It seems that Smartpls considers the missing value as a valid value in the dataset and includes it in the calculations. I faced discrepancy in correlations and alpha scores between spss and smartpls. However, when I selected the mean replacement function for missing values the correlations and alpha scores were almost similar in both softwares.
Furthermore, to test if smartpls considers -1 in its calculations as a valid value, I used the dataset with the -1 values in spss, since spss will consider the -1 as valid values. The correlations and alpha scores in spss looked similar to those in smartpls when I deselect the "apply missing value algorithm"
I hope my idea is clear
smartpls considers -1 as a valid value
-
- PLS Junior User
- Posts: 5
- Joined: Thu Sep 05, 2013 8:53 pm
- Real name and title: