Rank among Peers during Game Competition Affects the Tendency to Make Risky Choices in Adolescent Males

Jerome C. Foo,Kazuhiro Yoshiuchi,Kenji Morita,Kohei Nagase,Sawako Naramura-Ohno,Yoshiharu Yamamoto

doi:10.3389/fpsyg.2017.00016

Jerome C. Foo, Kazuhiro Yoshiuchi + Show 4 more

Open Access

https://doi.org/10.3389/fpsyg.2017.00016

Copy DOI

Abstract

It has been shown that adolescents take more risks when they are with peers than when they are alone, presumably because the presence of peers can be a social reward/punishment that can bias decision making. Competition is inherent in peer interactions, and recent work has demonstrated that winning/losing is an intrinsic social reward/punishment. Taken together, it can be hypothesized that competition amongst peers affects adolescents’ risky behavior. While there is much evidence that status amongst peers can relate to antisocial/aggressive behavior, it remains unclear whether risky behavior is affected. Moreover, the degree to which ‘temporary status,’ such as ranking in a short-term competitive game, affects behavior is uncertain, an important issue because adolescents might be sensitive to situations or factors which potentially destabilize existing hierarchies. In this experiment, these issues were directly explored in the classroom environment using smartphone technology and Wi-Fi setup. Male junior high school students (aged 14–15) performed a roulette game task on smartphones, playing either independently or against five classmates. In the latter case, the students’ current ranks within the group of six were constantly presented on smartphone screens. To dissociate the effects of the students’ reactions to ranks from their actual performances, unknown to the students, the ranks presented were actually predetermined so that about half of the students were continuously presented with high ranks whereas the other half were continuously presented with low ranks. We found that the students presented with low ranks made more risky plays than those not presented with ranks or those presented with high ranks. This result suggests that even temporary status significantly affects adolescents’ risky behavior, and also demonstrates the usefulness of smartphones in examining and manipulating peer interactions in classroom experiments.

Highlights

It has been shown that adolescents are inclined to engage in risky behavior, potentially resulting in serious problems (Steinberg, 2008; Somerville and Casey, 2010; Reyna and Brainerd, 2011; Defoe et al, 2015), and it has been found that this tendency for risky behavior is enhanced by the presence of peers (Steinberg, 2008)
In addition to the analyses based on the play ratio, we explored sensitivity to differences in expected values (EVs) in individual trials (11th to 120th, excluding the initial 10 trials) using a Generalized Estimating Equation (GEE) (Liang and Zeger, 1986) analysis in reference to an approach used in a previous study (Weller et al, 2015)
Similar to the results of the original GEE analysis for all the rank groups, significant main effects of EV and of Low-Rank, but no interaction between EV and Low-Rank were observed. These results indicate that presentation of low ranks generally increased the tendency to play, consistent with the results of the analyses described above at least to a certain extent, and that presentation of low ranks did not change the sensitivity to EVs

Summary

Introduction

It has been shown that adolescents are inclined to engage in risky behavior, potentially resulting in serious problems (Steinberg, 2008; Somerville and Casey, 2010; Reyna and Brainerd, 2011; Defoe et al, 2015), and it has been found that this tendency for risky behavior is enhanced by the presence of peers (Steinberg, 2008). Taken together with the abovementioned suggestion that peer presence affects risky behavior by modulating the reward system, we hypothesized that awareness of one’s place in competitive interactions amongst peers would further modulate the reward system, and potentially affect risky behavior more. It was difficult, to specify exact expectations of the possible effects from the studies introduced above; we discuss this issue in relation to a wider behavioral economics literature in the Section “Discussion.”

Methods

Results

Conclusion