When my respondents are from adopters firms. The alpha and CR value of the Construct 'Actual Usage' was 0.732 and 0.8. After inclusion of non-adopter in the data set, The alpha value and CR value of the construct 'Actual Usage' comes down as 0.455 and 0.422. Can i keep the construct for the further study. the convergent and discriminant validity of the construct is acceptable and satisfy all the criteria. Kindly suggest me.
Can you explain little bit what are you trying to do in your study ? whats the objective?
I wonder if the problem is that some of the the items can't be answered properly unless you are in an adopter firm. If the scale measures some aspect of usage, people who aren't users might have a hard time answering, and items won't relate to one another.
- Badges
- Science topic
- Similar topics
- Psychology
- Psychometrics