Saltar al contenido | Saltar al meú principal | Saltar a la secciones


Reasoning and problem solving

Sunday, October 02nd,   2011 [18:00 - 20:00]

PS_3.066 - Grasping isomorphism: review of Hinton’s “Learning distributed representations of concepts”

Varona-Moya, S. & Cobos, P. L.

Department of Basic Psychology, University of Málaga, Málaga, Spain

Multilayer perceptron networks’ ability to perform sensible inferences by analogy was pointed out by Hinton from the results of knowledge generalization tests between two isomorphic family trees. Due to its importance, a comprehensive review of this work was tackled to improve some methodological aspects in order to find statistically grounded answers to the questions posed by the author. Using the same network architecture and learning procedure, 500 simulations were trained in the task proposed by Hinton. In this review (1) the degree of isomorphism grasped by a simulation was computed through an ad hoc algorithm applied to principal component analysis scores of hidden units’ activation vectors and (2) the relationship between the network’s ability to perform inferences by analogy and its grasp of isomorphism was examined on a mixed factorial design basis, using corrected generalization tests. The main conclusions are these: (1) isomorphism grasp is not as consistent a property as Hinton suggested, since many simulations failed to build identical representations for both family trees, and (2) statistically significant interaction effects between isomorphism grasp and learning the second family tree proved that the more isomorphism a simulation grasped, the better it generalized from one tree to another.

PS_3.067 - Does the presentation frequency of obligatory and prohibitory traffic signs influence their information interpretation?

Vargas, C. , Moreno Rios, S. & Castro, C.

University of Granada. SPAIN

The frequency of a traffic sign could alter the difficulty of interpreting the situations referred to by the sign. Different traffic signs can be used at a T-junction to indicate that the road on the right is allowed: obligatory right, prohibitory left or both. Firstly, a right turn is referred and the “no left turn” is inferred. The opposite happens when signalling with a prohibitory left a right turn. This study test how we evaluate traffic scenarios under time pressure, where a car is shown in the situation signalled or in the inferred one. The frequency of presentation of prohibitory and obligatory signs was manipulated. When the signs were shown with a same frequency, responses to the referred situations were faster and more accurate, but the negative effect was found in the inferred situation. However, when the more frequent sign was shown, responses to the referred situations were faster and more accurate, and the negative effect was greater in the inferred situation. For example, when the obligatory signs were more frequent, the highest number of errors was found for the prohibitory left sign when a car was taking the right turn. Safety driving implications and suggestions will be further discussed.

PS_3.068 - Personality, individual difference & counterfactual thinking

Maloney, D.

Psychology department. Mary Immaculate College. Limerick, Ireland.

How we imagine what might have been, called counterfactual thinking, is influenced by several factors. Much of the existing literature has focused on the contextual or situational factors that goad people to think counterfactually. This research suggests that more extensively studied situational factors (e.g., how normal/controllable/mutable an event was) do not fully explain counterfactual thoughts, and investigates the role of personality and individual differences in how we construct alternative simulations of the past. Participants completed a series of difficult cognitive tasks (e.g. anagrams) and were then asked to think about how their performance on the tasks might have been better. Participants then repeated the cognitive tasks, and answered a battery of psychological personality measures. The results suggest that autonomy may be a particularly important personality trait in terms of counterfactual thinking and in how performance might improve as a result of considering alternative exemplars. Person’s high in autonomy generated significantly more self-regulating counterfactual thoughts than person’s low in autonomy and a greater total number of counterfactual thoughts than person’s low in autonomy. These results suggest that personality traits may influence both the activation and focus of counterfactual thinking. Results are discussed in relation to the functional theory of counterfactual thinking.

