A threat to construct would mean that there was a failure in the way that constructs -- or thoughts -- were formed. In the case of e-learning, if we were to show that e-learners are more open to not just non-traditional learning, but they are more open to non-traditional ideas, in general, a threat to construct validity would probably say that e-learners are not more open than any other learners and even if they are open to other types of learning, they are not more open to non-traditional ideas in general. A threat to construct validity would keep the two totally separated.

Validity issues could impact the research in a few different ways. First of all, there is much discussion about the efficacy of e-learning and external validity could generalize to a point that...

As well, the controversy involving construct validity would be brought up to question. E-learning is such a relatively new topic and thinking that it is efficient as traditional learning may be too far ahead of its game right now.
Therefore, external and construct validity could pose problems when it comes to researching e-learning and traditional learning. There is not room for generalizations right now in the research.