Causal thinkingPresents the causal-model approach to causal reasoning and learning. The causal-model approach is introduced as a theory which states that humans have the tendency to assume the existence of deep causal relations behind the surface, and is contrasted with traditional associationist theories. Several causal models are introduced which differ in structural aspects. Empirical studies conducted with humans and nonhuman animals testing the causal-model theory are presented which demonstrate that individuals do not just rely on covariational information in causal learning and reasoning, but instead infer a deeper causal structure. Furthermore, the studies also show that people are sensitive to the structural aspects of causal models, and coordination of identical learning input with different causal structures was observed. Moreover, it is stated that the strengths of individual links within causal models need to be learned and that causal strength does not necessarily correspond to observed covariation. Findings on limitations of causal reasoning suggest that causal models may overestimate the abilities of humans and nonhumans. Suggestions for further research are discussed.https://www.psych.uni-goettingen.de/de/cognition/publications/waldmann2010https://www.psych.uni-goettingen.de/@@site-logo/university-of-goettingen-logo.svg
and Michael Waldmann (2010)
Causal thinking
In: False, ed. Glatzeder, Britt, Goel, Vinod, von Müller, Albrecht (Eds.). Springer
Presents the causal-model approach to causal reasoning and learning. The causal-model approach is introduced as a theory which states that humans have the tendency to assume the existence of deep causal relations behind the surface, and is contrasted with traditional associationist theories. Several causal models are introduced which differ in structural aspects. Empirical studies conducted with humans and nonhuman animals testing the causal-model theory are presented which demonstrate that individuals do not just rely on covariational information in causal learning and reasoning, but instead infer a deeper causal structure. Furthermore, the studies also show that people are sensitive to the structural aspects of causal models, and coordination of identical learning input with different causal structures was observed. Moreover, it is stated that the strengths of individual links within causal models need to be learned and that causal strength does not necessarily correspond to observed covariation. Findings on limitations of causal reasoning suggest that causal models may overestimate the abilities of humans and nonhumans. Suggestions for further research are discussed.
