Comparative Decision Making Studies

College of Agriculture

Department of Agricultural Economics

Helen Pushkarskaya

Decision Making and Strategic Interaction Under Different Types of Environmental Uncertainty

In collaboration with Dr. Michael Smithson (Australian National University), Dr. Xun Liu (Mount Sinai School of Medicine), & Dr. Jane Joseph (Medical University of South Carolina); 2006-present.
More

The goal of this study was to investigate differences in individual choices and strategic interaction behavioral under four types of uncertain environments: risk, ambiguity, ignorance, and conflict. In particular, we question whether preferences toward different types of uncertainty (risk, ambiguity, conflict, and sample space ignorance) correlate (which might suggest the common psychological bases, i.e. recognition of computational limitations), or not (which might suggest that they have distinct psychological bases). We also investigate the gender difference in decision making and strategic interaction under different types of uncertainty.

Insurance Pricing Under Risk, Ambiguity, Conflict and Sample Space Ignorance

In collaboration with Dr. Jerry Skees (University of Kentucky) and Dr. Laure Cabantous (University of Nottingham); 2008 - present.
More

The project investigates how experienced professional actuaries from USA, UK and some other countries price insurance premiums under different types of uncertain environments (risk, ambiguity, conflict and sample space ignorance). The methodology builds on the Kunreuther and Hogath, (1992) and Cabantous (2006) papers, but we make several important modifications. First, to the used before levels of uncertainty-- risk and ambiguity (Kunreuther and Hogath, 1992), and conflict (Cabantous, 2006) -- we add sample space ignorance. Second (similarly to Cabantous, 2006), we ascertain whether actuaries recommend to offer insurance in this specific environment. Finally, keeping in mind a comparative ignorance hypothesis (Fox and Tversky, 1995), we present each subject only with one scenario.

Domain Specific Risk Taking or Tollerance Toward Different Types of Uncertainty?

2010 – present.
More

Using fMRI data collected while subjects played novel uncertainty-specific gambles I test the hypothesis that domain specific risk taking (Weber, et al, 2002) relates to different types of uncertainty (risk, ambiguity, conflict and ignorance tolerance). Preliminary results suggest that risk taking in different domains was (at least partially) associated with activation in the regions selectively or differentially associated with different uncertain environments. This might suggest that different risk domains might in fact depict different uncertain environments, or different combinations of different uncertain environment.

Department of Animal and Food Sciences

Jeffrey Bewley

Quantification of Physiological and Behavioral Indicators Of Disease In Dairy Cattle Using Precision Dairy Farming Technologies

In Collaboration with: Amanda Sterrett (MS Student, University of Kentucky), Dr. Michelle Arnold (University of Kentucky), Dr. Bill Silvia (University of Kentucky), Dr. Earl Aalseth (Aalseth Dairy Consulting), Dr. Mary Rossano (University of Kentucky), Dr. Eric VanZant (University of Kentucky) 2011 – present.

Impact of Animal Position Within Dairy Compost Bedded Pack Barns On Barn Performance

In Collaboration with: Randi Black (MS Student, University of Kentucky), Dr. Joseph Taraba (University of Kentucky), Dr. Tim Stombaugh (University of Kentucky), Dr. Eric VanZant (University of Kentucky); 2010-present

Relationships among management practices, barn design, temperature, moisture, bacterial counts, and animal hygiene in compost bedded pack barns

In Collaboration with: Randi Black (MS Student, University of Kentucky), Dr. Joseph Taraba (University of Kentucky), Dr. George Day (University of Kentucky), Flavio Damasceno (University of Kentucky); 2010-present

Potential for estrus detection in dairy cattle using reticular temperature monitors

In Collaboration with: Andy Smith (MS Student, University of Kentucky, Dr. Bill Silvia (University of Kentucky); 2009-present

College of Arts & Sciences

Department of Biology

Philip Crowley

Sex Allocation and Pollen Limitation in Outcrossing Hermaphroditic Plants

In collaboration with Prof. Evelyn Korn (U. Marburg, Germany), Dr. Nils Anthes (U. Tuebingen, Germany) and Dr. Andrea Case (Kent State U.)
More

We are using game theory to understand optimal sex allocation as it depends on population size and pollen limitation. It appears that a mix of male and female function (hermaphroditism with identical allocation among individuals) is optimal in the absence of pollen limitation, but limited access to pollen can make functionally separate sexes the more effective strategy. We are developing the modeling while looking for empirical consistency in the literature, with the potential to pursue empirical approaches in the future.

