No effect of short term exposure to gambling like reward systems on post game risk taking

Abstract

Is engaging with gambling-like video game rewards a risk factor for future gambling? Despite speculation, there are no direct experimental tests of this "gateway hypothesis". We test a mechanism that might support this pathway: the effects of engaging with gambling-like reward mechanisms on risk-taking. We tested the hypothesis that players exposed to gambling-like rewards (i.e., randomised rewards delivered via a loot box) would show increased risk-taking compared to players in fixed and no reward control conditions. 153 participants (M_age = 25) completed twenty minutes of gameplay-including exposure to one of the three reward conditions-before completing a gamified, online version of the Balloon Analogue Risk Task (BART). Self-reports of gambling and loot box engagement were collected via the Problem Gambling Severity Index, and Risky Loot-Box Index. Bayesian t-tests comparing BART scores across reward conditions provided moderate to strong evidence for a null effect of condition on risk-taking (BF = 4.05-10.64). Null effects were not moderated by players' problem gambling symptomatology. A Spearman correlation between past loot box engagement and self-reported gambling severity (r_s = 0.35) aligned with existing literature. Our data speak against a "gateway" hypothesis, but add support to the notion that problem gambling symptoms might make players vulnerable to overspending on loot boxes.

Figure 1

Mean BART scores according to reward condition.

Figure 2

Prior and Posterior Plots, and Robustness Checks for Bayesian T-Tests.

Figure 3

Sequential analyses for Bayesian T-tests.

Figure 4

Density plot of PGSI score distribution.