Prospects for a Dual-Process Account in the Development of Selective Trust

• In Event: 1-086 - Cognitive Mechanisms of Selective Learning

Integrative Statement

In the last decades, much research has been done on selective social learning in preschoolers. Yet, the findings remain inconclusive regarding the cognitive foundations of early selective trust: Some research suggests basic and non-sophisticated heuristics and biases. For example, children draw wide generalizations between unrelated domains, selectively learning novel words from attractive, rather than unattractive, models. Other evidence, however, points to adult-like rational capacities. For example, when introduced to someone strong and someone smart, children competently chose between these models for strength- und knowledge-related tasks, suggesting an understanding of the agents’ individual traits.
We propose that children may use multiple strategies in their selective trust: sophisticated rational processes based on trait reasoning and much simpler heuristics. Their interplay might be best described by a dual-process account (Hermes, et al., 2018a). Dual-process theories usually assume two qualitatively different kinds of processes. Type-I processes operate fast, automatically, are early-developing and relatively independent from cognitive resources. In contrast, Type-II processes are relatively slow, depend strongly on knowledge and cognitive resources, and develop later. In selective trust, an example for Type-I processes may be a simple trust-the-better heuristic that distinguishes between agents in evaluative terms, but in broad and undifferentiated ways. When confronted with a problem for which their Type-I heuristics provide no unique solution, children’s more sophisticated Type-II processes intervene and, for example, solve the problem with trait reasoning.
This account is compatible with existing data and gets more direct support from a recent study (Hermes et al., 2018b) designed to test it: children used a simple trust-the-better heuristic (Type-I) whenever it was applicable (when models differed in overall competence), but engaged in trait reasoning (Type-II) in problems where the heuristic was not applicable (when choosing between two experts) – yet only those children with sufficient executive function and background knowledge.

In this talk, I will introduce the dual-process account on selective trust and present two studies testing predictions of the account in novel ways, directly addressing the role of cognitive resources (study 1) and availability (study 2).
Since Type-II processes require cognitive capacities, Study 1 tests the prediction that adults usually engage in trait-based Type-II processes in selective trust decisions, but fall back on Type-I processes when cognitive resources are limited. Adults (N=200) performed selective trust tasks in which a simple heuristic and trait reasoning result in different solutions, with and without cognitive load/time pressure. The dual process account predicts that adults will generally choose models in line with their competencies, but fall back onto children’s performance under limited cognitive resources.
Study 2 (n=58 preschoolers, n=120 adults) tests a different kind of Type-I heuristic: availability. Participants chose in trait-related tasks between a model whose positive trait is more accessible to them (strong-and-shy) than the positive trait of a different model (weak-and-extroverted). We examine whether adults under cognitive load and children tend to answer heuristically, choosing the model inheriting the more accessible trait (strong/shy) for both, strength-related and extraversion-related tasks. Data collection is almost finished and the results will be presented in the symposium.

Authors

• Franziska Brugger, University of Göttingen
  Presenting Author

• Jonas Hermes, University of Goettingen
  Non-Presenting Author

• Tanya Behne, University of Goettingen
  Non-Presenting Author

• Hannes Rakoczy, University of Gottingen
  Non-Presenting Author