The research reported in this dissertation explores the coordination problem faced by economic agents in various strategic environments. The first chapter provides an experimental test of a theory of collusion in the presence of price-matching guarantees and thus throws light on the equilibrium selection problem embedded in this market game. The experiment yields important empirical information regarding the competitive nature of these low-price guarantees in the laboratory.In the second chapter, more general theoretical models are developed that undermine any collusive equilibrium in the presence of price-matching guarantees. Although the theory predicts that the competitive price should emerge in equilibrium in all these models, systematic discrepancy between the theoretical prediction and the observed behavior is found.In the third chapter, a well-known paradox is tested in the laboratory. Braess paradox (Braess, 1968) consists of showing that, in equilibrium, adding a new link that connects two routes running between a common origin and common destination may raise the travel cost for each network user. The experiment is designed to study whether the paradox is behaviorally realized in the critical minimal simulated traffic network. Results reject the hypothesis that the paradox is of marginal value and its force, if at all evident, diminishes with experience.In the last chapter, using controlled laboratory experiments, I study how the problem of coordination failure, as embedded in the 'minimum-effort' coordination game, can be overcome using structured, ex post feedbacks related to individual performance among members of a large group. I allow two types of performance feedback mechanisms, namely, negative and positive. I use 'disapproval' and 'approval' ratings about individual choices by group members as proxies for negative and positive feedback mechanisms, respectively. Results show that where participants are allowed to express only disapproval of others' choices, play converges towards the most efficient coordination. In contrast, where participants can express only approval of others' choices, inefficient coordination is obtained.

The main goal of this dissertation was to examine rhesus macaque and preschool child performance on invisible displacement tasks. In Study 1, rhesus monkeys were tested on a ramp task wherein an object rolled down a ramp behind four doors. An obstacle was placed along the ramp at one of four doors, and monkeys had to open the door nearest to the obstacle. Macaques were unable to locate the object until they were provided with cuing trials. In Study 2, macaques were tested on a vertical tube task. The tube had an occluding panel with three doors and an obstacle that was placed beneath one of the doors. Three of six monkeys were able to solve this task without any cuing. Subsequent studies manipulating the number of available doors identified the length of trajectory of the falling object as aiding performance. In Study 3, monkeys were tested on a trajectory task consisting of an apparatus with three tubes, each extending from the top of the apparatus into three widely spaced cups. The animals were shown a predetermined length of the object's trajectory (ranging from 15 cm to 35 cm) before it disappeared behind an occluding panel. Ten of eleven monkeys were able to solve the task. Young adults were initially affected by the trajectory length but were able to solve the task after minimal experience. In contrast, both elderly and juvenile macaques partially solved the task when given the maximum amount of information. Subsequent tests showed that performance of juvenile macaques was related to trajectory information whereas elderly macaques used a strategy of searching where the object was last seen. In the final four studies, we investigated the failure of preschool children to solve the ramp task. We conducted various manipulations to promote integration of the obstacle into the child's solution, none of which markedly improved performance. In conclusion, macaques appear to (1) have difficulty understanding the contact-mechanics between objects and obstacles and (2) use trajectory information to solve invisible displacement tasks. Children, however, appear unable to incorporate the obstacle into their solution even with the addition of cues.

Abstract not available

Language processing requires the organization of a complex physical pattern into a memory representation which captures the meaning relations intended in the utterance. Theories of language processing have focussed on the structure of syntax and semantics as the underlying organization of memory representations. Although research on prosodic information in spoken language has demonstrated that prosody can affect language processing, theories have usually treated prosodic information as a source of cues in developing a syntactic or semantic representation. This thesis proposes that prosody occupies a distinct level of representation in the language processing system. The organization of the prosodic representation determines the availability of information in memory as well as the processing units for higher levels of analysis. Two experiments are presented in support of this view. The first experiment compared listening comprehension time for sentences with good and poor prosodic information. The pattern of results was best explained by assuming an internal representation of the prosodic contours. The second experiment found similar effects for temporally organized sentences in reading. Two pilot experiments were conducted to test a specific representation model. These experiments served to evaluate some techniques for further research in auditory language processing.

