Possibilities pop up: Preoperational Children’s Everyday Conditional Reasoning in Social-communicative Contexts

• In Event: Poster Session 06
  In Poster Session: PS 06 Topic - Cognitive Processes

Abstract

Reasoning deductively is a characteristic of human reason and a major goal of education (Aristotle, 2012; Markovits, 2004; Osborne & Reagan, 1973). Nonetheless, according to accumulated evidence in the past five decades, our everyday reasoning is not purely deductive. When making inferences, we tend to rely on the heuristic type 1 processing rather than the reflective type 2 processing (Evans, 2016, 2003; Kahneman, 2011; Markovits et al., 2017, 2013; Stanovich & West, 2000). As humans are social animals, our everyday reasoning takes place typically in social contexts (Borghi et al., 2019; Evans, 2007; Johnson-Laird et al., 1972; Oaksford & Chater, 2020; Rasga et al., 2016). However, how we use the heuristic processing to make deductive inferences in social contexts remains unclear.
To figure out this question, we manipulated the complexity of social cues and the level of social interaction by varying the way and the point of view for presenting premises, and accordingly set two contexts at different social-communicative levels. In the high-level social-communicative context (HSC), the premises were presented with an experimenter’s first-person performance, while in the low-level social-communicative context (LSC), the premises were presented with a third-person card story.
In study 1 & 2, for each context, 96 preoperational children (4–7 years of age), whose rule-based reflective processing was still absent (Barrouillet, 2011; Piaget, 1964, 1971), completed a conditional reasoning task comprising 4 types of inference: Modus Ponens (MP), Modus Tollens (MT), Affirming the Consequent (AC), and Denying the Antecedent (DA). In study 1, after controlling antecedent-consequent order, age, and fluid intelligence, we found children’s conclusion accuracy in HSC was higher for AC & DA, but lower for MP & MT than in LSC, while there was no difference in the logicality of their justifications between the two contexts (Figure 1A, 1B). In study 2, we further controlled the content association strength, conducted an independent replication of study 1 (Figure 1C, 1D), and thus exclude a non-reasoning alternative explanation.
In study 3, we recoded children’s inferential conclusion and justification as uncertain conclusion index (UCI) and possible justification index (PJI), which reflects the uncertainty in conclusions and the possibility representation in justifications, respectively. Data from study 1 & 2 combined, the main findings were repeated (Figure 2A, 2B). Moreover, we founded a partial mediation from the social-communicative level of reasoning context, through PJI, to UCI (indirect effect = 0.20，SE = 0.05，p < .001, 95% CI = [0.11, 0.31]; Figure 2C).
In conclusion, our findings suggest the potential mechanism that how social-communicative contexts activate the heuristic system in our everyday reasoning. In social-communicative contexts, given the complex social environment and high-level social interaction, we are inclined to represent more possibilities, thus perceive more uncertainty, and eventually make more intuitive reactions. These findings also have educational implications that practitioners should attach significance to the integration of teaching logical rules, training divergent thinking (Chantal et al., 2019), and providing social-communicative contexts as scaffoldings for fostering logical thinking.

Authors

• Meng-Han Cong, Peking University
  Presenting Author

• Yanjie Su, Peking University
  Non-Presenting Author