The Size-Number Trade-Off in Polyembryonic Parasitoid Wasps

In collaboration with Dr. Paul Ode (Colorado State U.), PhD student Yoriko Saeki, and MS student Elizabeth Dusing
More

We have shown that clonal wasp broods of males and of females balance the trade-off between the size of offspring in the brood and the number of individuals in the brood differently. We have identified some mechanisms involved and are planning empirical studies in the greenhouse. Our models make predictions about relationships that determine fitness in nature; our goal is to continue the iterative process of model development and empirical testing with this powerful research system

Cooperation and Conflict in Hermaphroditic Seabass Mating Systems

In collaboration with PhD student Mary Hart
More

We have used a combination of game-theoretic models and empirical studies by SCUBA in Panama to measure sex allocation in relation to population density and predation risk. We have a model that demonstrates how the intensity of sexual conflict is expected to change along gradients of these two environmental variables. Pairs remain together for alternating fertilizations in male and female roles—both within 2-hr mating episodes and between days over weeks—despite repeated interventions by streaker individuals acting as males. The models predict some of these interesting behaviors.

David Westneat

The role of personality and flexibility in stochastic variation in behavior

In collaboration with Dr. Kurt Viele (Statistics, University of Kentucky) and Professor Jon Wright (Norwegian Institute of Science and Technology).
More

We are investigating the way in which different individual house sparrows express differences in parenting behavior. We already know individuals differ in the mean level of care they provide offspring (a parenting personality). All individuals also adjust their care to changing conditions, and it appears that individuals adjust to the same conditions in different ways. But there is variation as well in the underlying distributions of insects that parents find to feed their offspring, and we are investigating if stochasticity (uncertainty) in whether a parent can find a good food item affects their behavior. Of particular interest is whether parent birds differ in their tendency to search for risky to find items, if they adaptively shift their use of risky items (i.e., they manage the risk), and if individuals differ in their tendencies to shift from risky to less risky items as conditions change. We will then be interested in the impact of this on offspring, and the genetic or developmental conditions that influence these patterns of behavior.

Gatton College of Business and Economics

School of Management / Decision Science and Information Systems (DSIS) Area

De Liu

Interactive Digital Entertainment: Interpersonal Interaction and Its Implications

in collaboration with Dr. Xun Li (Nichols State U), Milton Shen (University of Kentucky), and Dr. Radhika Santhanam (University of Kentucky); 2005-present
More

Competition is a major captivating element of games, particularly in online games where players compete and challenge one another. In this project we examine the effect of competition on player effort levels, enjoyment, and performances and draw implications for online game designs and e-training design.

Competing Keyword Auctions

in collaboration with Dr. Jianqing Chen (University of Calgary) & Dr. Andrew B. Whinston (University of Texas at Austin); 2007-present.
More

This project examines the optimal bidding strategies for advertisers when they can choose from two keyword auctions that rank bidders differently.

The Time-Contingent Value of Social Capital: Structural Holes, Resources Richness, and Knowledge Production.

in collaboration with Dr. Juan Ling (Georgia College & State University, Management), Dr. Ajay Mehra (University of Kentucky, Management), Dr. Daniel Brass (University of Kentucky, Management), and Dr. Steve Borgatti (University of Kentucky, Management); 2007-present
More

The objective of our study was to build and test arguments about how social resources and social network structure explain subsequent knowledge production over time. We tested our ideas using data on co-authorships among 5286 authors in elite management journals (1956-2006).

