forum.smartpls.com

good evening everyone

this is the first time i'm using smart pls for pls-sem, i'm working on a study of the adoption of medical technologies "the case of the electronic medical record" by healthcare professionals in france.

I administered the questionnaire to the target population, i collected 390 responses, i noticed that i have two categories of respondent, users of the electronic medical record (310 responses) and non-users (80 responses).

So, my question is as follows: how should I deal with the 80 responses from non-users, given that I'm faced with missing data because of this category?
Should I eliminate the non-user responses and work with the data from the 310 respondents?[/b]

thank you in advance for your comments

You can also include them and use pairwise deletion as the missing value treatment. This way every relationship that has only users is estimated with the data of users only. If you have data for both the relationship is estimated for both.