WAIS prorating?

[Amina]

I realise the short answer to my question is to go and read the manual, but I won't have access for a few days...

So, a client gained a raw score of 0 on symbol search - I didn't substitute a subtest as I didn't realise at the time. Is it possible to prorate by just using scores from the other 9 subtests (and the table in the manual?), or is prorating only for working out scores on composites?

If not possible, I can just do the GAI.


Thank you in advance

[alexh]

What happened?

[miriam]

A score of zero might be useful data, if it represents a genuine large deficit. I'd only substitute if the subtest was invalid for some reason, or pro-rate if it was not administered.

[BenJMan]

Processing speed can't be prorated because it is only comprised of two subtests, you would need a substitute subtest (i.e. cancellation). You can just calculate GAI and report on the working memory and coding seperately, but there is a question of whether this is in the best interests of testing and interpretation, or done out of convenience.

As Miriam says substitution and prorating should also only be done if there is a non clinical reason why they scored a 0 on this subtest, there could be a good reason for the difference, in which case it becomes part of your assessment and write up just as much as a higher score.

[BenJMan]

That said I have a vague memory that you can calculate a FSIQ on 9 subtests somewhere in the manual... but I'd need the manual and possibly technical manual to know how reliable that is and what circumstances it is appropriate for..

[BlueCat]

I may be wrong, but I thought you would only use pro-rating where a test is spoilt (e.g. fire alarm, you get the procedure wrong, other incident) or cannot be administered? Did you correctly administer the sub-test, i.e. with the correct instructions and using the sample items to check the respondent understood the task? If you have correctly administered the sub-test, and the respondent has achieved a valid score of zero, then that is the score, surely? If the respondent has achieved a score of zero under valid administration conditions, then that is what you report, surely? Otherwise how is that any different to leaving out/discounting other subtest scores at a threshold above zero? I'm pretty sure that on most subtests, zero is a valid score and translates to a scaled score?