Doctor of Philosophy / Department of Psychology / James C. Shanteau / Unstructured decision making is a dynamic process where an individual must create an alternative because one is not available or provided. In this type of a decision, an individual may not have formed preferences or may not know the path to arrive at a solution. As opposed to selecting from existing alternatives, little research examines when decision makers create an alternative. Electronic commerce websites allow individuals to create a product by customizing it. A web-based simulation called Interactive Choice was developed for the investigation. It is an interactive naturalistic decision space permitting experimental controls such as random placement of participants into conditions and random display of stimuli. Participants customized three products (pizza, cell phones, shoes). Building on theoretical foundations of unfolding model and Image Theory, a model asserts the presentation of the information and preparation of the decision maker influences a decision maker. A phased examination explores decision makers’ cognitive processes by measuring participants’ evaluations of the product created and the process to create it. In the first phase, three experiments find, contrary to previous independent investigations, participants rarely retain a pre-selected default value. Logistic regression reveals that the odds ratio of predicting default retention is dependent on product type. In the second phase, results identify that problem solving instructions influence decision making. Analyses of multidimensional scaling and cluster analysis reveal patterns for default retention and problem solving instructions that define an electronic decision aid called Choice Builder. The dissertation suggests that when an individual creates a product, he or she has more control over the process that subsequently reduces the influence of the default. A new theoretical foundation is proposed identifying that for unstructured decisions individuals construct both decision strategies and preferences while creating an alternative. With an active process of acquiring and evaluating information, an individual forms a decision strategy and updates preferences to achieve an ideal outcome. This dissertation makes four contributions that include 1) a research tool, Interactive Choice, for exploration, 2) the identification of cognitive processes involved, 3) a proposal of a new theoretical approach, and 4) an electronic decision aid, Choice Builder.

Doctor of Philosophy / Department of Psychology / John J. Uhlarik / Usability practitioners need effective usability assessment techniques in order to facilitate development of usable consumer products. Many usability evaluation methods have been promoted as the ideal. Few, however, fulfill expectations concerning effectiveness. Additionally, lack of empirical data forces usability practitioners to rely on personal judgments and/or anecdotal statements when deciding upon which usability method best suits their needs. Therefore the present study had two principal objectives: (1) to validate a hybrid usability technique that identifies important and ignores inconsequential usability problems, and (2) to provide empirical performance data for several usability protocols on a variety of contemporary comparative metrics. The User-Centered Hybrid Assessment Tool (User-CHAT) was developed to maximize efficient diagnosis of usability issues from a behaviorally-based perspective while minimizing time and resource limitations typically associated with usability assessment environments. Several characteristics of user-testing, the heuristic evaluation, and the cognitive walkthrough were combined to create the User-CHAT. Prior research has demonstrated that the User-CHAT supports an evaluation within 3-4 hrs, can be used by individuals with limited human factors / usability background, and requires little training to be used competently, even for complex systems. A state-of-the-art suite of avionics displays and a series of benchmark tasks provided the context where the User-CHAT’s performance was measured relative to its parent usability methods. Two techniques generated comparison lists of usability problems – user-testing data and various inclusion criteria for usability problems identified by the User-CHAT, heuristic evaluation, and cognitive walkthrough. Overall the results demonstrated that the User-CHAT attained higher effectiveness scores than the heuristic evaluation and cognitive walkthrough, suggesting that it helped evaluators identify many usability problems that actually impact users, i.e., higher thoroughness, while attenuating time and effort on issues that were not important, i.e., higher validity. Furthermore, the User-CHAT had the greatest proportion of usability problems that were rated as serious, i.e., usability issues that hinder performance and compromise safety. The User-CHAT’s performance suggests that it is an appropriate usability technique to implement into the product development lifecycle. Limitations and future research directions are discussed.

