in children

Aggressive interactions between groups are frequent in human societies and can bear significant fitness costs and benefits (e.g. death or access to resources). During between-group competitive interactions, more cohesive groups (i.e. groups formed by individuals who cooperate in group defence) should out-perform less cohesive groups, other factors being equal (e.g. group size). The cost/benefit of between-group competition are thought to have driven correlated evolution of traits that favour between-group aggression and within-group cooperation (e.g. parochial altruism). Our aim was to analyse whether the proximate relationship between between-group competition and within-group cooperation is found in 3–10 years old children and the developmental trajectory of such a relationship. We used a large cohort of children (n = 120) and tested whether simulated between-group competition increased within-group cooperation (i.e. how much of a resource children were giving to their group companions) in two experiments. We found greater within-group cooperation when groups of four children were competing with other groups then in the control condition (no between-group competition). Within-group cooperation increased with age. Our study suggests that parochial altruism and in-group/out-group biases emerge early during the course of human development.

analyses to test the effect of kinship and familiarity on the two dependent variables in experiment #1.
Within each experimental condition, half of the groups played experiment #1 first and the other half played experiment #2 second. Children were given around 10 minutes to rest between the two experiments.
At the beginning of each experimental session we greeted and thanked the parents and their children for their participation. We asked the children to seat at the four sides of a table at the centre of the experimental room (parents waited outside). One of us (LM) led the experiment and collected data with the help of two assistants. Once the children at the experimental table, we recorded the ID code of each child (which was used after the experiment to get data on their date of birth, gender and whether they had special needs) and their position at the table. We described the game instructions to the children following the text below, depending on whether we had allocated the group of four children to the control or between-group competition condition, and on whether they were going to play experiment #1 or experiment #2 first. At the end of each experiment we let children to exchange their stickers with a range of reward stickers that they could 'buy' according to the rules described in the game instructions (3 star stickers 'buy' one reward sticker). A large range of reward stickers, suitable for children of different age and gender, was available for the children to choose on. If children had to play the second experiment (experiment #1 or #2 depending on which experiment they had played first), we gave them approximately 10 minutes to rest before starting the second experiment play. Once the break was over, we asked children to seat at the same position at the table where they were sitting for the first experiment and we read them the game instructions. At the end of the second experiment children could exchange the star stickers then had kept/gained in the second experiment with additional reward stickers, following the rules described above. Once the two experiments had been completed, we debriefed the children and their parents about the aims of the study and thanked them for their participation. Descriptive statistics on within-group cooperation performed by children in each round of the two experiments is shown in Table S2 below.
In order to control whether our dependent variables were over-dispersed, we compared the mean to the variance for each dependent variable and then we visually inspected the frequency distribution of the data. For experiment #1, the variance was less than two times the mean (number of stickers the player gave to children in their group: mean=1.09, variance=1.79; number of children the player gave stickers to in their group: mean=0.99, variance=1.60), indicating little over-dispersion. However, visual inspection of the frequency distribution of these two variables showed a marginal skew in favour of zero values. Thus, to further control for the effect of over-dispersion on our results for experiment #1, we ran two mixed-model negative binomial regressions (MMNBRs) with the same fixed and factors used in the two mixed-model Poisson regressions. The two MMNBRs gave very similar results to those obtained with the mixed-model Poisson regression. The number of stickers given, and the number of recipients of stickers, were significantly greater in the BGC than in the control condition (p=0.026 ad p=0.008, respectively). The relationship between age of the children and the number of stickers given was marginally non-significant (p=0.071) whereas older children gave stickers to a larger number of recipients than younger children (p=0.018). The other fixed factors had no significant effect. Moreover, the likelihood-ratio tests of the overdispersion parameter alpha linked to each the two MMNBRs were not significant (both p>0.49). Therefore, we concluded that over-dispersion did not affect the results of experiment #1. For experiment #2, the variance was lower than the mean of stickers given by children (mean=1.53, variance=1.28) and visual inspection of the frequency distribution of this variable showed no skew. Thus, we did not run any additional check on over-dispersion for experiment #2 Experiment 1 -Game instructions for the control condition 1. You are going to play a game. You will be in this same group during the whole game.
Please pay careful attention to the game instructions. If you have any question, please raise your hand. Otherwise, please be quiet and listen carefully, just like you would to your teacher in school. 2. We will randomly select one of you as the player of this game. If you are the selected child (the player), we will give you three star stickers. You have to decide whether you are going to keep the star stickers for yourself and/or you are going to give some/all of stickers to one or more children in the group. If you want to keep some/all stickers for yourself, you can leave them next to you on the table. If you want to give some/all of stickers to one or more children in the group, you can pass the stickers to the other child/children of your choice and leave the stickers in front of them on the table. Please do whatever you wish, either share your stickers or keep them; both are good options. 3. Once you have made your choice, we will ask children who have stickers to put them in the plastic glass provided so to avoid confusion with the stickers used in the next round. After this, we will start another round of the game and randomly select another player. 4. You can keep the stickers, that you have kept for yourself and/or that you have received from other children, next to you on the table. We will play an indefinite number of rounds of this game and you might be 'the player' once, more than once or never. 5. Once this game is completed, you can count all the stickers you kept or received from other children and exchange your star stickers with the reward stickers of your choice: you need three start stickers to get one reward sticker. 6. If you have any question, please raise your hand now. Otherwise, let's start the game.

