Understanding the What and When of Analogical Reasoning Across Analogy Formats: An Eye-Tracking and Machine Learning Approach.

Starting with the hypothesis that analogical reasoning consists of a search of semantic space, we used eye-tracking to study the time course of information integration in adults in various formats of analogies. The two main questions we asked were whether adults would follow the same search strategies for different types of analogical problems and levels of complexity and how they would adapt their search to the difficulty of the task. We compared these results to predictions from the literature. Machine learning techniques, in particular support vector machines (SVMs), processed the data to find out which sets of transitions best predicted the output of a trial (error or correct) or the type of analogy (simple or complex). Results revealed common search patterns, but with local adaptations to the specifics of each type of problem, both in terms of looking-time durations and the number and types of saccades. In general, participants organized their search around source-domain relations that they generalized to the target domain. However, somewhat surprisingly, over the course of the entire trial, their search included, not only semantically related distractors, but also unrelated distractors, depending on the difficulty of the trial. An SVM analysis revealed which types of transitions are able to discriminate between analogy tasks. We discuss these results in light of existing models of analogical reasoning.

Children's Failure in Analogical Reasoning Tasks: A Problem of Focus of Attention and Information Integration?

Children's improved performance with age in analogy tasks has been explained by an increase in semantic knowledge of the items and the relations between them or by the development of an increased ability to inhibit irrelevant information. We tested the so-called "unbalanced attentional focus hypothesis" that claims that a failure to choose the "analogical" match can be the result of a difficulty to focus on all the relevant information available. Previous eye-tracking research has suggested, in analogies of the A:B::C:D format, that 5-6 year-olds organize their search around the C item. They focused significantly less than adults on the A:B pair, thereby hindering their discovering the relation(s) between A and B. We hypothesized that inducing them to focus their attention on the A:B pair at the beginning of the trial would affect their performance. In Experiment 1, increasing children's focus on the A:B pair did, indeed, lead to better performance. In contrast, in Experiment 2, focusing their attention on the A:B pair impaired performance when the most salient relation holding between A and B was, in fact, irrelevant for the analogy. By contrast, the obvious-but-irrelevant relation in the A:B pair had no negative effect on performance when no explicit A:B focusing was induced. These results are discussed in terms of the temporal organization of the task and availability of information, and of children's difficulties to disengage from the main goal of the task, when necessary.

An evaluation of scanpath-comparison and machine-learning classification algorithms used to study the dynamics of analogy making.

In recent years, eyetracking has begun to be used to study the dynamics of analogy making. Numerous scanpath-comparison algorithms and machine-learning techniques are available that can be applied to the raw eyetracking data. We show how scanpath-comparison algorithms, combined with multidimensional scaling and a classification algorithm, can be used to resolve an outstanding question in analogy making-namely, whether or not children's and adults' strategies in solving analogy problems are different. (They are.) We show which of these scanpath-comparison algorithms is best suited to the kinds of analogy problems that have formed the basis of much analogy-making research over the years. Furthermore, we use machine-learning classification algorithms to examine the item-to-item saccade vectors making up these scanpaths. We show which of these algorithms best predicts, from very early on in a trial, on the basis of the frequency of various item-to-item saccades, whether a child or an adult is doing the problem. This type of analysis can also be used to predict, on the basis of the item-to-item saccade dynamics in the first third of a trial, whether or not a problem will be solved correctly.

Brown lemurs (Eulemur fulvus) can master the qualitative version of the reverse-reward contingency.

Behavioral flexibility that requires behavioral inhibition has important fitness consequences. One task commonly used to assess behavioral inhibition is the reverse-reward task in which the subject is rewarded by the non selected items. Lemurs were tested for their ability to solve the qualitative version of the reverse-reward task with the choice between identical quantities of different food items instead of different quantities of the same food. Two of four subjects mastered the task without a correction procedure and were able to generalize the acquired rule to novel combinations of food. One of the two subjects competent on the quality version of the task could transfer this ability to different quantities of the same food. Our results are compared to lemurs' performances when tested under the quantitative version in a previous study and those of capuchin monkeys tested under a similar paradigm. The whole results suggest that the qualitative version of the reverse-reward task may be easier to master than its quantitative counterpart and that lemurs perform better than capuchin monkeys as they were able to later transfer the learning rule to the quantitative version of the task.