Here is a video I made about why composite scores are more extreme than the average of two subtest scores. It is a companion to the paper I wrote with Kevin McGrew about the dangers of averaging subtest scores and thinking of them as composite scores.

The Excel spreadsheet I used in the video can be downloaded here.

### Like this:

Like Loading...

*Related*

Thanks, very nice work

Great video, great teacher. Makes the complex simple. It will be high on my list of recommended viewing

Thanks, Trevor!

Fantastic work! I’m adding this to my bookmarks and will be passing the link along. Thanks~

Very nice.

One question: does this representation/calculation work properly in the case of negative correlations?

Thanks!

Yes

Pingback: Why composite scores are more extreme than the average of their parts | Assessing Psyche, Engaging Gauss, Seeking Sophia