All Access forum API Error

Net Talk • Music Test Respondents Wrong?
Page 1 of 1

Music Test Respondents Wrong?

Posted: Thu Feb 12, 2015 9:45 am
by rogerwimmer
Doc: A research company recently did a music test for us and I went to watch the procedure. I’m not suggesting that the company did anything wrong, but it seemed to me that a few of the respondents didn’t belong in the test. Is there any way to check the results to identify respondents who (maybe) didn’t belong in the test? - Anonymous

Anon: Yes, there is a way and it’s a fairly simple procedure. If you don’t know how to do these things, then ask your research company to provide the information.

Here is what you need to do:

1. Compute means and standard deviations for each respondent (all scores for every song).
2. Compute Z-scores for the means and standard deviations.
3. Look at the Z-scores. First check to see if there are any respondents with Z-scores for mean or standard deviation that are ±2.00 or greater. If so, delete those respondents. If there are none in that category, then look at (and delete) respondents with Z-scores of ±1.50 or greater.

What this does is delete respondents who are significantly different from the rest of the group. This happens every so often, so don’t be alarmed. If you can’t do these computations, call your research company. The whole computation process takes about 10 minutes.

Please let me know if you have any other questions.

(Want to comment on this question? Click on the POSTREPLY button under the question.)