Statistical Consulting

Overview

The NDSU Center for Computationally Assisted Science and Technology (CCAST) provides statistical consulting to all faculty, staff, and students involved with research at NDSU. There is no charge to clients. 

The service is provided by Curt Doetkott using email and Zoom or Microsoft Teams. If you would like to use the service, initial contact should be made via e-mail to curt.doetkott@ndsu.edu.  Your e-mail should identify the kind of statistical help you are looking for--this may include a brief description of your research questions and possibly supporting materials such as relevant papers or data. This process allows follow-up appointments as needed via Zoom or Teams.  Appointments are available from 10 am to 5 pm, Monday through Friday.

We provides assistance to faculty, staff and graduate students regarding data analysis and general statistical questions related to their research. Consulting is not intended for class-related questions like homework or tutoring. If you are not sure if the questions you have are appropriate for statistical consulting, feel free to contact us and we will discuss your needs and whether or not our services may be helpful.

• your research goals and assist you in identifying aspects of your research that may benefit from the application of statistical principles.
• statistical methods that may be relevant for your project based upon your research questions and available data.
• the design of experiments; e.g., identification of response variables, factors, and type of statistical design appropriate for your research. 
• mechanisms for collecting, entering (e.g., Excel), and managing data prior to statistical analysis. 

• suggest appropriate statistical methods for data analysis. 
• assist in managing and processing data. 
• obtain data from clients and perform statistical analyses (if appropriate and with proper permissions). 

• recommend methods of data presentation, description and portrayal of results. 
• facilitate graphical display of data by pre-processing your data and generating transportable formats like png. 
• generate presentation graphics for clients as needed. 

Most of our consulting work is done using SAS. We also have experience with SPSS, JMP, and R. 

Our experience with statistical methods is fairly broad, but certain methods comprise a lot of our consulting efforts. These include: 

• regression (linear, nonlinear, logistic).
• ANOVA including repeated measures and mixed models.
• discrete data analysis methods (e.g., frequency tables with chi-square tests).
• multivariate methods.

Recent years have seen increases in the use of methods such as generalized linear mixed models (GLMMs), resampling techniques (bootstrapping, permutation tests and Monte Carlo methods) and graphical methods/visualization techniques. If you are interested in assistance with areas not listed, feel free to check with us anyway--we have used many other methods not listed and we are always looking to broaden our experience. 

Our staff have worked collaboratively with researchers in many disciplines. This has often included reviewing the descriptions of statistical methods and results in proposals and manuscripts as well as providing opinions on reviewer comments received in the publication process. We have been included as co-authors on manuscripts when our contributions were deemed worthy of inclusion. We agree to co-authorship only if we believe we have contributed significantly to the project.

Contacts 

Curt Doetkott: Quentin Burdick Building 242M. Phone: (701) 231-7031. E-mail: curt.doetkott@ndsu.edu