Experiment 1 -Game instructions for the between-group competition condition (text underlined is for instructions only used in the between-group competition condition)
1. You are going to play a game. You will be in this same group during the whole game.
Please pay careful attention to the game instructions. If you have any question, please raise your hand. Otherwise, please be quiet and listen carefully, just like you would to your teacher in school. 2. We will randomly select one of you as the player of this game. If you are the selected child (the player), we will give you three star stickers. You have to decide whether you are going to keep the star stickers for yourself and/or you are going to give some/all of stickers to one or more children in the group. If you want to keep some/all stickers for yourself, you can leave them next to you on the table. If you want to give some/all of stickers to one or more children in the group, you can pass the stickers to the other child/children of your choice and leave the stickers in front of them on the table. Please do whatever you wish, either share your stickers or keep them; both are good options. 3. Once you have made your choice, we will ask children who have stickers to put them in the plastic glass provided so to avoid confusion with the stickers used in the next round. 4. You can keep the stickers, that you have kept for yourself and/or that you have received from other children, next to you on the table. We will play an indefinite number of rounds of this game and you might be 'the player' once, more than once or never. 5. We are running a tournament where we compare what you do during this game with what children in other groups have done. The top three groups at the moment in our tournament are White, Black and Grey, as you can see on the league table on the wall. Let's see if you can beat these three top groups. 6. Once this game is completed, you can count all the stickers you kept or received from other children and exchange your star stickers with the reward stickers of your choice: you need three start stickers to get one reward sticker. 7. If you have any question, please raise your hand now. Otherwise, let's start the game.

Experiment 2 -Game instructions for the control condition
1. You are going to play a game consisting of various rounds. You will be in this same group during the whole game. Please pay careful attention to the game instructions, because the clearer the game instructions are the more star stickers you can earn, and the more star stickers you can earn the more reward stickers you can get at the end of the game). If you have any question, please raise your hand. Otherwise, please be quiet and listen carefully, just like you would to your teacher in school. 2. We will give each of you three star stickers before playing each round of the game. You have to decide whether you are going to keep the star stickers for yourself and/or you are going to give some/all of stickers to the common group project. 3. Next to you, you have two envelopes. The envelop that has "you" written on is where you can put the stickers you want to keep for yourself (if any) whereas the envelope that has "group" written on is where you put the stickers you want to give to the common group project. 4. Your choices have to be kept secret so please do not show whether and how many stickers you put in each envelope, and put the group envelope in the common group bowl even if it has no stickers in it. The best way to keep your decision secret is for you to move away from the table, at your side of the room, put the stickers in the two envelopes according to your decision, and then return to the table to continue the game. We do not think it would be better for you to give the stickers to the common group project or better to keep them for yourself. 5. When I clap my hands, you will need to make your choice. Once you have put the stickers in the two envelopes, you can leave the "you" envelop next to you and the "group" envelope in the common group project box at the centre of the table. Please put the "group" envelope in the box even if it doesn't contain any stickers and remember your choice is secret. 6. Once all of you have made your choice and positioned the two envelopes, we will open the "group" envelopes in the common group project box and count them. We will multiply the stickers in the common group project box by two and divide them equally among you, no matter how many (if any) stickers you have given to the common group project. This means that if, for example, in the common group project box there are 4 stickers in total, we will multiply this number by two, giving us eight stickers in total; each of you will receive two stickers from the common group project no matter how many stickers you put in the common group project box. 7. Once you receive the stickers from the common group project you can put them in the container at the side of your chair, together with the stickers you kept for yourself (the ones you put in the "you" envelope). Please remember that it is a secret how many stickers you have so do not show the content of this box to the other children. 8. Once these steps have been completed, we will give each of you three new star stickers and we will play another round following the same steps of the first round. We will play an indefinite number of rounds before completing the game. 9. At the end of each game, you can count all the stickers you have in the box at the side of your chair and exchange your star stickers with the reward stickers of your choice: you need three start stickers to get one reward sticker. 10. If you have any question, please raise your hand now. Otherwise, let's start the game.

Experiment 2 -Game instructions for the between-group competition condition (text underlined is for instructions only used in the between-group competition condition)
1. You are going to play a game consisting of various rounds. You will be in this same group during the whole game. Please pay careful attention to the game instructions, because the clearer the game instructions are the more star stickers you can earn, and the more star stickers you can earn the more reward stickers you can get at the end of the game). If you have any question, please raise your hand. Otherwise, please be quiet and listen carefully, just like you would to your teacher in school. 2. We will give each of you three star stickers before playing each round of the game. You have to decide whether you are going to keep the star stickers for yourself and/or you are going to give some/all of stickers to the common group project. 3. Next to you, you have two envelopes. The envelop that has "you" written on is where you can put the stickers you want to keep for yourself (if any) whereas the envelope that has "group" written on is where you put the stickers you want to give to the common group project. 4. Your choices have to be kept secret so please do not show whether and how many stickers you put in each envelope, and put the group envelope in the common group bowl even if it has no stickers in it. The best way to keep your decision secret is for you to move away from the table, at your side of the room, put the stickers in the two envelopes according to your decision, and then return to the table to continue the game. We do not think it would be better for you to give the stickers to the common group project or better to keep them for yourself. 5. When I clap my hands, you will need to make your choice. Once you have put the stickers in the two envelopes, you can leave the "you" envelop next to you and the "group" envelope in the common group project box at the centre of the table. Please put the "group" envelope in the box even if it doesn't contain any stickers and remember your choice is secret. 6. Once all of you have made your choice and positioned the two envelopes, we will open the "group" envelopes in the common group project box and count them. We will multiply the stickers in the common group project box by two and divide them equally among you, no matter how many (if any) stickers you have given to the common group project. This means that if, for example, in the common group project box there are 4 stickers in total, we will multiply this number by two, giving us eight stickers in total; each of you will receive two stickers from the common group project no matter how many stickers you put in the common group project box. 7. We are running a tournament where we calculate children's contribution to the common group project. The top three groups at the moment in our tournament are White, Black and Grey, as you can see on the league table on the wall. You have been selected to take part in this tournament. At the end of the game, if your group ends up being one of the top three groups, we will calculate the difference between the total contribution to the common group project of your group and that of the group standing immediately below yours. We will double that difference and distribute the stickers equally among you. For example, if you end up being above the White group and the difference between your group and the White group is 4 stickers in total, we will multiply this number by two, giving us eight stickers in total; each of you will receive two stickers from the common group project no matter how many stickers you put in the common group project box. 8. Once you receive the stickers from the common group project you can put them in the container at the side of your chair, together with the stickers you kept for yourself (the ones you put in the "you" envelope). Please remember that it is a secret how many stickers you have so do not show the content of this box to the other children.