Course Structure and Topics
This course combines lectures on theory and concepts with significant time practicing statistical tools within the R environment. The course concludes with a week-long project session where participants work independently to conduct a full analysis of a provided dataset and present their results. The format is designed to make it as accessible as possible to both professionals and graduate students who may be navigating complicated work and family schedules. Content is typically posted one week early for those looking to work ahead while time is available.
Participants should expect to spend between 8-10 hours per week working through the course material and assignments. Each week will include:
- 2-4 hours of recorded presentations (slides with audio) introducing new theory and concepts
- One or more well-commented demonstration code scripts teaching new analytical tools
- At least one analysis assignment, where participants will adapt the demonstration code to complete a novel analysis and answer a series of questions regarding the results
- An optional live Q & A session with the instructor (Friday 9:30-10:20 am ET)
- An optional live review session to walk through assignment solutions (Monday 10:30-11:45 am ET)
- Two opportunities for “virtual office hours” with instructor where time blocks can be reserved by individual participants (Thursday 1:30-2:30 pm; Friday 10:30-11:30 am ET)
This course covers:
- Introduction to probability theory, random variables, and statistical distributions
- Linear models (univariate, multivariate)
- Generalized linear models (Poisson, quasi-Poisson, negative binomial, and binomial)
- Linear mixed models and generalized linear mixed models
- Testing for model assumptions and model fit
- Proper use of data transformations to improve model performance and fulfill assumptions
- Evaluating model performance with diagnostic tools
- Preparing quality graphics and interpreting results
- Dealing with common problems in data including missing data and collinearity
- Tips and tricks of programming in R
- Organizing analyses into data exploration, descriptive statistics, model application, model diagnostics, and discussion of results.
Software and Programs
All exercises and some lectures will be performed in R, through user-friendly interface RStudio. Pre-course work will be suggested for all participants not familiar with the program R and emailed to participants at least 3 weeks prior to the course. All participants should have a basic familiarity working in the R environment by the time the course begins.
