R help - Jun 2005 - Factanal loadings as large as 1.2 with promax -- how unusual?

I am performing a large (105 variable) factor analysis with factanal,
specifying promax rotation.  I kow that some loadings over 1.0 are not
unsual with that rotation, but I have some as large as 1.2, which seems
extreme.  I am skirting the assumptions of the model by using responses
on a 7-point rating scale as data; I may have to go back and compute
polychoric correlations instead of product moment, but before doing that
I would like to know if others have had equally large factor loadings
using data that are truly interval level data on continuous scales.
 
Thanks for suggestions or information,
 
Ben Fairbank
 

	[[alternative HTML version deleted]]

Dear Ben,

To get big factor loadings like this, I'd guess that you have large
communalities (not a bad thing, of course, and probably less rather than
more likely with crudely measured variables) and strong correlations between
factors. I suppose that if the latter get too large that might suggest that
you could get by with fewer factors.

I hope this helps,
 John

--------------------------------
John Fox
Department of Sociology
McMaster University
Hamilton, Ontario
Canada L8S 4M4
905-525-9140x23604
http://socserv.mcmaster.ca/jfox 
-------------------------------- 
> -----Original Message-----
> From: r-help-bounces at stat.math.ethz.ch 
> [mailto:r-help-bounces at stat.math.ethz.ch] On Behalf Of Ben Fairbank
> Sent: Monday, June 20, 2005 4:10 PM
> To: r-help at stat.math.ethz.ch
> Subject: [R] Factanal loadings as large as 1.2 with promax -- 
> how unusual?
> 
> I am performing a large (105 variable) factor analysis with 
> factanal, specifying promax rotation.  I kow that some 
> loadings over 1.0 are not unsual with that rotation, but I 
> have some as large as 1.2, which seems extreme.  I am 
> skirting the assumptions of the model by using responses on a 
> 7-point rating scale as data; I may have to go back and 
> compute polychoric correlations instead of product moment, 
> but before doing that I would like to know if others have had 
> equally large factor loadings using data that are truly 
> interval level data on continuous scales.
>  
> Thanks for suggestions or information,
>  
> Ben Fairbank
>  
> 
> 	[[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