Deliberative and Affective Risky Decisions in Teenagers: Different Associations with Maladaptive Psychological Functioning and Difficulties in Emotion Regulation?

Abstract

Using network analysis, we investigated the relationships between maladaptive psychological functioning, difficulties in emotion regulation, and risk-taking in deliberative and affective behavioral decisions. Participants (103 adolescents aged between 13 and 19 years, 62% boys) took the Cold (deliberative) and Hot (affective) versions of the Columbia Card Task and completed the Youth Self-Report (YSR) and the Difficulties in Emotion Regulation Scale (DERS). In contrast to the view that risk propensity increases from preadolescence to middle adolescence and decreases at later ages, our study revealed no age-specific trend. YSR syndrome scales were significantly correlated with risk propensity, but only in the Cold version. The YSR Thought Problems scale was the most central node in the network, linking internalizing and externalizing problems with risk propensity in the Cold CCT. Lack of emotional Clarity was the only DERS consistently linked with risk-taking both in correlation and network analyses. Maladaptive psychological functioning and difficulties in emotion regulation were linked with risk propensity in affective risky decisions through deliberative processes. The statistical significance of direct and indirect effects was further examined using nonparametric mediation analyses. Our study highlights the role of cognitive factors that in each variable set might account for risk-taking in teenagers.

Keywords: Columbia Card Task; deliberative and affective processes; difficulties in emotion regulation; executive functions; network analysis; risk propensity; thought problems; youth self-report.

Figure 1

Average number of flipped cards per trial in the Hot and Cold Columbia Card Task as a function of (a) potential gains, (b) potential losses, and (c) penalty cards.

Figure 2

Average number of flipped cards per trial in the (a) Hot and (b) Cold Columbia Card Task as a function of age and gender.

Figure 3

A network containing the eight syndrome scales of the Youth Self-Report, the six subscales of the Difficulties in Emotion Regulation Strategies, and risk propensity in decision-making tasks. The edge thickness represents the strength of an association.

Figure 4

Node strength (a), betweenness (b), and closeness centrality (c) estimates for eight syndrome scales of the Youth Self-Report (YSR), six subscales of the Difficulties in Emotion Regulation Strategies (DERS), and risk propensity in the decision-making tasks.

Figure 5

A network containing Externalizing Problems (EP), Internalizing Problems (IP), Cognitive Problems (CP), Difficulties in Regulating Emotions (DRE), Difficulties in Identifying Emotions (DIE), and risk propensity in the Hot and Cold CCT. The edge thickness represents the strength of an association. Legend: IP = Internalizing Problems; EP = Externalizing Problems; MCP = Mixed Cognitive Problems; DRE = Difficulties in Regulating Emotions; DIE = Difficulties in Identifying Emotions.

Figure 6

Node strength, betweenness, and closeness centrality estimates for three Youth Self-Report (YSR) indexes, two Difficulties in Emotion Regulation Strategies (DERS) indexes, and risk propensity in decision-making tasks.

Figure 7

Path model in which maladaptive psychological functioning and difficulties in emotion regulation predicted deliberative and affective risky decisions in teenagers. Note: Standardized estimates are in the picture. Rectangles represent observed variables. Unidirectional straight arrows represent the regression paths. Bidirectional curved arrows represent covariance. Dotted lines indicate nonsignificant model parameters paths. Legend: IP = Internalizing Problems; EP = Externalizing Problems; MCP = Mixed Cognitive Problems; DRE = Difficulties in Regulating Emotions; DIE = Difficulties in Identifying Emotions. ** p< 0.01; p = p-value; df = degrees of freedom; Chi² = Model’s chi square; TLI = Tucker-Lewis fit index; CFI = Comparative Fit Index; RMSEA = Root Mean Square Error of Approximation; SRMR = Squared Root Mean Square Residual; R² = R squared.