|
| Predicting Risk Taking on the Balloon Analogue Task: A Dynamic Analysis of Choice |
| Monday, May 26, 2014 |
| 11:00 AM–11:50 AM |
| W176c (McCormick Place Convention Center) |
| Area: EAB; Domain: Basic Research |
| Chair: Elias Robles (Arizona State University) |
| Abstract: The Balloon Analog Risk Task (BART; Lejuez et al., Journal of Experimental Psychology: Applied, 2002, 8, 7584) is a behavioral assessment developed to estimate individual propensity to take risks. Responding on the BART is probabilistically followed by wins or loses over a fixed number of opportunities (balloons). BART scores positively correlate with other behavioral and self-report measures of impulsivity and risk taking. The presentations that make up this symposium explore the utility of dynamically analyzing local response patterns as a function of wins, losses, nature of the reward, and amount of exposure to a modified balloon analogue task (BAT). The first study looks at the predictive validity of the task in relation to real and hypothetical rewards; The second looks at the effects of individual balloon explosions on the behavior of risk takers and non-risk takers; and the third describes an exponential model that effectively predicts responding on individual trials. Taken together, these studies show how differential propensity to make risky choices interacts with the contingencies of reinforcement embedded in the task to produce systematic response patterns, and characteristic overall risk taking scores. |
| Keyword(s): balloon analogue, choice, dynamic analysis, risk taking |
| |
Predictive Validity of the Balloon Analogue Task (BAT): Choice Between Real and Hypothetical, Certain and Uncertain Cash |
| MUCHEN ZHU (Arizona State University), Sineenuch Wongsomboon (Arizona State University), Araceli Moreno (Arizona State University) |
| Abstract: Balloon analogue tasks are behavioral simulations developed to estimate individual propensity to take risks. Responding on the BAT is probabilistically followed by wins or loses over a fixed number of opportunities (balloons). BAT scores positively correlate with other behavioral and self-report measures of impulsivity and risk taking. Following random assignment to 1 of 3 choice tasks, young adults (n=61) played the BAT (30 balloons) for points with no monetary value, and then were offered a choice between $5 in cash and betting to win $0 to $15. A fourth group was offered a hypothetical choice between $5 and $0 or $10 based on a coin flip. As expected, the mean number of adjusted responses per balloon in the BAT predicted choice for the riskier monetary option, regardless of whether the choice was real or hypothetical. In addition, the specific task offered to participants in each group significantly contributed to the choice between certain and uncertain cash. |
| |
| Dynamic Adaptation to Wins and Losses on the Balloon Analogue Task (BAT) |
| ROBERT ROSS (Arizona State University), Mikayla Gilbreath (Arizona State University) |
| Abstract: The effects of 1, 2, and = 3 consecutive explosions on subsequent risk taking on the BAT were assessed in 30- and 60-balloon sessions. One hundred college students played the BAT for either cash or points without monetary value. Individual responding within balloons was analyzed in terms of the number of immediately preceding consecutive explosions and the distance between each unexploded balloon and the last explosion. A systematic decrease in responding was observed after explosions, followed by recovery to above average levels. In 4 out of 6 cases this pattern was well described by a linear model accounting for = 75% of the variance; data for the remaining function did not have enough cases. Both, the observed decrease in responding following explosions and the subsequent recovery to each subject’s mean were a function of the frequency and recency of the immediately preceding explosions. The effects on risk taking of earning cash and points without monetary value are discussed. |
| |
| Modeling Within- and Between-Trial Patterns of Responding on the Balloon Analogue Task (BAT) |
| AMANDA SMALL (Arizona State University), Brandon McColley (Arizona State University) |
| Abstract: Mean adjusted responses (those occurring on unexploded balloons) follow a negatively accelerated pattern over the course of a BAT session. Based on 30-balloon sessions, we tested (n = 61) an exponential raise to the maximum model, r = a (1- e ^(-bi)), where r is the number of responses on balloon in position i, and a is the intercept corresponding to the subject’s overall risk taking score. The simple model effectively predicted (R2 = .75) responding on a different sample of college students (n = 40) playing a 60-balloon BAT session. Consistent patterns of responding within balloons were also observed. For risk takers, responding during unexploded balloons was often characterized by a high steady rate followed by a decrease in response rate right before cashing points to the permanent bank; conversely, for non-risk takers, responding within balloons was often characterized by a wavy pattern resulting from alternating periods of higher and lower rates, and a lower mean number of adjusted responses. These patterns are consistent with rates of risk evaluation associated with varying degrees of loss aversion. |
|
| |