MINERVINO Ricardo Adrian
congresos y reuniones científicas
Analogical Retrieval Mediated by Everyday Schema-Governed Categories
OLGUÍN, MARIA VALERIA; LUCÍA MICAELA TAVERNINI; PACELLA, LAUTARO; MINERVINO, R. ADRIÁN
Congreso; Fourth International Conference on Analogy.; 2017
Institut du Cerveau et de la Moelle
Analogical Retrieval Mediated by Everyday Schema-Governed Categories Two situations are considered analogous if they can be assigned to a common schemagoverned category (SGC; Minervino, Oberholzer & Trench, 2013). For example, a situation where an ex-girlfriend scribbled her ex-boyfriend´s books could be considered analogous to another situation where an ex-girlfriend threw her ex-boyfriend´s cloths away because of being cases of revenge, despite the lack of similarity between corresponding relations and objects. Tavernini, Trench, Olguín and Minervino (2017) found that when two facts are framed by a SGC, they are perceived as more analogous to the extent that they exhibit close values along relevant SGC´ dimensions. For example, the first situation would be more analogous to the second than to a third situation where an ex-girlfriend burned her ex-boyfriend´s motorcycle, since the former ones are closer in severity. Laboratory studies show that retrieval based solely on structural similarity is infrequent unless an abstract encoding of the analogs is promoted (see Trench & Minervino, 2017). Several authors (e.g., Hofstadter & Sander, 2013) cast doubt on this superficial bias in simple dailylife analogies, in which we naturally tend to encode both analogs in terms of a common-sense schema. The first aim of the present study was to determine if in everyday analogies retrieval based only on structural features is more frequent than in laboratory studies. A second objective was to asses if the closeness of the analogs on their dimensional value along the schema to which they belong plays a role in schema-based retrieval. We employed ?me too? analogies (see Hofstadter & Sander, 2013), in which a person tells a story and a second one replies by reporting an analogous episode. Materials were eight sets of stimuli, each of which comprised a pair of target analogs. Both of them could be assigned to the same SGC (e.g., expression of love), but differed in their value along a relevant dimension of the SGC (e.g., magnitude). For example, the Target analog 1 was I painted a portrait for my girlfriend, a very big expression of love, and the Target 2 was I cut a rose for my girlfriend, a smaller expression of love. Forty psychology students received one target analog from each set as if it were told to them by someone else, and were asked to reply by recalling an analogous episode. Participants were then asked in which sense they considered that the given and the retrieved events were analogous. Participants´ responses were classified as reporting a SGC or not. Results showed that in 91% of the retrieved analogs belonged to the same SGC as the targets. A posterior analysis of the SGC analogies was performed by two independent judges, who had to rate separately on 5-point Likert scales (1 = highly dissimilar; 5 = highly similar) the similarity between base-target relations and base-target objects. The similarity of each pair of analogs was obtained averaging the two similarity scores given by each judge. Then, analogy ratings of both judges were averaged. Analogies yielding lower scores than 2 were sorted as "dissimilar". In 73% of the cases analogies were superficially dissimilar, which indicates that in everyday analogies recalls based on structural properties are more frequent than in laboratory studies. Two additional judges had to rate on 5-point Likert scales the value that each base analog exhibited along a pre-specified dimension of the corresponding SGC. Base analogs recalled in response to target analogs with high dimensional values received higher scores than those recalled in response to target analogs showing lower dimensional values, thus demonstrating an effect of dimensional values on analogical retrieval. The implications of our results for theories of analogical retrieval are discussed. References Hofstadter, D. R., & Sander, E. (2013). Surfaces and Essences: Analogy as the fuel and fire of Thinking. New York: Basic Books. Minervino, R., Oberholzer, N., & Trench, M. (2013). Global similarity overrides element similarity when evaluating the quality of analogies. Journal of Cognitive Science, 14, 287-317. Tavernini, L. M., Trench, M., Olguín, V., & Minervino R. A. (in press). Similarities between objects in analogies framed by schema-governed categories. Proceedings of the 39th Annual Conference of the Cognitive Science Society. Austin, TX: Cognitive Science Society. Trench, M., & Minervino, R.A. (2016). The surface bias in analogical retrieval. En B.H. Ross (Ed.), The Psychology of Learning and Motivation (Vol. 66, pp. 1-41). Cambridge, MA: Elsevier.