Dear all,
For my PhD courses I'm asked to perform a CFA in SAS.
The task description last two questions concerning the CFA are the following:
In the SAS data file "psy.sas7bdat" (or in "psy.txt" on the website) you can find the correlation matrix based on nine psychological tests which were administered to 145 children. The variables are denoted by the labels v1 through v9.
The hypothesis is that these are indicators of three latent factors which are called ‘Visual’,
‘Verbal’ and ‘Speed’ (v1-v3 for ‘Visual, v4-v6 for ‘Verbal’ and v7-v9 for ‘Speed’). The factors are assumed to correlate.
Does this model fit the data?
Test whether the loadings of v5 and v6 are the same.
Test whether the correlation between the factor ‘Visual’ and the factor ‘Speed’ = 1.
TIP: Sometimes it can be necessary to increase "maxiter" in the proc calis procedure in SAS to achieve convergence.
does anyone know how to solve the two last questions (in bold)
To read the data you can use this statement:
filename webdata url "http://feb.kuleuven.be/martina.vandebroek/public/Statdata/psy.txt";
data testimport (type=corr);
infile webdata;
input _TYPE_$ _NAME_ $ v1 v2 v3 v4 v5 v6 v7 v8 v9;
proc print data = testimport; run;
Does anyone know how to test whether two loadings are the same, and how to test correlation between two latent factors in the CFA?
Thanks in advance,
Maarten
