Causal Models Interact with Structure Mapping to Guide Analogical Inference (original) (raw)

Causal models guide analogical inference.

Computational models of analogical inference have assumed that inferences about the target are generated solely by "copying with substitution and generation" from the source, guided by a mapping based on similarity and parallel structure. In contrast, work in philosophy of science has stressed that analogical inference is based on causal models of the source and target. In two experiments, we showed that reducing analogical overlap by eliminating a matching higherorder relation (a preventive cause) from the target increased inductive strength even though it decreased similarity of the analogs. Analogical inference appears to be mediated by building and then "running" a causal model.

The Role of Causal Models in Analogical Inference.

Computational models of analogy have assumed that the strength of an inductive inference about the target is based directly on similarity of the analogs and in particular on shared higher order relations. In contrast, work in philosophy of science suggests that analogical inference is also guided by causal models of the source and target. In 3 experiments, the authors explored the possibility that people may use causal models to assess the strength of analogical inferences. Experiments 1-2 showed that reducing analogical overlap by eliminating a shared causal relation (a preventive cause present in the source) from the target increased inductive strength even though it decreased similarity of the analogs. These findings were extended in Experiment 3 to cross-domain analogical inferences based on correspondences between higher order causal relations. Analogical inference appears to be mediated by building and then running a causal model. The implications of the present findings for theories of both analogy and causal inference are discussed.

Integrating analogical inference with Bayesian causal models

2009

ABSTRACT We present an integration of analogical transfer with Bayesian causal models, focusing on distinctions between causes versus effects, generative versus preventive causes, and causal predictions versus attributions. When drawing analogical inferences, people can use the source to build a causal model for the target, which then provides a basis for making causal inferences about the target. We report experiments in which people estimated the probabilities of various causal predictions and attributions.

Inference processes in causal analogies

Proceedings of the second international conference …, 2009

In recent papers, Lee & Holyoak (2007, 2008a, 2008b) argue that extant models of analogy fail to explain how people draw inferences from causal analogies. In the current research, we argue that Structure-mapping Theory sufficiently explains the analogical inferences drawn from ...

Transfer of Structure-Related and Arbitrary Information in Analogical Reasoning

The Psychological Record, 2001

Analogies can aid learners in understanding a new domain, yet misunderstandings may occur if they are applied too broadly. The present studies examined transfer of two types of information. Participants read analogical source and target stories. The source stories in Experiments 1-3 included two additional sentences that could be transferred to the target. One of the sentences was related to the analogical structure, while the other was more arbitrary. Participants transferred the structure-related information significantly more often than the arbitrary information both when retrieving source stories from memory (Experiment 1) and when having access to them (Experiment 2). Participants in Experiment 3 were explicitly encouraged to consider both types of information for transfer. Results showed the structure-related information was selected as the appropriate transfer sentence. Experiment 4 examined the possibility that reading both types of information in the source stories influenced transfer rates. Some participants received stories with both the structure-related and arbitrary information while others received stories with only one type of information. Again, participants transferred the structure-related information to a greater extent than the arbitrary information. Furthermore, no differences in transfer were found between participants who received both types of information in the source domain versus those who received only one type of information. Overall , the results of the studies provide evidence that learners will preferentially transfer information related to the shared analogical structure. Experiment 1 was conducted as part of a dissertation submitted in partial fulfillment of the requirements for the PhD degree at the University of Massachusetts at Amherst, 1995. I thank my dissertation committee: Marvin Daehler, Chair, Arnold Well, Carole Seal, and John Clement. Experiments 2 and 3 were presented at the annual meeting of the American Psychological Society, June, 2000 in Miami , Florida. I thank Catherine Clement, Marvin Daehler, and Jennifer Yanowitz for their helpful and extensive comments on earlier versions of this article. Special thanks also to Catherine Clement for generously providing copies of materials she generated for her research .

Constraints on Analogical Mapping: A Comparison of Three Models

Cognitive Science, 1994

Three theories of analogy have been proposed which are supported by computational models and data from experiments on human analogical abilities. In this paper, we show how these theories can be unified within a common metatheoretical framework which distinguishes between levels of informational, behavioural and hardware constraints. This framework makes clear the distinctions between three computational models in the literature (the Analogical Constraint Mapping Engine, the Structure-Mapping Engine and the Incremental Analogy Machine) . The paper then goes on to develop a methodology for the comparative testing of these models. In two different manipulations of an analogical-mapping task we compare the results of computational experiments with these models against the results of psychological experiments. In the first experiment, we show that increasing the number of similar elements in two analogical domains, decreases the response time taken to reach the correct mapping

Structural Constraints and Real-World Plausibility in Analogical Inference

Theoretical accounts of analogy have largely agreed that structural constraints play a substantial role in the mapping process. Less is known, however, about the robustness of these constraints in the inference process and the way in which particular content influences the use of structural constraints in analogical inference. We conducted two studies testing whether the plausibility (or implausibility) of an inference influences adherence to general structural principles in analogical reasoning. We found substantial reliance on the predicted structural constraints, but also an influence of the plausibility of the inference.