Question related to standardization in PROC FACTOR Method=principal

[berkowr]

Hello!

I am using principal component analysis in SAS V.9.4 for the first time, and I noticed that when I output the first factor it was standardized (mean of 0, standard deviation of 1). However, the continuous variables I entered into the original procedure were not centered originally. In reading through SAS procedure guide for PROC FACTOR, I am having trouble telling whether or not I should have included NOINT in the data step (to indicate that the variables being combined were not centered already) or if PROC FACTOR standardizes the variables as a part of the behind-the-scenes programming. My concern is that the results are inaccurate (without including the NOINT data option) because the variables were not standardized before being included in the procedure but it is assumed that they are.

Page 6 of this SAS-related resource seems to suggest that the variables are transformed through the PCA procedure and are then standardized, perhaps explaining why the resulting factor is standardized -- https://www.sas.com/storefront/aux/en/spsxsfactor/61314_excerpt.pdf

Could anyone clarify this? Below is the code that I used which resulted in a standardized factor 1:

proc factor data=SET corr scree ev method = principal out=SET_FACTOR nfactors=1;
var item1 item2 item3 item4 item5 item6 item7 item8;

run;

[PaigeMiller · Accepted solution]

I believe both PROC FACTOR and PROC PRINCOMP default to standardizing your input variables to have a mean of 0 and a variance of 1.

Unless you have a very good reason for including NOINT in the code, I would not use NOINT.

[berkowr]

Thank you for clarifying!