Hi Hieu,
I agree with your point that convergent validity should come first because a construct needs to be measured clearly before we compare it with other constructs. ...
I agree with your point that convergent validity should come first because a construct needs to be measured clearly before we compare it with other constructs. ...
Hi Hieu,
I agree with your point that convergent validity should come first because a construct needs to be measured clearly before we compare it with other constructs. Your explanation about weak convergence making later analysis unreliable is very convincing.
In my own research on HUFLIT students’ perceived English speaking fluency, I also think convergent validity is especially important at the beginning. If the items in my fluency scale do not work well together, then it would be difficult to say that the construct is being measured properly. At the same time, I also see your point that discriminant validity is still necessary, especially when two constructs may overlap in meaning. So overall, I agree that both matter, but convergent validity forms the foundation.
I agree with your point that convergent validity should come first because a construct needs to be measured clearly before we compare it with other constructs. Your explanation about weak convergence making later analysis unreliable is very convincing.
In my own research on HUFLIT students’ perceived English speaking fluency, I also think convergent validity is especially important at the beginning. If the items in my fluency scale do not work well together, then it would be difficult to say that the construct is being measured properly. At the same time, I also see your point that discriminant validity is still necessary, especially when two constructs may overlap in meaning. So overall, I agree that both matter, but convergent validity forms the foundation.