Information use in risky decision making: Do age differences depend on affective context?

Abstract

The current study focused on the degree to which decision context (deliberative vs. affective) differentially impacted the use of available information about uncertainty (i.e., probability, positive and negative outcome magnitudes, expected value, and variance/risk) when older adults were faced with decisions under risk. In addition, we examined whether individual differences in general mental ability and executive function moderated the associations between age and information use. Participants (N = 96) completed a neuropsychological assessment and the hot (affective) and cold (deliberative) versions of an explicit risk task. Our results did not find a significant Age × Hot/Cold Condition interaction on overall risk-taking. However, we found that older adults were less likely to use the full decision information available regardless of the decision context. This finding suggested more global age differences in information use. Moreover, older adults were less likely to make expected-value sensitive decisions, regardless of the hot/cold context. Finally, we found that low performance on measures of executive functioning, but not general mental ability, appears to be a risk factor for lower information use. This pattern appears in middle age and progressively becomes stronger in older age. The current work provides evidence that common underlying decision processes may operate in risk tasks deemed either affective or deliberative. It further suggests that underlying mechanisms such as information use may be paramount, relative to differences in the affective context. Additionally, individual differences in neuropsychological function may act as a moderator in the tendency to use available information across affective context.

Burst Beliefs – Methodological Problems in the Balloon Analogue Risk Task and Implications for Its Use

“…For example, the models by Wallsten et al (2005) break down behavior into risk-taking, response consistency, and learning. In addition, computational models can be used to take into account censoring and to provide an index of uncensored risk-taking in the BART (Dijkstra et al, 2020;Tobin, 1958;Weller et al, 2019). A second way of dealing with the BART's limitations is by modifying the task, for example by rigging it (Figner et al, 2009;Slovic, 1966), providing additional feedback, or automating the responses (Pleskac et al, 2008).…”

“…Some of them employ Bayesian (generalized) linear mixed-effects regression (Weller et al, 2019;Young & McCoy, 2019); others use maximum likelihood estimation, adding a cumulative distribution function to the likelihood function to account for censoring (Dijkstra et al, 2020;Tobin, 1958). Such models perform significantly better (i.e., have less biased predictions) than those that do not account for censoring.…”

Untitled

“…For example, the models by Wallsten et al (2005) break down behavior into risk-taking, response consistency, and learning. In addition, computational models can be used to take into account censoring and to provide an index of uncensored risk-taking in the BART (Dijkstra et al, 2020;Tobin, 1958;Weller et al, 2019;Young & McCoy, 2019). A second way of dealing with the BART's limitations is by modifying the task, for example by rigging it (Figner et al, 2009;Slovic, 1966), providing additional feedback, or automating the responses (Pleskac et al, 2008).…”

Background Uncertainty Can Increase Risk Aversion in Decision Making

“…We also report the trial by trial analysis in the electronic supplementary materials C, as some authors (De Groot, 2020;Pleskac et al, 2008) strongly advise against using the aggregated adjusted BART scores as this overestimates the degree of risk aversion and treats missing values as randomly missing which is inaccurate. We followed recent suggestions (De Groot, 2020;Weller et al, 2019;Young & McCoy, 2019), and additionally conducted all analyses using Bayesian multilevel Poisson regression analysis accounted for the right-censored data implemented in the brms package for R (Bürkner, 2018).…”