About this paper symposium

---

Panel information
Panel 4. Cognitive Processes

Paper #1
Children Differentiate Between Different Domains of Risk: Economic, Social, and Physical
Author information	Role
Rachael Silberstein, Boston University, US	Presenting author
Peter R. Blake, Boston University, US	Non-presenting author
Abstract
Understanding how children perceive and evaluate risks across various situations is crucial for enhancing safety and long-term outcomes (Boyer, 2006; Defoe et al., 2015). Despite substantial research on risky choices throughout development, theories differ on whether risk-taking is domain general or domain specific (varying across types of risk) (Hanoch et al., 2006; Blais & Weber, 2006), and few studies have directly compared children’s evaluations of different risk types. Here, we investigated whether and how 4- to 9-year-old children differentiate between three risk types: physical, social, and economic. In two pre-registered studies, children saw two scenarios of each risk type (counterbalanced). In each, a character chooses between a risky option with two possible outcomes and a certain option. For example, a physical risk involved a choice between a long, boring bike ride to school (certain) or a faster, steeper ride with risk of injury (risky). Social risks involved possible peer exclusion or embarrassment, and economic risks involved choosing between a set amount of candy or a gamble to get more candy or none. Children 1) rated all scenarios on a 7-point Risk Scale (1= not risky, 7 = very risky), 2) rank-ordered three scenarios (one per risk type) from most to least risky, and 3) indicated which option they would choose. Study 2 increased the magnitude of risk for economic and social scenarios. We predicted that children would rate physical risks highest, followed by social, then economic, and that their differentiation between risk types would increase with age. We predicted that the self-choice measure would mirror the risk ratings, with children avoiding activities they rated most risky. A total of 193 children (Study 1 n = 94, Study 2 n = 99, Mage = 7;00, range = 4;02–9;11, predominantly White and mid-to-high SES) participated on Zoom or in-person. Consistent with our hypotheses, children in both studies rated physical risks as more risky than social (Study 1 β = 2.08, p < .001; Study 2 β = 2.20, p <.001) and economic risks (Study 1 β = 1.98, p < .001; Study 2 β = 1.22, p < .001), and ranked physical risks as most risky. Social and economic risk were rated the same in Study 1, but economic risk was rated higher in Study 2 (Figure 1). On the self-choice measure, children were more likely to choose risky options in social (Study 1 β = -1.88, p < .001; Study 2 β = -2.73, p <.001 and economic (Study 1 β = -2.04, p < .001; Study 2 β = -1.99, p < .001) scenarios than in physical ones, as we predicted (Figure 2). In partial support of the domain specific hypothesis, only social and economic risks were correlated in Study 1, but in Study 2 no risk types were correlated. These results suggest that children have distinct perceptions of different types of risks, indicating that risk preferences may be domain-specific early in development. Further research will need to test whether some domains of risk are more predictive of later outcomes than others.

Paper #2
Longitudinal Patterns of Risk Taking and Risk Perception Across Adolescence in Seven Countries
Author information	Role
Dr. Natasha Duell, Ph.D., Cal Poly San Luis Obispo, United States	Presenting author
Sami Griesdorf, Cal Poly San Luis Obispo, US	Non-presenting author
Christina Bertrand, University of Massachusetts Amherst, US	Non-presenting author
Ann Folker, University of Massachusetts Amherst, US	Non-presenting author
Kirby Deater-Deckard, University of Massachusetts Amherst, US	Non-presenting author
Abstract
Introduction: Experimental and epidemiological reports show that risk taking increases during the second decade of life, yet much of this research has been cross-sectional. Although heightened risk taking is thought to be a “universal” characteristic of adolescence, there is diversity in opportunities for risk as well as attitudes towards risks in different parts of the world. As such, the present study examined longitudinal trends in behavioral and self-reported risk taking as well as the perceived dangers of risk behaviors in a cross-national sample of adolescents. Hypotheses: It was hypothesized that self-reported and behavioral risk taking would increase linearly over time, whereas the perceived riskiness of both health and antisocial risk behaviors would decrease over time. Study population: Participants included 1,093 adolescents (50% female) from Colombia, Italy, Jordan, Kenya, the Philippines, Thailand, and the United States, as part of the longitudinal Parenting Across Cultures study. Data from Waves 3, 6, and 9 were used in the present analysis. Youth were M = 10.72 years (SD = 0.67) at Wave 3, M = 14.51 years (SD = 0.96) at Wave 6, and M = 17.75 years (SD = 1.02) at Wave 9. Methods: Self-reported risk taking and risk perception were measured using the Benthin scale. Risk taking was divided into health risks (e.g., binge drinking, smoking, having unprotected sex) and antisocial risks (e.g., vandalism, sneaking out of the house, fighting). Participants reported on the frequency with which they engaged in each risk in the past six months. The risk perception questions examined the extent to which adolescents perceived each of the health and antisocial risks as being scary, how much the potential benefits outweighed the potential costs, and how serious the consequences of the risk would be if something “bad” happened as a result. Behavioral risk taking was also examined using the Balloon Analog Risk Task (BART) and the Stoplight Task. Latent growth curve modeling in Mplus v.8.1 was used to estimate average linear growth for each measure in the full sample and within each country. Results: Results from the full-sample analyses indicated that self-reported health risk taking increased linearly between ages 11-18 years, whereas self-reported antisocial risks did not change. The perceived riskiness of both health and antisocial risks decreased between ages 11-18 years. For the behavioral indices of risk taking, results indicated that risk taking on the BART increased linearly, whereas risk taking on the Stoplight decreased linearly. Individual variations were also observed. Conclusion: Although all adolescents perceived fewer dangers to risk behaviors as they got older, patterns of risk taking over time were more variable. As such, risk taking in adolescence is not a monolith. Nuances related to opportunities, motivations, and cultural norms will be discussed.

