Factor analysis: values of raw component covariance matrix

classic Classic list List threaded Threaded
5 messages Options
Reply | Threaded
Open this post in threaded view
|

Factor analysis: values of raw component covariance matrix

luisp
Hello all,

I've been asked to find which theories (6 in total) have a higher "weight" for 200 participants which answered a question composed of 20 items in a questionnaire. Items that refer to the same theory are supposed to be related between them. I've selected factor analysis with principal component method for extraction and covariance matrix without rotation. Theories are acting as “latent variables” (they can't be directly observed) but, instead, they are inferred through the items of the question X in the questionnaire, acting as “observed variables". The results are:

1st: Social Norms (59,35% of variance explained)
2nd: Political Regime (53,66% of variance explained)
3rd: Social Movement (43,98% of variance explained)
4th: Relative Deprivation (35,95% of variance explained)
5th: Rational Choice (29,37% of variance explained)
Contagion Theory can’t be measured (only 1 item)

I also got the values of raw component covariance matrix (and, for all cases, between .615 and 1.433). How can I interpret the values? Or should I rely only in % of variance explained for each latent variable?

Is there any other possibility in order to know which are the most important theories (latent variables)? Thanks in advance.

Luis
Reply | Threaded
Open this post in threaded view
|

Re: Factor analysis: values of raw component covariance matrix

Maguin, Eugene
I'm mostly confused by what you have done and are asking. You have six theories. Were there 20 questions for each theory (a total of 120) or were the same 20 questions used for all six theories? What do you mean by "weight"? What would I look at on the output to assess weight? I think there are more questions to ask but please begin with these.
Gene Maguin


-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of luisp
Sent: Wednesday, October 12, 2016 3:45 AM
To: [hidden email]
Subject: Factor analysis: values of raw component covariance matrix

Hello all,

I've been asked to find which theories (6 in total) have a higher "weight"
for 200 participants which answered a question composed of 20 items in a questionnaire. Items that refer to the same theory are supposed to be related between them. I've selected factor analysis with principal component method for extraction and covariance matrix without rotation. Theories are acting as “latent variables” (they can't be directly observed) but, instead, they are inferred through the items of the question X in the questionnaire, acting as “observed variables". The results are:

1st: Social Norms (59,35% of variance explained)
2nd: Political Regime (53,66% of variance explained)
3rd: Social Movement (43,98% of variance explained)
4th: Relative Deprivation (35,95% of variance explained)
5th: Rational Choice (29,37% of variance explained) Contagion Theory can’t be measured (only 1 item)

I also got the values of raw component covariance matrix (and, for all cases, between .615 and 1.433). How can I interpret the values? Or should I rely only in % of variance explained for each latent variable?

Is there any other possibility in order to know which are the most important theories (latent variables)? Thanks in advance.

Luis



--
View this message in context: http://spssx-discussion.1045642.n5.nabble.com/Factor-analysis-values-of-raw-component-covariance-matrix-tp5733238.html
Sent from the SPSSX Discussion mailing list archive at Nabble.com.

=====================
To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD
Reply | Threaded
Open this post in threaded view
|

Re: Factor analysis: values of raw component covariance matrix

Bruce Weaver
Administrator
I was also confused.  But based on my (quite possibly incorrect) understanding of what the OP is trying to do, I think confirmatory factor analysis (CFA) would be more appropriate, with model fits being compared via AIC or BIC, perhaps.  But of course, SPSS-Statistics has no means of carrying out CFA.

Maguin, Eugene wrote
I'm mostly confused by what you have done and are asking. You have six theories. Were there 20 questions for each theory (a total of 120) or were the same 20 questions used for all six theories? What do you mean by "weight"? What would I look at on the output to assess weight? I think there are more questions to ask but please begin with these.
Gene Maguin

--- snip the original post ---
--
Bruce Weaver
bweaver@lakeheadu.ca
http://sites.google.com/a/lakeheadu.ca/bweaver/

"When all else fails, RTFM."

NOTE: My Hotmail account is not monitored regularly.
To send me an e-mail, please use the address shown above.
Reply | Threaded
Open this post in threaded view
|

Re: Factor analysis: values of raw component covariance matrix

Jon Peck
AMOS

On Wed, Oct 12, 2016 at 7:42 AM, Bruce Weaver <[hidden email]> wrote:
I was also confused.  But based on my (quite possibly incorrect)
understanding of what the OP is trying to do, I think confirmatory factor
analysis (CFA) would be more appropriate, with model fits being compared via
AIC or BIC, perhaps.  But of course, SPSS-Statistics has no means of
carrying out CFA.


Maguin, Eugene wrote
> I'm mostly confused by what you have done and are asking. You have six
> theories. Were there 20 questions for each theory (a total of 120) or were
> the same 20 questions used for all six theories? What do you mean by
> "weight"? What would I look at on the output to assess weight? I think
> there are more questions to ask but please begin with these.
> Gene Maguin
>
> --- snip the original post ---





-----
--
Bruce Weaver
[hidden email]
http://sites.google.com/a/lakeheadu.ca/bweaver/

"When all else fails, RTFM."

NOTE: My Hotmail account is not monitored regularly.
To send me an e-mail, please use the address shown above.

--
View this message in context: http://spssx-discussion.1045642.n5.nabble.com/Factor-analysis-values-of-raw-component-covariance-matrix-tp5733238p5733242.html
Sent from the SPSSX Discussion mailing list archive at Nabble.com.

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD



--
Jon K Peck
[hidden email]

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD
Reply | Threaded
Open this post in threaded view
|

Re: Factor analysis: values of raw component covariance matrix

luisp
In reply to this post by Maguin, Eugene
Thanks to all for your reply. I don't have AMOS (just R + Lavaan) but my main goal with this analysis is to show which are the main theories (latent variables), relying on the FA.
I did the analysis with SPSS version 24 for Mac OS. Next you can see the results for Rational Choice Theory (13 items).
In total, there are 22 items (some of them are shared between different theories).