This Data Analysis Practicum is intended for graduate students in statistics who already have a comprehensive background probability, inference, and linear modeling. Our focus for this semester will be on applying the techniques you've learned in graduate school to a single real-world data analysis problem. This semester, we will be considering data on Russian twitter messages from the time of the 2016 US presidential election.
Room: Science & Math Learning Center (SMLC), Rm. 352
Time: Tuesday & Thursday, 2:00pm – 3:15pm
Prerequisites: STATS 561 (Probability), STATS 553 (Inference), STATS 540 (Regression), STATS 545 (ANOVA)
Syllabus: dap_s19_syllabus.pdf
At the link above, you can access the original Russian tweet data as compiled by Drs. Linvill & Warren at Clemson University. The data set consists of 13 .csv files and a readme file.
The data for this class were published online by the website 538 during the summer of last year. This write-up provides a basic introduction to the data and its historical context.
A pre-publication report on the data by the researchers who collected it: Darren L. Linvill and Patrick L. Warren of Clemson University. This provides some additional context for, and information about, the data.
Dr. Christensen's presentation on statistical consulting. There are also additional readings on Type-3 errors, consulting skills, experimental planning, and professionalism for the interested student.
You may also find the R code Dr. Christensen used in class to be helpful. Find it here.
These two short guides give you tips for how to approach a data analysis project. Mine (Fletcher's) focuses on the larger scientific context in which the analysis is conducted. Ron's focuses on the analysis itself and making sure it is done acceptably.
This presentation is a guide to how to give good statistics presentations in various contexts.
Final Report Handout -- Due May 10th
Office: SMLC 328
Spring 2019 Office Hours: