Factor Analysis

[ArjumandBano]

Hello,

Recently I conducted a pilot study. The sample size was 36 and the Questionnaire contains 41 items. I ran PCA but unfortunately the out put showed that the correlation matrix is not positive definite. The one reason I understand is the number of cases X variables. Is there any other way to validate the instrument? or please suggest me something as I am not very much good in statistical methods and techniques.

Thank You

 

[GretaGarbo]

Maybe you have a few missing values. That can cause it to be non positive definit.

You can do a simple test and just run it with say 30 items, and see what happens.

 

[ArjumandBano]

Thanks for your reply let me test this way.. I made sure there are no missing values..