Mathematics, Computer Science, & Physics, 2013
  • Modeling the Spread of White-Nose Syndrome in Hibernating North American Bat Populations
    Sarah Bogen, Isaac Ressler
    Mentor: Paula Federico


    North American bat populations are currently being threatened by an emergent infectious fungal disease known as white-nose syndrome (WNS) which causes mass mortality in hibernating colonies. Since it was first detected in New York in 2006, WNS has spread rapidly in the United States and Canada and killed over 5.5 million bats. Control of WNS is of major concern to both the scientific and caving communities, and the disease and mechanisms of transmission are still not well understood. We developed an individual-based model at the county level to gain insight into the spatial and temporal spread of the disease. We assume the probability of infection for each county in a given year is a function of the density of caves, the estimated cave temperature, and relative proximity to other infected counties. Model parameters were estimated by means of maximum likelihood. We compared model predictions with known infection data from 2006 until 2011. The model imitates the overall spatial and temporal patterns of the data and may be improved by decreasing the number of “false alarm” predictions in future extensions of the study.

    Architectural Acoustics and Capital University’s Mezzanine
    Cameron Girard
    Mentor: Patrick Shields, Mark Lochstampfor, Dina Lentsner


    The Mezzanine, located on the third floor of Capital University’s Campus Center, hosts assorted musical and non-musical acts. Unfortunately, the space is plagued by a number of acoustic problems. The poor sound propagation threatens speech intelligibility and musical clarity. I propose a number of acoustic treatment options so the Mezzanine may better serve the variety of shows and events that are held there. Using what I have learned from my research, I ran acoustic tests and using modeling software to accurately quantify the issues. I also researched non-acoustic factors, like pricing and aesthetics, for a variety of treatment options. All of this is culminating into a number of comprehensive treatment propositions akin to what an acoustic consultant would present to a client. The “package” will be presented to organizations that could potentially carry out the project. Treatments that would work based on my findings include a system of curtains, diffusers and absorbers. These will reduce unnecessary reverberation, giving the space a more cohesive frequency distribution. The improvement of an acoustically critical space is not only a great fusion of my courses of study (music technology and physics), but has great potential to impact the events held at the Mezzanine.

    Law and Statistics: The OJ Simpson Case
    Lindsay Pester
    Mentor: Keith Wilkinson


    Statistics have played and will continue to play a crucial role in several legal cases. When used correctly, statistics can be a major part of a case where physical evidence is lacking. Unfortunately, there is a common misconception that all statistical calculations are accurate. Because of this, there are several cases where a verdict has been determined by miscalculated probabilities. This paper researches some of those cases and explores where the calculations went wrong. The highly publicized O.J. Simpson trial is highlighted to show how the defense attorney misinterpreted probabilities and convinced the jury that the fact that Simpson had physically abused his wife was irrelevant to the murder of his wife. In fact, when using Bayes’ Formula, the probability that the husband is the murderer of his physically abused wife is .81. This paper details the calculations that lead to this conclusion. Statistical calculations can be an asset for either side of a legal case, but only when calculations are made by experts in the field of statistics.

    A Computational Model of Insulin Release from the Endocrine Pancreas
    Ariel Webb
    Mentor: Catherine Boulant, Paula Federico


    Because glucose is the primary energy substrate for metabolism, homeostatic control of blood glucose levels is critical. Insulin, a peptide hormone secreted from the pancreas, is the primary regulator of blood glucose levels. The objective of this project is to develop a computational model of blood glucose and insulin levels during a hypothetical Glucose Tolerance Test (GTT) and to compare those values to actual values obtained during a modified GTT. The software package, STELLA v 10.0, was used to develop the model. It is software that is designed to facilitate the building and use of System Dynamics models. The model reflects values presented in the published literature and favorably compared to the values obtained during a modified GTT. Input variables representing glucose set point, glucose release, glucose uptake, insulin secretion, insulin breakdown, and insulin usage fraction can be manipulated to model the dynamic control of glucose levels by insulin. An accurate model of glucose-insulin interrelationship may be used as an educational tool and could demonstrate how computational models may be used to enhance understanding of physiological systems. Future manipulations of the model may include different forms of diabetes. Additional time-dependent factors influencing insulin release may be integrated into the model.