Seminar Schedule



Estimating Judicial Accomplishment for the U.S. Supreme Court 

Marcus Hendershot, Department of Political Science

This talk centers on a collaborative research effort that is taking place between faculty members within the Political Science and Statistics departments and that utilizes resources from OSU's High Performance Computing Center (HPCC).  This particular project bridges the existing divide between the landmark policy indicators of the U.S. Congress and the U.S. Supreme Court. It applies the legislative accomplishment measurement strategy to the Court’s orally argued cases (1899 to 2004 terms) and invokes an expansive array of dichotomous indicators of contemporary salience and retrospective import to create continuous level judicial accomplishment estimates. Tests of reliability and validity for this new measure suggest that it performs well versus contemporaneous, retrospective and network-based measures of landmark status. These new values can be utilized to better understand the institutional development of the Court as well as to discern differences within the policy-making processes of elected and life tenured branches.

September 16, 2016
MSCS 310



Characterizing uncertainty in agricultural systems model predictions of climate impacts

Phillip Alderman, Department of Plant and Soil Sciences

Dynamic simulation modeling is a useful approach to understanding how agricultural systems respond to external factors (e.g. climate) over time.  Recent international efforts focused on climate impact assessment have brought renewed emphasis to the comparison of different agricultural systems models and their combination into model ensembles. To date, analyses of model-ensemble simulated results has not clearly differentiated between prediction uncertainties due to model structural differences per se and those due to parameter value uncertainties. The objectives of this presentation will be to introduce agricultural systems modeling, explain its application for climate impact assessment, and illustrate the use of Bayesian analysis in characterizing components of overall agricultural model prediction uncertainty.

September 30, 2016
MSCS 310




Josie Akosa

Dept. of Statistics

Josie will share her experience and other information about summer interships.  All are encouraged to attend. 

Refreshments in the conference room after the seminar.

October 7, 2016
MSCS 310



Kalyani Nagaraj - rescheduled for the Spring 2017 semester
Dept. of Industrial Engineering and Management




Simple Data Analytic Tools for Psychologists and other Social Scientists

James W. Grice, Ph.D.
Professor, Department of Psychology
For over 50 years psychologists have relied almost exclusively on null hypothesis significance testing to draw inferences from their data. At the same time surveys have repeatedly shown that majorities of psychologists do not properly understand the meaning of "statistical significance." Consequently, modern psychology finds itself in a serious crisis, as shown most dramatically by recent failed replications that were widely publicized. In this talk I present our efforts to develop and test simple analytical tools that are more suitable to the types of data collected by psychologists. These tools encourage psychologists (and other social scientists) to consider classification accuracy as more important than statistical significance. They also encourage researchers to develop a more sophisticated view of model creation, evaluation, and development.
November 4, 2016
MSCS 310


Bayesian Nonparametric Spatial Modeling for Mesonet Data Using Dirichlet Process Mixing

Longji Sun, Department of Statistics

Statistical modeling plays an important role in spatial data analysis. While the Gaussian processes (GP) are often used to model continuous spatial data, the Dirichlet processes (DP) models are able to handle non-Gaussian and non-stationary spatial data. Within a Bayesian framework, a multivariate spatial DP model is proposed for soil moisture measurements collected from Oklahoma Mesonet. The proposed model takes into account the spatial dependence of soil moisture among Mesonet stations and the cross-variable dependence. Posterior inference on model parameters are obtained by using Markov chain Monte Carlo (MCMC) to draw samples from the posterior distribution of the parameters given data. Spatial predictive mapping is obtained using posterior samples in a parallel setting, for locations where measurements are not observed. The proposed approach is illustrated using both simulated data and real soil moisture measurements.

November 18, 2016
MSCS 310


Utilizing Observations and Ranks in Kernel Estimation

Nicholas Kaukis, Department of Statistics

For univariate i.i.d. samples, analyses are usually performed on observations themselves, or on the ranks of the observations when they are ordered. This is true in both the parametric and non-parametric setting. However, observations and ranks can be utilized simultaneously. This work concentrates on developing methods that combine observations and ranks in kernel-based estimation methods. The standard kernel density estimator, transformation kernel density estimator, and the Nadaraya-Watson kernel regression estimator are modified using ranks. Excluding the modified transformation kernel density case, the asymptotic properties are investigated for the estimators. For all estimators, various bandwidth selection methods are explored. Simulation is used to compare the modified estimators to the standard estimators. It is demonstrated that when samples are taken from skewed distributions, the modified kernel density estimator outperforms the standard estimator. Additionally, the modified Nadaraya-Watson outperforms the standard estimator in a variety of cases. The modified transformation density estimator did not outperform the unmodified estimator.

December 2, 2016
MSCS 310