Doctor of Philosophy / Department of Psychology / James C. Shanteau / A training protocol was developed to teach an implicit reasoning strategy to encourage the consideration of alternatives, specifically in behavioral trap decision environments. Engaging the strategy would thereby decrease the effect of focusing on traps, resulting in more rational behavior. In two studies, training was delivered in an instructor-less environment using paper-pencil and multimedia examples. The main training components consisted of analogical problem solving and counterfactual thinking. The potential moderators between training and performance outcomes consisted of an information processing disposition Need for Cognitive Closure, an individualized approach to decisions, Decision-Making Style, and a capacity to process information Working Memory Capacity. Arousal and mood were also measured before, during, and after the training as both have been linked with learning. In Study 1, participants engaged in analogical problem solving, additive counterfactual thinking, subtractive counterfactual thinking, or none of these (i.e., control group). Results revealed that the training was minimally effective, although some comparisons revealed a large shift from pre- to post-training in commitment score away from trap options. Likewise, the Need for Cognitive Closure was the best predictor of decision behavior revealing that a predisposition for amount of information processed during decision making is indicative of behavioral outcomes in this decision environment. Based on results from Study 1, the training was reformatted in Study 2 to obtain the maximum potential benefit. Analogical problem solving was coupled with each form of counterfactual thinking so participants engaged in both critical thinking processes. When training was effective, the two forms were differentially effective as related to behavioral trap problem type. Forward-looking training assisted problem types that force explicit cost recognition and immediate decision outcomes. Past-looking training assisted problem types that force little cost recognition and delayed decision outcomes. Results of this project could be used to enhance the acquisition of critical thinking as well as improve educational practices. Both information processing disposition and decision approach style predicted learning whereas capacity to process information and training manipulations did not. Future projects will examine how long the training effects last and if critical thinking training can be successfully applied to other decision environments.

Master of Science / Department of Psychology / Jerome Frieman / Male and female Phodopus campbelli were investigated for kin recognition abilities, assessed by kin discrimination of bedding, urine, fecal, or midventral glandular secretion odors. Using a habituation-discrimination paradigm, subjects were presented either the odors of a pair of their own brothers (C[subscript 1]) or a pair of brothers that were unrelated to them (C[subscript 2]). All subjects were then presented with a final test discrimination odor from a male donor unrelated to themselves or the first donors. All subjects differentiated between the bedding and glandular odors of two unrelated males, regardless of condition assignment. Subjects did not statistically differentiate between the odors of two brothers. These results suggest Phodopus campbelli are capable of recognizing kin based on bedding and midventral gland odors.

Doctor of Philosophy / Department of Psychology / Stephen W. Kiefer / Exposing rats to differential rearing conditions, during early post-weaning development, has been shown to produce changes in a number of behaviors displayed during adulthood. The purpose of the current study was to investigate whether rearing alcohol-preferring (P) and non-preferring (NP) rats in an environmental enrichment condition (EC), a social condition (SC), or an impoverished condition (IC) would differentially affect the consumption of and operant responding for 10% ethanol. In Experiment 1 rats were tested for both limited access and free access (two bottle choice between water and ethanol) consumption of 10% ethanol. For, Experiment 2 rats were trained to respond in an operant chamber for ethanol and then provided concurrent access to 10% ethanol (right lever) and water (left lever). After concurrent access, rats were required to respond over a gradually increasing fixed-ratio schedule for 10% ethanol and finally a progressive ratio schedule for 10% ethanol, 15% ethanol, and 10% sucrose. For Experiment 3 rats were trained to respond for 10% sucrose and then assessed for the maintenance of operant responding for 10% sucrose. The data from this series of experiments shows that EC P rats consumed, responded for, and preferred 10% ethanol significantly less than their IC P counterparts. Also, EC P rats did not significantly differ from NP rats during any aspect of testing for all experiments. Experiment 3 failed to reveal a significant effect of rearing although there was a line effect that has been previously observed in the literature. Thus, it would appear from these results that rearing in an EC condition acts to protect alcohol-preferring rats from increased levels of consumption of, preference for, and responding for ethanol compared to rearing in an impoverished environment.