Yes...speed is coded in the opposite direction from memory and reasoning.

So, for memory and reasoning, higher scores equal BETTER performance.

For speed, higher scores indicate POORER (slower) performance.

All of the scores are in a "Z-like" distribution (Blom distribution), so their scaling is kind of arbitrary.

-----Original Message-----
Sent: Thursday, February 12, 2009 10:45 PM
To: Michael Marsiske
Subject: Re: FW: multivariate hw clarification

Hi Dr. Marsiske!

Regarding the first dataset, when examining the "speed" post-test
variable, is it correct that a positive score indicates slower

Thank you for your time,