Factor Analysis
Locked
 
|
Hi all,
I am doing EFA for 15 variables. Each variable range from 0 to 56. The goal is to reduce the 15 variables into a few number of factors. I am using Principal Components using varimax rotation. Any constructive comments about the method I used are welcome. Thank you. Eins --------------------------------- Get your preferred Email name! Now you can @ymail.com and @rocketmail.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 |
Re: Factor Analysis
|
|
It is hard to say much not knowing more about your research. However, i will say that i have found few real world applications for factor solutions that result in uncorrelated factors. I find varimax and other orthogonal solutions to be rather artificial and i don't use them without a compelling reason. Most of the factors we see in nature and most certainly in the social and behavioral sciences are intercorrelated, even if only minimally. I find it more realistic to preserve this intercorrelation by applying an oblique rotation. My 2-cents worth on this hectic election day Mark *************************************************************************************************************************************************************** Mark A. Davenport Ph.D. Senior Research Analyst Office of Institutional Research The University of North Carolina at Greensboro 336.256.0395 [hidden email] 'An approximate answer to the right question is worth a good deal more than an exact answer to an approximate question.' --a paraphrase of J. W. Tukey (1962) To: [hidden email] |
|
In reply to this post by E. Bernardo
A lot depends on what you are doing and what you are going to do with
the results. Are you trying to create scales for future use? Are you trying to look for variables that go together to guess at underlying constructs? Is the unique variance of a variable part of what you want to account for? Is divergent validity important to you? What is the nature of your variables? Why do you consider them to be a domain? What is the nature of your cases? Is it possible that you have the relatively rare situation where you want your measures to be correlated? Is your set of cases large enough that FA is reasonable? Art Kendall Social Research Consultants Eins Bernardo wrote: > Hi all, > > I am doing EFA for 15 variables. Each variable range from 0 to 56. The goal is to reduce the 15 variables into a few number of factors. I am using Principal Components using varimax rotation. Any constructive comments about the method I used are welcome. > > Thank you. > Eins > > > --------------------------------- > Get your preferred Email name! > Now you can @ymail.com and @rocketmail.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
Art Kendall
Social Research Consultants |
|
|
In reply to this post by Mark A Davenport MADAVENP
Hi All,
Practically, Mark is right that most of the factors we see in nature and most certainly in the social and behavioral sciences are intercorrelated, even if only minimally. And that it is more realistic to preserve this intercorrelation by applying an oblique rotation. Follow-up question: any idea on the appropriate extraction method for continuous variables? Is Principal component analysis "OK"? Thank you. Eins Mark A Davenport MADAVENP <[hidden email]> wrote: It is hard to say much not knowing more about your research. However, i will say that i have found few real world applications for factor solutions that result in uncorrelated factors. I find varimax and other orthogonal solutions to be rather artificial and i don't use them without a compelling reason. Most of the factors we see in nature and most certainly in the social and behavioral sciences are intercorrelated, even if only minimally. I find it more realistic to preserve this intercorrelation by applying an oblique rotation. --------------------------------- Get your preferred Email name! Now you can @ymail.com and @rocketmail.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 |
«
Return to SPSSX Discussion
|
1 view|%1 views
| Free forum by Nabble | Edit this page |