Determination of relationships between distributions of stimuli and distributions of judgments under instructions of differing specificity

Thumbnail Image



Journal Title

Journal ISSN

Volume Title

Repository Usage Stats



INTRODUCTION: Basic to the judging process is the relating of a given item to a group of items* The simpler case of judging is one in which an item is compared with another which is simultaneously present while the more complex case consists in comparing an item with previously experienced items* Analysis of the latter process was given impetus by Wever and Zener (8) who introduced a method of investigation applicable to this problem of judgment in time* Positing that even simple comparisons draw heavily upon an extended context of experiences, these investigators demonstrated that their method of presenting for judgment single members of a stimulus series gives data comparable to that obtained with the traditional method of constant stimuli. Wever and Zener and investigators who subsequently utilized the method of single stimuli have demonstrated that subjects are able to make consistent judgments which are sensitive to small increments of change in the stimulus series* Additional studies have investigated some of the influences that modify judgments such as changes in end stimuli or stimulus density, and aspects of the stimulus distribution to which judgments are anchored. Several reviews of the research in this area are available (5,6,7). In addition to laboratory findings everyday life offers many examples of the utilization of judgments which reflect previous experiences with the stimulus dimension involved. The basis for such characterizations as “a tall man”, “a fascinating lecture”, “ a good meal” is admittedly more involved than the basis for usual laboratory judgments but the same general principles may be assumed to underlie both. In both the laboratory and the social situation the process of relating one item to a non-present set of items is dependent upon a temporal integration of the effects of previous contacts with items of that set. It is meaningful, therefore, to examine the functional dependence of distributions of judgments upon previous experience with items of the same set as the ones being judged. This problem is implicit in several different lines of research such as investigations of shifts in judgments, where the underlying assumption is that changes in judgment reflect changes in the fundamental character of the stimulus distributions, and empirical studies of anchoring, which in general follow the pattern of modifying essentially rectangular stimulus distributions. Both types of investigation represent efforts to discover the aspects of a stimulus distribution to which judgments are related. The present study is composed of several experiments which 4 were designed to investigate systematically general relationships obtaining between different distributions of stimulus items and distributions of judgments elicited by these items with attention to such factors as differences in the instructions, the number of judgment categories and the step-interval between items. In all experiments the subjects were required to judge the length of singly presented horizontal lines. The first group of four experiments represents an effort to discover the form of the basic functional relationship in relatively unstructured situations which are representative of most judging tasks. The initial experiment consisted of separate groups of subjects judging one of five different distributions of stimulus items. All the distributions (rectangular, symmetrical unimodal, bimodal, positively skewed, negatively skewed) had the same range and density of items and two categories of judgment (longer or shorter) were available to the subjects. The second experiment was designed to investigate the influence of the factor of stimulus distribution on judgments rendered by subjects who experience successively more than a single stimulus distribution, since in life situations individuals do not typically experience one clearly defined distribution of similar stimulus items. Rather they have a variety of contacts with items whose distribution may vary over a period of time. The aspect of the judging situation which was altered in the third experiment was the number of judgment categories. In order to determine the effect of the distributional properties of the stimulus items on judgments in multiple category situations the number of categories available to the subjects was increased from two to three (longer, medium, shorter). In the fourth experiment the step interval between stimulus Items was increased from a barely supraliminal to a clearly discriminable one. This was done in order not to restrict the findings of the study to situations such as those of the traditional psychophysical experiments where the step-interval is in the region of the Ilmen. In the first four experiments the instructions to the subjects were very general, and thus the question is raised whether the relationships obtained under these conditions depend upon varying individual interpretations of the task. The last two experiments in this study were designed to investigate the effect of more explicit instructions with the aim of obtaining results which could be compared with the relationships found between distributions of stimuli and distributions of judgments in the more representative unstructured situations.



This thesis was digitized as part of a project begun in 2014 to increase the number of Duke psychology theses available online. The digitization project was spearheaded by Ciara Healy.



Published Version (Please cite this version)


Dukes student scholarship is made available to the public using a Creative Commons Attribution / Non-commercial / No derivative (CC-BY-NC-ND) license.