• February 26, 2017

# Chronicle Forums

February 26, 2017, 3:43:19 pm
 News: Talk online about your experiences as an adjunct, visiting assistant professor, postdoc, or other contract faculty member.
 Pages: [1]
 Author Topic: Effect Size for Paired T-test  (Read 21516 times)
august_leo
Distinguished Senior Member

Posts: 1,339

 « on: April 19, 2010, 10:24:12 am »

How do you calculate effect sizes for paired t-tests?

I usually use Cohen's d for my effect sizes, which I calculate manually in excel for one-sample t-tests against chance. I sometimes use this calculator to do them quickly for conference submissions, etc. I don't have a routine for paired t-tests.

I found one calculator for this online, but the values seem too low....
 Logged

Your environment sounds vaguely toxic.  Or maybe just characteristically British.
I heart august_leo.
resis
Junior member

Posts: 97

 « Reply #1 on: April 19, 2010, 3:53:32 pm »

r = square root of (t-squared / [t-squared + degrees of freedom])

or if you prefer the coefficient of determination,

r-squared = t-squared / (t-squared + degrees of freedom)

Source:

p. 19 of Rosenthal, Robert (1991) Meta-analytic procedures for social research (revised). Newbury Park, CA: Sage.

 Logged
drspouse
Distinguished Senior Member

Posts: 1,621

 « Reply #2 on: April 23, 2010, 12:13:05 pm »

Hi august_leo

I use a variety of online calculators including:

http://www.cognitiveflexibility.org/effectsize/

http://www.stat.uiowa.edu/~rlenth/Power/index.html

You can usually find what you need here

http://statpages.org/
 Logged
august_leo
Distinguished Senior Member

Posts: 1,339

 « Reply #3 on: April 23, 2010, 6:18:24 pm »

Thanks, resis and Dr Spouse!

(my internet hasn't been working at home, so it took a while to get back to the fora.)
 Logged

Your environment sounds vaguely toxic.  Or maybe just characteristically British.
I heart august_leo.
 Pages: [1]