Give-Now

ASA 175logo

IYS

 

 

 

 

 

WIS2015

 Welcome to the OSU Department of Statistics

The Statistics Department at Oklahoma State University is the only statistics department in the state of Oklahoma.  The department offers courses in both undergraduate and graduate degree programs.  Our graduates are in demand in the areas of:  education, actuarial science, health care, agriculture research, marketing research, money and banking and many more disciplines.

OSU Department of Statistics
301 MSCS (Math Statistics and Computer Sciences)
Stillwater, OK  74078-1056

Office:  405-744-5684

FAX:  405-744-3533

 

Statistics Seminar:   April 8, 2016, 2:30pm MSCS 310

Hailey Altom
Master's Student

"Zero Variance Component Estimates and the Influence on Tests of Fixed Effects"

Abstract: In a mixed effects model it is important to appropriately estimate variance components. Specific options available in SAS software using the Restricted Maximum Likelihood (REML) approach of variance component estimation are investigated. Of particular interest are situations in which the variance component estimates are zero. SAS software has various options for degrees of freedom adjustments available when a zero component estimate is yielded. In this study we investigate how these adjustments influence the power and size of the test.

Alex King
Master's Student

"Combining Multiple K-S Test Statistics via Permutation Tests"

Abstract: One of the driving forces of academia and research is money. Money forces researchers to find shortcuts and cheaper methods to analyze particles. This is the driving force behind the development of a permutation version of an extension of the Kolmogorov-Smirnov test for comparing multiple estimated CDFs. We will look at the motivational problem that arose in the Engineering Department that caused a need for this extension. After the motivational problem, we will discuss the cases that will be considered via simulations and possible future research questions.

 

  

Click here to like us on Facebook 

FBLike