Measures of correlation allow you to tell the extent to which two variables systematically change, without assuming any functional relationships. Depending upon the data, correlation analyses can be conducted by assuming parametric assumptions (such as normality of the raw data), as non-parametric approaches (based upon rankings), and via permutation analyses.
At the end of this topic, you should be able to:
- Understand the differences between parametric, non-parametric, and randomization methods for hypothesis testing.
- Test the normality of data using the Shapiro-Wilks test.
- Perform basic transformations on data to attempt to get your data more representative of normal data sets.
- Use the
cor.test()function for parametric and non-parametric data.
This topic has the following general content.
- Slides for the topic.
- A larger narrative on correlation and permutation.
- An in-class activity
- A video recording of the lecture content (as I'll be out of town).
If you need to contact me, I am available at:
- Email: rjdyer@vcu.edu
- Twitter: @dyerlab