If all they're doing is describing a trend, then I understand your subsequent points. I will be watching for a secondary analysis that performs a cross-sectional regression.Thanks for this comment. I agree with you to an extent. I think it depends on what kind of study we are talking about. In the case of a study where the underlying model is the measurement of latent constructs and the aim is to relate them to each other, such studies should indeed begin with a theoretical framework and set hypotheses. However, in the case of the MixMag Survey, the researchers are doing something different - they are taking a convenience sample of drug users from year to year and describing their use patterns and harms. There will likely be some planned analyses where questions are repeated year to year, but also some additional questions specific to this sample that will simply be exploratory/descriptive.
Agreed. I just haven't figured out if the methodological compromises (that potentially doom it to to an unsubstantiated headline in the print media) mean that it is a worthwhile exercise.One of the concerns with large surveys using convenience sampling is that people will be wowed by the numbers without considering the sample frame is still convenience. Still, year to year, we can get a general sense of trends, and for this application, it is good enough in my opinion. You also have to ask what other option do researchers have?
If all they're doing is describing a trend, then I understand your subsequent points. I will be watching for a secondary analysis that performs a cross-sectional regression.![]()
Part of me wishes researchers avoided the low-hanging fruit like this. Telling us whether the consumption of drugs that the majority of the public have never heard of, might be going up or down (possibly with inappropriate CIs?), based on a sample of people who are either self-selecting or incapable of filling in a bloated form, doesn't really move the debate forward.
Or does it?
I'd like to see a survey that targets people who use research chemicals carefully or use psychedelics therapeutically.
It's something we could make happen too![]()
Xorkoth, the former PD senior mod who built the massive index, had a survey in the works. He even showed us draft versions of it. It was going to be a pretty long and comprehensive survey.
Now that's an idea. I hadn't even thought of a 'Bluelight-produced' survey. I suppose if magazines can host surveys and publish the results, there's not much reason a harm reduction forum couldn't either.![]()