Graphs versus numbers: How information format affects risk aversion in gambling

Abstract

In lottery gambling, the common phenomenon of risk aversion shows up as preference of the option with the higher win probability, even if a riskier alternative offers a greater expected value. Because riskier choices would optimize profitability in such cases, the present study investigates the visual format, with which lotteries are conveyed, as potential instrument to modulate risk attitudes. Previous research has shown that enhanced attention to graphical compared to numerical probabilities can increase risk aversion, but evidence for the reverse effect — reduced risk aversion through a graphical display of outcomes — is sparse. We conducted three experiments, in which participants repeatedly selected one of two lotteries. Probabilities and outcomes were either presented numerically or in a graphical format that consisted of pie charts (Experiment 1) or icon arrays (Experiment 2 and 3). Further, expected values were either higher in the safer or in the riskier lottery, or they did not differ between the options. Despite a marked risk aversion in all experiments, our results show that presenting outcomes as graphs can reduce — albeit not eliminate — risk aversion (Experiment 3). Yet, not all formats prove suitable, and non-intuitive outcome graphs can even enhance risk aversion (Experiment 1). Joint analyses of choice proportions and response times (RTs) further uncovered that risk aversion leads to safe choices particularly in fast decisions. This pattern is expressed under graphical probabilities, whereas graphical outcomes can weaken the rapid dominance of risk aversion and the variability over RTs (Experiment 1 and 2). Together, our findings demonstrate the relevance of information format for risky decisions.

Display of major risk categories for android apps.

“…In gambling, a lottery includes two parts, the probability and the outcome, either of which can be displayed with numbers or graphs. As an example, Dambacher, Haffke, Groß, and Hübner (2016) investigated how information display format affects risk aversion in gambling (i.e., preference of an option with higher winning probability but lower reward outcome). Participants were asked to choose from two lotteries with probabilities or outcomes displayed in numerical or graphical (bar graph or icon array) formats.…”

“…The above reviewed studies suggest that the graphical format increases the salience of the displayed information (Dambacher et al, 2016). Yet, whether to adopt a graphical or numerical/table display depends on the specific task (e.g., choice vs. estimation; Feldman-Stewart et al, 2000;Hawley et al, 2008) and the purpose of the presentation (e.g., risk avoidance or not; Stone et al, 1997Stone et al, , 2003.…”

“…We hypothesized that (1.1) risk ratings would be higher when the risk scores increased, and (1.2) self-reported risk concerns would interact with the risk categories in a way that the most concerned risk category would be weighted more in the ratings. On the basis of studies in other domains, we also hypothesized that (1.3) participants would be more risk aversive with bar-graph than with table risk displays (Dambacher et al, 2016;Stone et al, 1997Stone et al, , 2003, and (1.4) vertical bar-graph display would yield higher accuracy rate and/or faster response times (Feldman-Stewart et al, 2007;McCaffery et al, 2012). No direc-This document is copyrighted by the American Psychological Association or one of its allied publishers.…”

Attention biases the process of risky decision‐making: Evidence from eye‐tracking

“…Visual factors are usually reflected in the visual features attached to an option, including color, size, location, saliency, and visual format. Studies have shown that the influence of visual factors on bottom‐up attention is revealed through early attention, usually the first fixation (Dambacher et al, 2016 ; Spitmaan et al, 2019 ). Options' color, location, and size can capture the first fixation to the more visually salient option, regardless of cognitive factors or option value (Fisher, 2021 ; Kwak & Huettel, 2018 ; Roberts et al, 2021 ; Vanunu et al, 2021 ).…”

Stepwise training supports strategic second-order theory of mind in turn-taking games

“…Importantly, Marble Drop is game-theoretically isomorphic to Matrix Games and thus requires essentially the same reasoning (see next subsection). Instead of using numerical payoffs, as commonly used in experimental games, we chose colored marbles to counter numerical but non-optimal 8Visual presentation has been shown to support other judgment and decision tasks, see (Dambacher, Haffke, Groß, & Hübner, 2016).…”