Paper #3
A preference for epistemic over physical uncertainty across development: Evidence from three diverse cultures
Author information	Role
Julia M. Smith, Duke University, US	Presenting author
Hadi Mohamadpour, Iran, University of Tabriz	Non-presenting author
Jan Engelmann, Arizona State University, US	Non-presenting author
Helen Elizabeth Davis, University of California, Berkeley, US	Non-presenting author
Justine Krieger, Texas A&M University, US	Non-presenting author
Bettina Soerensen, National Science Foundation, US	Non-presenting author
Jeremy Koster, National Science Foundation, US	Non-presenting author
Soomaayeh Heysieattalab, University of Tabriz, Iran	Non-presenting author
Dorsa Amir, Duke University, US	Non-presenting author
Abstract
We regularly make decisions under uncertainty, but the same decision can feel psychologically different depending on the source of the uncertainty. Physical uncertainty arises when an outcome of interest has not yet occurred and a decision maker has to guess what will happen. On the other hand, epistemic uncertainty refers to a lack of knowledge on the part of a decision maker about an outcome that has already occurred. For example, the results of an election are physically uncertain before the votes have been cast, and epistemically uncertain after votes are cast but before they are counted. Past research suggests that children prefer epistemic to physical uncertainty, presumably because they find it easier to imagine an outcome that has already occurred than an outcome that has yet to occur (Beck, Robinson, & Rowley, 2012). However, it is unclear whether these imagined outcomes become less compelling as children develop better metacognitive abilities. Studies that asked adults about their preferences in hypothetical scenarios suggest that preferences for epistemic uncertainty may diminish or even reverse over the course of development, but studies asking adults to make in vivo choices suggest that they, like children, prefer epistemic uncertainty. Moreover, preferences for different types of uncertainty have never been examined outside of Western societies, leading to open questions about the generalizability of these patterns across diverse settings. In this study, we examined children and adolescent's preferences for epistemic versus physical uncertainty in three populations: among urban communities in Iran (N = 100) and the United States (N = 56), in addition to hunter-horticulturalist Tsimane communities in Bolivia (N = 54). We demonstrated two types of dice games: a physical “guess then roll” game in which children guessed what number would be shown on a die before it was rolled, and an epistemic “roll then guess” game in which children guessed the outcome after the die was rolled but before it was revealed. In all three cultural groups, children (ages 6-19) were more likely to prefer guessing the outcome of a dice toss after the die had been thrown (epistemic uncertainty), as opposed to before (physical uncertainty). There were no age-related differences in preferences for epistemic uncertainty, challenging the idea that this preference is reduced or reversed over the course of childhood development. The observed consistency across cultures, including those with no previous experience with dice, further suggests that the preference for epistemic uncertainty is not an artifact of familiarity with objects such as dice. Taken together, these results suggest a seemingly stable preference for epistemic uncertainty over physical uncertainty across development and across diverse cultural settings.

⇦ Back to session