I did PCA for some variables which were Likert scaled questions (1-5). After predicting, I get negative for some of the values. Is there something wrong? Do I take absolute values of the predicted variable ?

Note: The Eigen values were positive

The values in the new variable (after predicting) is given as attachments

The stata output are as follows:

pca v1 v2 v3 v4

Principal components/correlation Number of obs = 302

Number of comp. = 4

Trace = 4

Rotation: (unrotated = principal) Rho = 1.0000

-------------------------------------------------------------------------------------------------------------

Component | Eigenvalue Difference Proportion Cumulative

-------------+-----------------------------------------------------------------------------------------------

Comp1 | 1.50368 .51615 0.3759 0.3759

Comp2 | .987535 .16249 0.2469 0.6228

Comp3 | .825045 .141311 0.2063 0.8291

Comp4 | .683735 . 0.1709 1.0000

-------------------------------------------------------------------------------------------------------------

Principal components (eigenvectors)

---------------------------------------------------------------------------------------------------------

Variable | Comp1 Comp2 Comp3 Comp4 | Unexplained

-------------+----------------------------------------+------------------------------------------------

v1 | 0.6017 0.0522 -0.3627 -0.7097 | 0

v2 | 0.5426 -0.4256 -0.3741 0.6200 | 0

v3 | 0.5034 -0.1359 0.8531 -0.0192 | 0

v4 | 0.3001 0.8931 -0.0273 0.3340 | 0

--------------------------------------------------------------------------------------------------------

More Tazkeer Azeez Chaudhuri's questions See All
Similar questions and discussions