[Frannie Noble]

This question come from our recent IDELA webinar about Translation and Adaptation.

When you change what constitutes and acceptable answer how do you know the validity and reliability of that? Or do you now include data from adapted tests in ongoing discussions on reliability and validity?

Topic

[Nivida Chandra]

Hi,

Since there is no norm against which we compare the data, the questions of reliability and validity hold a different meaning in this context. I know Jonathan has responded to this in better detail, and I entirely agree with him.

I’d just like to add that one way of making sure that your adaptation is sound and evenly scored is inter-rater reliability. And to keep that going for as long as possible.
In using it in Delhi (India), we kept a close eye on the data and every enumerator was double scored 4-5 times. If the master trainers / supervisors felt they needed closer attention, then even more.

[Author]

Replies