- Data management
- Reproducible research (basics of best practices on coding, folder structure, running scripts, etc.)
- How to select the appropriate stat test
- How to handle and find errors and bugs in code
- Proper coding practices
- Tutorial on how coding/programming software works (its important to understand to grasp other concepts (e.g. macros, ODS, etc.))
-
Intro to the basics of SAS
-
Basic statistics in SAS
-
Macros (basic, intermediate, advanced)
-
ODS output
- Mixing in data manipulation/editing
- Like do one of the other workshops this way, teaching two things at once to show that git/VCS should be an integral part of doing any analysis. Like one group figure out what ODS to use, an other to build the macro.
-
Possible flow of sessions:
- Intro + data wrangle
- Basic stats
- Macros
- ODS
-
Intro to the basics of R
-
Basics of statistics in R
-
ggplot
-
Data visualization (I would argue this is more important to teach before other stats methods)
-
Functions
-
Report generation (R (knitr), pandoc, markdown)
-
Possible flow of sessions:
- Intro + data wrangle
- ggplot, data viz
- Basic stats
- Functions
- Report generation
-
Git or version control in general
- Maybe do an interactive (using etherpad), team assignment to show the benefits of using VCS. Like all create a github account, create a practice assignment, and merge/work together to do an analysis. Like split the group up into a few teams, get one group to generate data, another to analyze it, and another to make plots or something simple of course.
-
Possible flow of lesson(s):
- Pair up and alternate between pushers and pullers (owners and collaborators)
- At end of session (last 1 hour)