Michael Harris, over at EASI blog has provided a great reference for situations where you need to illustrate what a correlation coefficient means--to students, stakeholders, whoever.
The tool can be found here and allows the user to model N values ranging from 5 to 100 and r values from -1.0 to +1.0.
It also allows you to generate as many samples as you want to illustrate how points might be distributed AND analyze the leverage that particular data points have on the coefficient AND allows you to move each data point and see the impact! Check it out.
No comments:
Post a Comment