R help - Aug 2005 - converting a t statistic to r2

HI
 
I wonder if anyone can help.  I have a longitudinal sample of 100 subjects:
200 data points were acquired starting at different ages and at irregular
intervals (subjects have different numbers of repeated data points, so some
have only one data point).  I have been examining the relationship over time
(it is quadratic) of continuous variables A on variable B.  To model this I
have been using linear mixed models in R.  
 
B~A*age +A*I(age^2) + random term (for individual)
 
I get t values associated with A, age and A*age.
 
How (or can) the t value for A be converted to a correlation (r) or variance
value?  
 
I recall that R2 = t2/(df+t2)
 
But can this be applied to linear mixed models and what are the degrees of
freedom?
 
I hope this is reasonably clear,
 
Many thanks for any opinions
 
Philip
 

	[[alternative HTML version deleted]]

The formula R2 = t2/(df+t2) applies only if a single intercept and a 
single slope are estimate with simple linear regression in something 
like B~A or B~age, but not with interaction nor quadratic term in age.

	  For further information, I just got 4 hits from 'RSiteSearch("R^2 in
lme")', two of which seemed relevant to your question: 
"http://finzi.psych.upenn.edu/R/Rhelp02a/archive/17572.html", and 
"http://finzi.psych.upenn.edu/R/Rhelp02a/archive/34377.html".

	  If you still want more help after this, please submit another 
question -- after reading the posting guide! 
"http://www.R-project.org/posting-guide.html".  The Posting Guide
serves
two purposes:  (a) It helps people get better answers to their questions 
quicker.  (b) It makes it easier for people who try to answer such 
questions to understand what the questioner wants.  It seems to succeeds 
fairly well on both counts when it is used.  I think I can see in a 
question whether the submitter has paid adequate attention to the 
posting guide:  The questions tend to be better focused, more complete, 
and easier to understand and reply to.  If you want free consulting, you 
have to pay for it.

	  spencer graves
-- 

Shaw, Philip (NIH/NIMH) wrote:
> HI
>  
> I wonder if anyone can help.  I have a longitudinal sample of 100 subjects:
> 200 data points were acquired starting at different ages and at irregular
> intervals (subjects have different numbers of repeated data points, so some
> have only one data point).  I have been examining the relationship over
time
> (it is quadratic) of continuous variables A on variable B.  To model this I
> have been using linear mixed models in R.  
>  
> B~A*age +A*I(age^2) + random term (for individual)
>  
> I get t values associated with A, age and A*age.
>  
> How (or can) the t value for A be converted to a correlation (r) or
variance
> value?  
>  
> I recall that R2 = t2/(df+t2)
>  
> But can this be applied to linear mixed models and what are the degrees of
> freedom?
>  
> I hope this is reasonably clear,
>  
> Many thanks for any opinions
>  
> Philip
>  
> 
> 	[[alternative HTML version deleted]]
> 
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide!
http://www.R-project.org/posting-guide.html
-- 
Spencer Graves, PhD
Senior Development Engineer
PDF Solutions, Inc.
333 West San Carlos Street Suite 700
San Jose, CA 95110, USA

spencer.graves at pdf.com
www.pdf.com <http://www.pdf.com>
Tel:  408-938-4420
Fax: 408-280-7915