Define 'tails' and 'type' better than the TTEST help file.

Please define 'tails' and 'type' better than the TTEST help file. Online
statistics glossarys are of little use.

thanks.

Before collecting the data, did you know which group would be larger if
there was a difference? If so, then you can use tails=1. Otherwise
(the difference could have been in either direction) you use tails=2.
This is discussed in more detail in any introductory statistics book.

Help on TTEST seems quite clear on type, assuming that you have a
background comparable to having covered t-tests in an introductory
statistics course. The choice is based on how your experiment was
designed, and involves more nuances than are easily addressed in a
simple newsgroup reply.

Jerry

goodmonkey wrote:

> Please define 'tails' and 'type' better than the TTEST help file. Online
> statistics glossarys are of little use.
>
> thanks.