Faculty and staff from the BERD Unit both curate and contribute to the articles found below. To receive notifications when we post new articles, or to suggest new topics or content, email berd@uthsc.edu.
Newsletter
Statistical tests
- First Annual Survey on Management Software used in Statistical Consulting - Abstract: In a survey on software used to manage statistics consulting projects and billing, we received responses from 38 statistics and biostatistics departments. Of these, 31.2 % of departments reported no system for tracking projects and 34.2% reported no software in use for billing; 39.5% of departments used some combination of a REDCap/Google/Microsoft Form to capture client/project information; 8 organizations (21.1%) use… Read More
- Seminar Slides | The Perfect Doctor | Dr. Fridtjof Thomas - The-Perfect-Doctor-TN-CTSI-2019-05-28-handout Time: May 28th 12:00-1:00PM Location: Simulation Building Room 102 Presented by Fridtjof Thomas, PhD The Perfect Doctor: An introduction to Causal Inference Consider a Perfect Doctor that has the magical ability to pick from two treatments always the one that is better for each given patient. Can we learn about treatment effects by observing the… Read More
- Seminar Slides | Multiple Comparisons and the False Discovery Rate | Dr. Saunak Sen - Please click below for the slides from Dr. Saunak Sen’s presentation May 21st, 2019. Multiple Comparisons and the Falsefdr Discovery Rate
R, RStudio, Shiny & SAS
- How to Read a SAS Dataset Into R – The Right Way - The problem: while packages for reading in SAS datasets into R exist, they do not many formats, especially custom formats. Hence, a user must manually enter those in R. This becomes particularly onerous with survey datasets involving custom Likert scales. Solution: SAS-R scripte. This handy script from an anonymous contributor generates R code to set… Read More
- Equivalent of SAS Macros in R – Loops and Functions by Courtney Gale - If you’re an avid SAS user, you’re likely very familiar with SAS macros. SAS macros are a key component to creating efficient and concise code. Although you cannot use macros in R, R offers other features like functions and loops that can perform the same tasks as SAS macros. Using apply() to loop over variables… Read More
- A SAS Macro/R Code Comparison: Generating Multiple Datasets by Tristan Hayes - This is a short comparison of SAS and R code in the context of generating multiple datasets. In this example, the mpg dataset from R is used to show how one might use a SAS Macro to subset the data by car class, followed by the R equivalent. You will see that in SAS, the macro… Read More
P-values
- Seminar Slides | Multiple Comparisons and the False Discovery Rate | Dr. Saunak Sen - Please click below for the slides from Dr. Saunak Sen’s presentation May 21st, 2019. Multiple Comparisons and the Falsefdr Discovery Rate
- Seminar Slides: P-Value Discussion | Dr. Saunak Sen - Please see below for the sides from Dr. Sen’s seminar discussion. p-value-discussion
- P-values – What they are and what they are not – Seminar Presentation Slides – Fridtjof Thomas - P-values: What they are and what they are not, will look in detail at good examples of using p-values and how to interpret them. After reviewing widely understood problems with p-values, attention is drawn to regularly encountered use of p-values where it is less clear what their correct interpretation actually is. Furthermore, we demonstrate why… Read More
- American Statistical Association’s statement on p-values [outside article] - The American Statistical Association (ASA) has released a “Statement on Statistical Significance and P-Values” with six principles underlying the proper use and interpretation of the p-value. The ASA releases this guidance on p-values to improve the conduct and interpretation of quantitative science and inform the growing emphasis on reproducibility of science research. The statement also… Read More
Julia
- Interactive Visualizations Using the Julia Package Pluto - The Julia package Pluto makes it easy to write “reactive notebooks” that intersperse code with text, figures, and interactive elements. They have been used for teaching, blog posts and data analysis reports. Below we present some interactive Pluto notebooks developed by UTHSC researchers. At the moment they are hosted on Binder and may take a… Read More
- Julia Debugging Basics by Gregory Farage - This is a practical how-to guide on best practices for debugging code in Julia using the Gallium package. We explore two methods: REPL and Juno-Atom. Installation To use the debugger Gallium in Julia 0.6+, the following packages should be installed: Gallium and ASTInterpreter2. julia> Pkg.add("Gallium") julia> Pkg.clone("https://github.com/Keno/DebuggerFramework.jl") julia> Pkg.clone("https://github.com/Keno/ASTInterpreter2.jl") There are two possible ways to debug with Gallium,… Read More
- Missing values in Julia by Milan Bouchet-Valat [outside article] - Starting from Julia 0.7, missing values are represented using the new missing object. Resulting from intense design discussions, experimentations and language improvements developed over several years, it is the heir of the NA value implemented in the DataArrays package, which used to be the standard way of representing missing data in Julia.
Computation, statistical learning, and optimization
- Real Close to the Machine: Floating Point in D | Using Floating Point Without Losing Your Sanity by Dan Clugston [outside article & video] - An entertaining look at floating point computation and absurdities that you might encounter if you are not careful. The video is entertaining; the details are in the article. Computers were originally conceived as devices for performing mathematics. The earliest computers spent most of their time solving equations. Although the engineering and scientific community now forms… Read More
- Missing values in Julia by Milan Bouchet-Valat [outside article] - Starting from Julia 0.7, missing values are represented using the new missing object. Resulting from intense design discussions, experimentations and language improvements developed over several years, it is the heir of the NA value implemented in the DataArrays package, which used to be the standard way of representing missing data in Julia.
- Gradient Descent: The mother of all algorithms? by Aleksander Mądry [outside lecture] - More than half a century of research in theoretical computer science has brought us a great wealth of advanced algorithmic techniques. These techniques can be combined in a variety of ways to provide us with sophisticated, often beautifully elegant algorithms. This diversity of methods is truly stimulating and intellectually satisfying. But is it also necessary?