How to defend a significant association of demographic on uni- dimension construct but not significant association with muti dimensional construct?
Can you tell us more about your independent and dependent variables, and what was specifically involved in your comparison of uni- and multi-dimensional constructs?
Hi Sir,
I am doing the study on the customer Satisfaction in telecom sector (Uni and Multi Dimensional constructs) and association with demographic like occupation and education.
Customer satisfaction constructs are network, value, care, experience and delight.
The result of the study reveals that occupation and education plays a significant role satisfaction (multi-dimensional model), but It is to note that occupation and education is not significantly associated with the network, value, care, experience and delight.
How to defend this , Kindly assist
Regards,
If you are creating a scale from your separate variables, then it will have a higher reliability than any of the individual variables. This single, more reliable indicator will also have a higher correlation with the independent variables, because it has less random error than the individual variables.
One analogy is to think of the individual variables as separate sticks and the multi-item scale as a bundle of sticks, which is stronger than the smaller sticks it is made from.
- Badges
- Science topic