Change degrees of freedom in error term

Moderators: statman, Analyst Techy, andris, Fierce, GerineL, Smash

Psycho7
Posts: 6
Joined: Tue Jan 13, 2015 8:24 pm

Change degrees of freedom in error term

Postby Psycho7 » Tue Feb 10, 2015 5:18 pm

I had to estimate several values in my dataset. I read somewhere that you need to adjust the degrees of freedom from the error term in that case. Is there a way to calculate the new F, p and eta-partial with SPSS?
(I analyse my data with a mixed design ANOVA)
GerineL
Moderator
Posts: 1477
Joined: Tue Jun 10, 2008 4:50 pm

Re: Change degrees of freedom in error term

Postby GerineL » Wed Feb 11, 2015 11:12 am

more info needed.
Are you talking about corrections for multiple testing?
Psycho7
Posts: 6
Joined: Tue Jan 13, 2015 8:24 pm

Re: Change degrees of freedom in error term

Postby Psycho7 » Wed Feb 11, 2015 4:29 pm

No. I replaced outliers by the next highest value. So these are 'estimations'. I have to perform a mixed design anova (1 between and 1 within subject factor). Because of the estimations, I need to correct the degrees of freedom for the error term (i.e. df2 in this formulation: F(df1, df2)). However, if you correct degrees of freedom, your F-value, p-value and eta squared partial change. Calculating these values by hand is complex.
So I was wondering whether there is an easy way to correct these degrees of freedom for the error term in SPSS so that these values are calculated automatically?
GerineL
Moderator
Posts: 1477
Joined: Tue Jun 10, 2008 4:50 pm

Re: Change degrees of freedom in error term

Postby GerineL » Thu Feb 12, 2015 8:59 am

ah ok. Not that I know of, but I am not 100% sure
There are tons of calculators online that let you do that by the way, so you could just enter your numbers in one of those?
Psycho7
Posts: 6
Joined: Tue Jan 13, 2015 8:24 pm

Re: Change degrees of freedom in error term

Postby Psycho7 » Thu Feb 12, 2015 5:01 pm

Okay, thank you!

Who is online

Users browsing this forum: No registered users and 1 guest

cron