Sunday, May 09, 2004

low on weblog material, he decides at least to publish an idle graph from his ongoing research

On the WLS* mail survey, there is this 10-item scale regarding problem drinking. It contains one of my five least favorite questions on the entire survey, "Does having a few drinks help decrease your shakiness or tremors?" The question is annoying because it seemingly presumes the respondent does have shakiness or tremors, which of course most respondents don't. I complained about this question at every opportunity, even though no revision could really be done since we wanted to replicate the standard scale (the S-MAST-G) of which it is part. Lo, the graph above shows the % of respondents who skipped each of the ten items of the S-MAST-G (excluding those who either rightly or wrongly skipped the S-MAST-G entirely). I will leave it to the reader to guess which of the above is the shakiness or tremors item.

Item #10 is perhaps a lesser example of the same kind of problem, incidentally. The item is "When you feel lonely does having a drink help?," which presumes the respondent sometimes feels lonely. The morose and worldly among us might think that everyone feels lonesome, sometimes, although I'm not so sure, especially for the purpose of answering survey questions.

* Wisconsin Longitudinal Study

No comments: