Individual–Group Discontinuity in Group–Individual Interactions: Does Size Matter?

Abstract

The consistent fi nding that interactions involving groups are more competitive than interactions between individuals is known as the discontinuity effect. We investigated the effects of group size in order to determine whether the effect is a true discontinuity or a continuous function of group size. We also asked whether dyads behave like larger groups. Four hundred and eighty seven students volunteered for the experiment. Individuals played 10 trials of two-choice prisoner's dilemmas against other individuals or against groups of sizes two through eight. Major fi ndings included (a) individual-individual interactions were more cooperative than individual-group interactions, but there were no differences among group sizes; (b) dyads were indistinguishable from larger groups; and (c) data on expectations revealed that, before interacting, groups were not only distrusted by individuals but also distrusted the individuals. The discontinuity between individuals and groups-including dyads-appears to be a true discontinuity.

Sharing responsibility with a machine

“…Second, the identi ability hypothesis proposes that deciding in groups provides a shield of anonymity which could also drive sel sh decision-making. ird, the ingroup-favouring norm could put some pressure on decision makers to behave in a way which bene ts the group before taking 3 is has been shown in a number of experimental games such as the Trust Game (Kugler et al, 2007), the Ultimatum Game (Bornstein and Yaniv, 1998), the Coordination Game (Bland and Nikiforakis, 2015), the Signaling Game (Cooper and Kagel, 2005), the Prisoners Dilemma (McGlynn et al, 2009), the Gi Exchange Game , the Public Good Games (Andreoni and Petrie, 2004) as well as in lo eries (Rockenbach et al, 2007) and Beauty Contests (Kocher and Su er, 2005;Su er, 2005). 4 Bland and Nikiforakis (2015) study a coordination game with third party externalities and nd sel sh behaviour among the joint decision makers even when the sel sh option imposes a strong negative externality on a third-party.…”

Rabble of Robots Effects

“…Although this was less than ideal, evidence on the discontinuity effect [26] suggests that it would not have a great influence on responses to robots. Further, our results do not suggest that the pattern of responses to groups of robots is directly affected by the number (two or three) of robots.…”

“…Previous research on the 'discontinuity effect' found that groups of various sizes (two to eight members) show similar patterns of interaction, with both group and non-group members [26]. For the Group conditions in this study, we used two to three robots as numbers that would contrast with the single-robot conditions.…”

The Influence of Robot Number on Robot Group Perception—A Call for Action

“…A precondition for effectively deploying robots in society and for understanding and predicting user reactions is the knowledge about the boundary conditions for the perception of a given number of robots in terms of a "group." This is important when trying to understand user behavior toward robots, as behavior toward individuals differs from behavior toward groups [5] in that people act much more aggressively and less cooperatively in intergroup situations compared to interindividual situations (e.g., References [20][21][22]). Thus, it must be systematically determined to what extent groups of humans and robots are similar and to what extent they differ.…”