Message Image  

Welcome to the Bayesian Statistical Science Section

The Bayesian Statistical Science Section of the ASA provides a forum for statisticians and people who have interest in the Bayesian paradigm. The broad objectives of the Section are: to encourage research on theory and methods of statistical inference and decision making associated with Bayes' theorem and to encourage the application and proper use of Bayesian procedures in the behavioral, biological, managerial, engineering, environmental, legal, medical, pharmaceutical, physical, and social sciences.



Current Affairs 


Upcoming Web-Based Lectures

 

Title: Effective Power and Sample Size Analysis with SAS
Presenter: John Castelloe, PhD, Research Statistician Developer, SAS
Date and Time: Wednesday, April 27, 2016, 12:00 p.m. - 1:30 p.m. Eastern time
Sponsor: Section for Statistical Programmers and Analysts

Registration Deadline: Monday, April 25, at 12:00 p.m. Eastern time

Description:
Sample size determination and power computations are an important aspect of study planning; they help produce studies with useful results for minimum resources in diverse application areas including clinical trials, marketing, and manufacturing. This webinar presents numerous examples to illustrate the components of a successful power and sample size analysis for proportion tests, t tests, confidence intervals, equivalence and noninferiority, survival analyses, logistic regression, repeated measures, and nonparametric tests. Attendees will learn how to compute power and sample size, perform sensitivity analyses for factors such as variability and type I error rate, and produce customized tables and graphs using the POWER and GLMPOWER procedures in SAS/STAT software.

 

Title: Enhancing the Value of Qualitative Research Using the Total Quality Framework (TQF)
Presenter: Margaret R. Roller (Roller Research) and Paul J. Lavrakas (Independent Consultant)
Date and Time: Thursday, June 9, 2016, 1:00 p.m. - 3:00 p.m. Eastern time
Sponsor: Survey Research Methods Section

Registration Deadline: Tuesday, June 7, at 12:00 p.m. Eastern time

Description:
Oftentimes a research question cannot be answered well through the use of quantitative research or through the exclusive use of quantitative approaches. For example, quantitative survey data may leave a researcher with unanswered questions about the reasons that underlie the responses or the particular contexts in which respondents framed their answers. That is why statisticians and other quantitative researchers are on occasion involved in conceptualizing, conducting, interpreting, and/or reviewing research projects that include the use of qualitative research methods.

Qualitative research goes beyond the expedient to gain a richer, more intricate appreciation of the research issue. Deriving these complex and contextual data, however, presents unique challenges to researchers who attempt to combine the essence of qualitative research with reliable and valid approaches that maximize the usefulness of their research. It may be because of these challenges that quality-design issues related to qualitative research - such as coverage, sample selection, nonresponse (including missing data), and researcher bias - have heretofore received relatively modest consideration by the qualitative research community.

In this presentation, we introduce a new approach that brings greater rigor to qualitative research. That approach is the Total Quality Framework (TQF) (Roller & Lavrakas, 2015). The TQF provides researchers with a systematic yet highly flexible way to (a) give explicit attention to reliability and validity issues in qualitative research, (b) critically examine the possible sources of bias and inconsistency in qualitative methods, (c) incorporate features into qualitative research designs that try to mitigate these effects, (d) acknowledge and take their implications into consideration during analysis, and (e) thereby maximize value of the research outcomes.

Our presentation: 1) presents a brief overview of what makes qualitative research uniquely different from quantitative; 2) explains the TQF and its value for conceptualizing, implementing, interpreting, and reviewing qualitative research; and 3) illustrates the application of the TQF by way of two qualitative methods, in-depth interviews and focus group discussions. It is intended that our presentation will help quantitative researchers think more critically and confidently about the value that qualitative methods can bring to their studies.

 

Title: Pushing the Frontier of TFL Automation and Dynamic Visualization with R/Shiny
Presenters: Danni Yu, Eli Lilly and Company and Tuan Nguyen Sr., Eli Lilly and Company
Date and Time: Thursday, July 7, 2016, 11:00 a.m. - 12:30 p.m. Eastern time
Sponsor: Section for Statistical Programmers and Analysts

Registration Deadline: Tuesday, July 5, at 12:00 p.m. Eastern time

Description:
Producing TFLs can be a tedious, time-consuming, expensive and painful process. It has been challenging to automate until the arrival of new technologies. Shiny is a R tool for building web-based GUI for statistical analyses and is well-suited for automation. In addition, Shiny is built for dynamic visualization/analyses; this key feature allows us to interact with data dynamically, thus enabling proactive engagement with physicians/scientists. We will give some examples on how R/Shiny is used to lead innovation in drug development.

About the Presenters:
Danni Yu (Research Scientist, Oncology Biomarker Statistics, Eli Lilly and Company) and Tuan Nguyen (Sr. Research Scientist, Oncology Biomarker Statistics, Eli Lilly and Company). Dr. Yu and Dr. Nguyen and colleagues developed a comprehensive automation platform based on the R and Shiny tools that allows for fast, dynamic, scalable, inexpensive and reproducible visualization/analyses and generation of TFLs in drug development.

  

Title: Enabling Reproducibility in Statistical Analyses Using R Markdown
Presenter: Eric Nantz, Eli Lilly and Company
Date and Time: Thursday, July 14, 2016, 11:00 a.m. - 12:30 p.m. Eastern time
Sponsor: Section for Statistical Programmers and Analysts

Registration Deadline: Tuesday, July 12, at 12:00 p.m. Eastern time

Description:
Reproducibility in statistical analyses has always been an important topic in many fields of statistics, but has gained even more attention in the last few years. In the past, the software tools enabling reproducibility in statistical programming required a large investment in time and effort. However, a new ecosystem around reproducibility has emerged within the R statistical language. In this talk, I will demonstrate specific examples using R in combination with rmarkdown and additional packages to make analysis reproducibility easy to set up and maintain throughout the life cycle of a project.

About the Presenter:
Eric Nantz is a senior research scientist supporting advanced analytics within the immunology unit at Eli Lilly and Company. Eric has utilized R in a wide variety of analyses involving clinical and novel biomarker data sets. Eric is also the creator and host of the R-Podcast, an audio podcast that provides valuable information for both new and experienced R users to accomplish data analyses.

  *** More information and registration at the webinar page.

 

 


From the SBSS Mixer at JSM 2015

 

Ed George with Savage finalists


SBSS student paper award winners

 

Nimar Arora - Mitchell Prize Recipient

 

David Dahl  and Leo Duan

   




















 

Latest Discussions

There are no discussions to which you have access

Either the content you're seeking doesn't exist or it requires proper authentication before viewing.

Recent Shared Files

No data found

Either the content you're seeking doesn't exist or it requires proper authentication before viewing.

Recent Blogs

Log in to see this information

Either the content you're seeking doesn't exist or it requires proper authentication before viewing.