Markets, Hierarchies, and the Coordination of Knowledge.

in collaboration with Dr. Bruce Skaggs (University of Massachusetts Amherst, Management), Dr. Alfred Boccia (Western New England College, Management), and Peter Mills (University of Oregon, Management); 2008-present
More

Through simulation modeling, the present paper represents a first attempt at understanding the relative impact of internal market on the coordination of knowledge. Our results show that while the internal markets can improve the efficacy of knowledge use by firms and resolve issues of appropriability by knowledge workers, they do so at the expense of firm profitability.

The Design and Implementation of Optimal Keyword Auctions.

in collaboration with Jun Li (University of International Business and Economics, Economics) and Dr. Shulin Liu (University of International Business and Economics, Economics); 2009-present
More

This paper characterizes the optimal keyword auction mechanism under the premise that it is costly for search engine serve additional advertisements. We obtain the explicit optimal allocation and payment rules for two special settings in which the optimal mechanism is characterized by simple scoring rules.

Information Asymmetry and Payment Schemes in Online Advertising.

in collaboration with Dr. Siva Viswanathan (University of Maryland); 2008 – present.
More

This project examines the role of information asymmetry between advertisers and publishers in determining the optimal payment scheme choices for publishers. We highlight the role of payment schemes as a means of leveraging private information available to publishers and advertisers. We identify the conditions under pay-per-impression or pay-for-performance schemes should be adopted. Our results provide insights into a number of commonly observed publisher strategies.   http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1698524.

Radhika Sannthanam

Emergency Management Information Systems: Could Decision Makers be Supported in Choosing Display Formats?

In collaboration with Milton Shen (University of Kentucky), Kyle Bailey (Lexmark International) and Dr. Melody Carswell (University of Kentucky); 2009-present .
More

This project investigates the design and use of sophisticated three-dimensional display formats in emergency management information systems (EMIS), a decision-support system that facilitates decision making in crisis situations. We conduct two experiments and find that, as prior research has suggested, decision makers, in crisis situations, do not choose the most appropriate format, but their performance improves with training developed on a decisional guidance framework. We discuss implications of these findings for EMIS design, for the training of emergency management professionals, and for future research on display formats and decisional guidance.

College of Education

Special Education and Rehabilition Counseling

Malachy Bishop

Rheumatoid Arthritis Project

in collaboration with Dr. Judy Goldsmith (Computer Science), Dr. Kristine Lohr (Internal Medicine), Dr. Jamie Studts (Behavioral Science), Dr. Nancye McCrary (Education Curriculum and Instruction), Radu Paul Mihail (Computer Science Ph D Graduate Student)
More

Rheumatoid arthritis (RA) is a serious disease, both painful and life-altering. Sufferers often find themselves unable to perform basic tasks, including dressing themselves, opening containers, using silverware, rising from a chair. There are medications available that, in many cases with high probability can enable the RA patient to accomplish such tasks. However, these medications often have low-probability but scary side effects. Patients often choose to not take drugs, because of highly unlikely scenarios. Although clinicians tell patents the probabilities of side effects, it seems that the patients do not grasp the unlikelihood of those possibilities. We propose a decision support system to help them understand probabilities of effects and side effects of RA drugs by "experiencing" them in a computer-game-like setting. The proposed system will allow patients to view and manipulate an avatar that suffers a similar level of RA, and can choose to be treated with drugs the patient is considering. The system will run simulations of every-day tasks, showing multiple windows with varying levels of success. The levels will be determined by the patient's disease level, the choice of treatment (for the avatar), and the known probabilities of efficacy and side effects.

College of Engineering

Department of Computer Science

Judy Goldsmith

Rheumatoid Arthritis Project

in collaboration with Dr. Kristine Lohr (Internal Medicine), Dr. Malachy Bishop (Special Education ), Dr. Jamie Studts (Behavioral Science), Dr. Nancye McCrary (Education Curriculum and Instruction), Radu Paul Mihail (Computer Science Ph D Graduate Student)
More