PS_3.069 - The influence of the analogical context on the similarity judgements of compared events

Oberholzer, N. 1, 2 , Trench, M. . 1, 2 & Minervino, R. A. 3, 2

1 Universidad de Buenos Aires
3 Universidad Nacional del Comahue

An experiment was carried out to demonstrate that judgments of analogical relatedness between two events vary as a function of the analogy in which they take part. In a control condition, participants used a 7-point Likert-scale to rate the analogical similarity of two events (e.g., Mary bought a play station vs. John borrowed a Wii). In the two experimental conditions, participants received these same events framed within base and target analogs aimed at rendering both events more or less similar. For instance, whereas in the pro-similarity condition the protagonists criticize each other’s actions as having broken their agreement to devote exclusively to their studies, in the less-similarity condition the protagonists criticize each other’s actions as having broken their agreement to cut unnecessary expenses. As predicted, similarity scores assigned by participants under the pro-similarity condition were much higher than scores given by participants under the less-similarity condition, with scores from the control condition falling somewhere in between. Data proved that the analogical similarity between two compared situations depends on the way in which they are related to the broader analogy in which they participate, thus posing a challenge to the role of semantics within traditional theoretical approaches of analogical mapping.

PS_3.070 - Cognitive processes in reasoning about moral dilemmas

Gubbins, E. 1, 2 & Byrne, R. 1, 2

1 School of Psychology, Trinity College Dublin, University of Dublin, Ireland
2 Institute of Neuroscience, Trinity College Dublin, University of Dublin, Ireland

We report the results of two experiments that examine cognitive processes in moral reasoning using different versions of the well-known ‘trolley’ problem. Participants were asked to imagine they were at the wheel of a runaway train approaching a fork in the tracks. They were asked to judge if they would allow the train to continue to the left, killing five railway workmen on that track; or hit a switch turning the train to the right, killing a single workman. In the first experiment, participants judged that they would push the switch for this standard version of the dilemma, more so than for a version with just a single workman on each track; but they judged they would look for an alternative solution in a version in which the train would reach the fork in thirty minutes. In the second experiment, participants judged they would push the switch in a version that described good qualities of the single workman, but not in a version that described the single workman as a close relative. We discuss implications of the results for alternative theories of moral reasoning.

PS_3.071 - The probability of events in a conditional: how this changes the way it is represented

Rodríguez-Gualda, I. & Moreno Ríos, S.

Facultad de Psicología. Universidad de Granada (Spain)

Does the probability of the events mentioned in conditionals influence the way the conditionals are represented? From the suppositional theory, the comprehension of a conditional depends on the perceived probability of occurrence of the consequent, given the antecedent. From the mental model theory, the comprehension of a conditional is based on the initial representation derived from it (not on the probability of the events). In the present study, the contents were constructed in order to create more probable or less probable situations based on the empirical experience. Thus, three kinds of content were used, depending on their frequency of occurrence in the world: very frequently, not very frequently or their occurrence is arbitrary in the world. Three tasks were used to evaluate how people understand semifactual (even-if) conditionals that connect two events (antecedent and consequent). Participants were tested with different measures: judgments about the probability of each event mentioned in the conditionals, the probability of the even-if statements and the degree of strength of a causal sentence. Results are shown and discussed from those theories.

PS_3.072 - Problem structure mediates the effects of numeracy in Bayesian reasoning

Johnson, E. 1 & Tubau, E. 1, 2

1 Departament de Psicologia Bàsica, Universitat de Barcelona, Barcelona, Spain
2 Institute for Brain, Cognition and Behavior, Universitat de Barcelona, Barcelona, Spain

Understanding probabilistic information is essential for success in school and in the job market, as well as for making good health and financial decisions. Despite its importance, probabilistic reasoning remains difficult even for mathematically informed university students. Recent studies have shown that individual differences in numeracy modulate the effects of numerical format in the context of risk perception, decision-making, and Bayesian reasoning. Inconsistent results have been found, however, in the direction of this interaction, with frequency formats sometimes benefiting those higher in numeracy and at other times facilitating performance for the less numerate. The present research addresses these inconsistencies through a series of Bayesian reasoning problems differing in numerical complexity, a standard measure of numeracy, and a number comparison task thought to index basic numerical representation. Extending previous studies using the same numeracy scale, results reveal an interaction between individual numeracy and problem structure: the effect of numeracy only clearly arises in problems involving more complex numerical relations. Preliminary results also indicate that higher-level probabilistic reasoning ability can be at least partially explained in terms of basic numerical representation. Accordingly, the current research provides a more complete account of the role of individual numerical ability in reasoning with probabilistic information.

©2010 BCBL. Basque Center on Cognition, Brain and Language. All rights reserved. Tel: +34 943 309 300 | Fax: +34 943 309 052