A COMPARISON OF TWO PROCEDURES FOR PROGRAMMING THE DIFFERENTIAL REINFORCEMENT OF OTHER BEHAVIORS

The relative effectiveness of two methods of programming DRO schedules of reinforcement was examined in two experiments. In these two methods, reinforcement is delivered if inappropriate responding is not occurring (a) at the end of an interval (momentary DRO), or (b) throughout the entire interval (whole-interval DRO). In Experiment 1, the effects of these schedules on disruptive responding of three retarded students were assessed in a multiple-baseline design. For two students, the momentary schedule occurred first and was ineffective, whereas the whole interval that followed was effective; for the third student, the whole-interval schedule occurred first and was effective, and reduced responding was maintained under the momentary schedule. In Experiment 2, baseline and the two DRO schedules were each presented in random order each day to one student in an alternating treatments design. The momentary DRO schedule reduced responding, but the whole-interval schedule was more effective.     

How one gambles if one must: Effects of differing return rates on multistage betting decisions

In the multistage betting game (MBG), a decision maker (DM) is provided with some capital x which he is required to bet over m (m > 1) mutually exclusive and collectively exhaustive alternatives, each of which occurs with probability p_i (p_i > 0, i = 1,…, m; Σ_{i = 1}^mp_i = 1). If y_i is bet on alternative i (y_i ≥ 0, Σ_{i = 1}^my_i = x) and alternative i obtains, the DM's capital for the next stage is y_ir_i, (r_i > 0). The MBG lasts until either the DM loses his capital or N stages elapse, whichever comes first. Each of six subjects participated in six sessions consisting of several hundred 3-alternative MBG stages. A within-subject design assigned negative expected value (EV) bets to the first three sessions and positive EV bets to three more sessions. Significant effects were found due to return rate, capital size, homogeneous runs of either wins or losses, and individual differences. Four maximization of expected utility and two minimization of risk models were presented and tested. A modified logarithmic utility model is proposed, which provides the best fit to the data. The implications of the results and directions for further research are briefly discussed.     

A factor analysis of the Stanford-Binet at four age levels

Standardization scores on the 1937 Revision of the Stanford-Binet are analyzed by the Thurstone centroid factor technique. Correlation matrices at age levels seven, nine, eleven, and thirteen, each based on two hundred subjects, are analyzed. The factors after rotation are interpreted in terms of the items exhibiting high factor saturations. In contrast to the findings of McNemar (1942) no evidence for a general factor in the Stanford-Binet battery is found in the present study. Rather, I.Q. variance can be explained in terms of psychologically meaningful group factors at the various age levels. The results indicate a need for cautious interpretation and use of single score indices of intelligence such as the Stanford-Binet I.Q.This study was accomplished in partial fulfillment of requirements for the Master of Science degree at the University of Washington, 1948. The author is greatly indebted to Dr. Lloyd G. Humphreys for suggesting the topic and for his invaluable guidance throughout the project. He is also appreciative of the cooperation of Dr. Quinn McNemar, who generously supplied the initial data and has offered helpful advice in the preparation of this paper.     

WHAT DOES A PIGEON SEE IN A PICASSO?

The present paper is primarily a response to Watanabe, Sakamoto, and Wakita's 1995 paper “Pigeons' Discrimination of Paintings by Monet and Picasso.” Our main criticism is that the key term concept, and in particular the way in which a class of stimuli is specified, remains equivocal. We argue that it is only useful to study categorization if the organism in question can be expected to have access to the information that is used in defining the classes, and that evidence for categorization requires selective responses that cannot be attributed to similarity in features that are not essential to the classification.     

Coherence as an ideal of rationality

Probabilistic coherence is not an absolute requirement of rationality; nevertheless, it is an ideal of rationality with substantive normative import. An idealized rational agent who avoided making implicit logical errors in forming his preferences would be coherent. In response to the challenge, recently made by epistemologists such as Foley and Plantinga, that appeals to ideal rationality render probabilism either irrelevant or implausible, I argue that idealized requirements can be normatively relevant even when the ideals are unattainable, so long as they define a structure that links imperfect and perfect rationality in a way that enables us to make sense of the notion of better approximations to the ideal. I then analyze the notion of approximation to the ideal of coherence by developing a generalized theory of belief functions that allows for incoherence, and showing how such belief functions can be ordered with regard to greater or lesser coherence.Many people influenced the present version of this essay. Ban van Frassen, Richard Jeffrey, David Lewis, Mike Thau, and Alan Hájek provided extensive and invaluable written comments on the entire essay. Mark van Roojen provided helpful comments on Sections 3 and 4. Mike Than and John Barker provided essential aid when I was formulating the proofs in section 6. Finally, I am grateful for valuable discussion of the essay with Ned Hall, Fiona Cowie, Jim Woodward, David Hilbert, and Frank Arntzenius.     

Typicality effects in logically defined categories

Subjects learned to categorize geometric designs by some rule combining two features, x and y. When xy stimuli (stimuli with both features) are always positive, the rule (concept) is inclusive disjunctive; when xy stimuli are always negative, the rule is exclusive disjunctive. In three other experimental conditions, the probability of xy stimuli being positive during acquisition was .25, .50, and .75. In a variety of postacquisition tests, xy stimuli were chosen as prototypical despite their less than consistent occurrence in the positive category. Stimuli with one relevant feature (x or y), although consistently assigned to the positive category, were often evaluated as poorer examples of the concept. These results are interpreted in terms of a schema model in which information acquired during learning is organized according to probability density functions over feature dimensions. This theory appears to be general enough to accommodate the evidence from laboratory studies of both logical and natural concepts.