Rheumatoid arthritis (RA) is a serious disease, both painful and life-altering. Sufferers often find themselves unable to perform basic tasks, including dressing themselves, opening containers, using silverware, rising from a chair. There are medications available that, in many cases with high probability can enable the RA patient to accomplish such tasks. However, these medications often have low-probability but scary side effects. Patients often choose to not take drugs, because of highly unlikely scenarios. Although clinicians tell patents the probabilities of side effects, it seems that the patients do not grasp the unlikelihood of those possibilities. We propose a decision support system to help them understand probabilities of effects and side effects of RA drugs by "experiencing" them in a computer-game-like setting. The proposed system will allow patients to view and manipulate an avatar that suffers a similar level of RA, and can choose to be treated with drugs the patient is considering. The system will run simulations of every-day tasks, showing multiple windows with varying levels of success. The levels will be determined by the patient's disease level, the choice of treatment (for the avatar), and the known probabilities of efficacy and side effects.

Probabilistic Computational Social Choice

In collaboration with Nicholas Mattei (Computer Science Ph D Student), Jörg Rothe (Universität Düsseldorf), Henning Fernau (Universität Trier), Gabor Erdelyi (Universität Düsseldorf)
More

One of the most common preference aggregation methods—the one most familiar to Americans—is election by majority. Other preference aggregation methods are not always recognized as such, for example, (sports) tournaments. One can view a sports tournament as an election where the best team wins. We can affect the outcome of a vote or tournament by voting and playing truthfully and to the best of our ability, etc., or by manipulating the aggregation process. There are several methods by which aggregation schemes can be manipulated. The most intuitive and well known is by influencing individual agents (through payments or other means). In real-world systems, typically not everything (the influence, the vote, the result) is observable by the manipulator. With this project, we focus on uncertain outcomes: What happens if the manipulator has access only to probabilities of vote outcomes and/or to probabilities of agents' responses to attempts to influence them. We achieve this through new model methods for established problems which take into account an agent's uncertainty about aspects of the aggregation procedures. Once we have developed these new models we study the complexity of lobbying and other influence methods in this uncertain world.

Decision-Theoretic Academic Advising

In collaboration with Joshua Guerin (Ph D Student), Robert Crawford (Ph D Student), Daniel Michler (Undergraduate Student), Tom Dodson (Undergraduate Student), Nicholas Mattei(Ph D Student).
More

During the course of an student's undergraduate education many decisions are encountered which may impact short and long term academic success as well as relative enjoyment and (perceived) utility that are obtained by the student. Human advisors help the student advisee make decisions that can have a positive major effect on their educational experience. The advisor's task is complicated by a potential lack of knowledge of the individual student's goals and preferences. Further, the potential long-term effects of actions may not be obvious even to experienced academic advisors. In order to deal with the difficulties encountered in academic advising our research group is developing tools and methods for generating stochastic models of an academic domain, and for fast stochastic planning and generation of advice. The project is divided into three areas: model construction, planning, and interface design. The academic domain poses challenges in each of these areas.

College of Medicine

Division of Rheumatology

Kristine Lohr

Rheumatoid Arthritis Project

in collaboration with Dr. Judy Goldsmith(Computer Science), Dr. Malachy Bishop (Special Education ), Dr. Jamie Studts (Behavioral Science), Dr. Nancye McCrary (Education Curriculum and Instruction), Radu Paul Mihail (Computer Science Ph D Graduate Student)
More

Rheumatoid arthritis (RA) is a serious disease, both painful and life-altering. Sufferers often find themselves unable to perform basic tasks, including dressing themselves, opening containers, using silverware, rising from a chair. There are medications available that, in many cases with high probability can enable the RA patient to accomplish such tasks. However, these medications often have low-probability but scary side effects. Patients often choose to not take drugs, because of highly unlikely scenarios. Although clinicians tell patents the probabilities of side effects, it seems that the patients do not grasp the unlikelihood of those possibilities. We propose a decision support system to help them understand probabilities of effects and side effects of RA drugs by "experiencing" them in a computer-game-like setting. The proposed system will allow patients to view and manipulate an avatar that suffers a similar level of RA, and can choose to be treated with drugs the patient is considering. The system will run simulations of every-day tasks, showing multiple windows with varying levels of success. The levels will be determined by the patient's disease level, the choice of treatment (for the avatar), and the known probabilities of efficacy and side effects.