Emotion and decision-making: affect-driven belief systems in anxiety and depression

Abstract

Emotion processing and decision-making are integral aspects of daily life. However, our understanding of the interaction between these constructs is limited. In this review, we summarize theoretical approaches that link emotion and decision-making, and focus on research with anxious or depressed individuals to show how emotions can interfere with decision-making. We integrate the emotional framework based on valence and arousal with a Bayesian approach to decision-making in terms of probability and value processing. We discuss how studies of individuals with emotional dysfunctions provide evidence that alterations of decision-making can be viewed in terms of altered probability and value computation. We argue that the probabilistic representation of belief states in the context of partially observable Markov decision processes provides a useful approach to examine alterations in probability and value representation in individuals with anxiety and depression, and outline the broader implications of this approach.

Figure 1

This figure shows a simple gamble consisting of two options (A and B) with probable outcomes and shows the effect of modulating the value function (left column), the probability weighting function (middle column) and the resulting utility of the two available options (right column). Specifically, the objective value is transformed according to [Subjective Value] = k [Objective Value]^b and the probability is transformed according to [Subjective Weight] = [Probability]^1-a/([Probability]^1-a + (1− [Probability])^1-a), in accordance with proposals by Mukherjee [29] and Hsee [6]. The dark grey lines signify a larger distortion due to presumed affectively driven modulation of objective value or probability. The bar graphs indicate the overall utility of option A or B; a relatively larger subjective utility of A over B is assumed to result in a preference for A over B. In the first two rows, examples are given such that alterations of the parameter determining the subjective weight, a, can reverse (first line) or not reverse (second line) the preference of the gamble. In the third and fourth row a similar example is provided for alterations of the value parameter, b, that reverses or does not reverse the preference of the gamble. Taken together, these simple calculations show that alterations of probability and value in accordance with empirical and theoretical approaches to understanding the effect of emotion on decision-making can have significant preference reversal effects. We propose that the influence of emotion, in general, and of anxiety or depression, in particular, is that of changing the nonlinearity of the weighting and the value function, such that preference reversals occur. These reversals help to explain performance differences on risk-related decision-making tasks.

Figure 2

A schematic representation of partially observable Markov decision processes [89, 90]. A POMDP consists of a belief state, which summarizes previous experiences, is represented in a probabilistic framework, and is updated by a belief-state estimator. The updating occurs between the last action, the current observation, and the previous belief state based on Bayes’ rule. Decision-making occurs as a consequence of a decision policy that maps the current belief state onto actions. Emotions can affect this process in two ways. First, the observed rewards are hypothesized to be transformed into values based on the subjective value function as shown in Figure 1. Second, probabilities are hypothesized to be transformed into weights and can, therefore, affect the updating via the belief-state estimator. In other words, faulty updating by the belief-state estimator because of attenuated valuation or exaggerated representation of low probabilities can result in suboptimal estimation of the current state and, therefore, poor selection of a decision policy. For example, greater weighting of threat-related states in anxiety may result in avoidance of action. Alternatively, attenuated representation of subjective value may result in diminished learning and updating of the belief systems that provide the basis for making optimal decisions in depression.