Joe (and others),
The approach I've found most useful in teaching "effect size" is to
make two transparencies of a normal distribution, each with standard
deviations marked, and to superimpose them on an overhead projector (it
works best if they're two different colors). Then, by sliding them
relative to one another one can easily illustrate various effect sizes.
My students are often struck by how much two distributions can overlap,
but still have "large" d values. Hope this helps.
---
You are currently subscribed to tips as: [EMAIL PROTECTED]
To unsubscribe send a blank email to [EMAIL PROTECTED]