The fundamental question to be answered in this dissertation is, What sort of structure controls the timing of an utterance during sentence production? Timing here refers to the ordering of elements (i.e., words) within an utterance, and the specification of the duration of each element and any pauses. The structure which controls timing is thus a plan or motor program for some segment of speech. The experimental paradigm used in this dissertation is a sentence repetition task in which subjects either read sentences aloud or repeat sentences temporarily held in a short-term memory buffer. Various measures such as word and pause duration as well as initiation time (time to begin saying an utterance) were examined. The first experiment demonstrates that a syntactic structure does not have the necessary properties to control utterance timing, contrary to claims made by Cooper and Paccia-Cooper (1980). Instead, it is necessary to posit a phonological level of representation constructed from the syntactic structure; it is a phonological structure which guides utterance timing. However, the first experiment also demonstrates that a phonological structure with the properties ascribed to it by Gee and Grosjean (1983) does not have the right properties to account for timing either. I present an alternative view of phonological structure, and a model of the way in which it is created from a syntactic structure. The most important characteristic of the model is that it separates planning-based acoustic effects from more rhythmically determined ones (timing-based effects). Creation of the phonological structure produces the planning effects; execution of the structure produces the rhythmic ones. The phonological structure has the characteristics of a metrical grid: Information concerning timing relations within an utterance are specified in a way that abstracts from the actual segmental content of the utterance.

Abstract not available

Despite assumptions about television addiction in the literature on television viewing, there have been no systematic attempts to document it. The present study was the first. A review of theories on addiction processes concluded that the theoretical consensus on addiction permits the possibility of the existence of this phenomenon in TV viewing. Also, the theories point to the effective domain as an area in which to expect differences between TV addicts and normal viewers. Both the scientific and popular conceptions of TV addiction were also reviewed. A content analysis of the popular literature provided specific guidelines for the operational definition of TV addiction used in this study. A television viewing questionnaire which included a self-rating scale of items derived from descriptions of addicts in the popular literature, as well as many other measures, was administered by mail to a random sample of residents of Springfield, Massachusetts, in the spring of 1982. The response rate was 65%. The results of confirmatory factor analyses failed to support the hypothesis that TV addiction exists as it is described in the popular literature. None of the respondents scored at or above the criterion for the existence of the addiction factor in their responses. An alternative two-factor model was better able to account for the item covariences. Descriptive statistics on the addiction scale items indicate that most respondents feel that TV-addictive behavior is completely uncharacteristic of them. Yet, 65% agreed that "Television is addictive" and eleven respondents called themselves addicts. It was observed that the average number of hours reported in viewing time for this group was over double that of the entire sample. Future analyses of this data will test alternative conceptions of television addiction. An attempt was made to account for the popularity of the notion of television as a "plug-in drug" in terms of (1) our sparse scientific knowledge about the nature of the viewing experience, and (2) an abiding fear or ambivalence in American culture about technology and its effects.

This paper describes an experimental study offered by Designs of Experiments (DOE) within the defined factor domains to evaluate the factor effects of simultaneous characteristics on the benchmarked hard disk drive performance by proposing well-organized statistical models for optimizations. The numerical relations of the obtained models permit to predict the behaviors of benchmarked disk performances as functions of significant factors to optimize relevant criteria based on the needs. The experimental data sets were validated to be in satisfying agreement with predicted values by analyzing the response surface plots, contour plots, model equations, and optimization plots. The adequacy of the model equations were verified effectively by a prior generation disk drive within the same model family. The retained solutions for potential industrializations were the concluded response surface models of benchmarked disk performance optimizations. The comprehensive benchmarked performance modeling procedure for hard disk drives not only saves experimental costs on physical modeling but also leads to hard-to-find quality improvement solutions to manufacturing decisions.

In this thesis, the current vs. magnetic field relationships for four magnets, M1, M2, M3, and Hyperon of FNAL Experiment E781 (the charm baryon study) havel,een studied on the basis of algorithms involving sign corrections pertaining to remedying incorrect FNAL EPICURE readout system's current and magnetic field values for E78 I's 4 magnets.After analyzing previous BSU personnel's work involving re-averaging of the current and magnetic field output values for E78 1, further research was carried out in this thesis activity regarding the correction of the algebraic signs of certain current and magnetic field values in the FNAL EPICURE read out values inputted into E781' data bank.Firstly, a code was developed to merge the four files of current and magnetic fields for the four magnets of E78 1, all into one file. Then, an algorithm for a sign correction program was developed in order to both sort the merged data and correct the current and magnetic field signautomatically.In order to obtain a better understanding of the effect of the incorrect sign for a magnet's current and magnetic field on analysis results, a statistical result was summarized. Finally, some conclusions were obtained. / Department of Physics and Astronomy

Typescript (photocopy). / Digitized by Kansas Correctional Industries