User: wiwa

Re:Web based survey

On: American Workers: Lazy or Creative?

Amazingly, the research indicates that the usual means of selecting respondents (they usually don't go for purely randomized samples, but stratified samples; i.e. a fair selection of people distributed over different variable - such as age group, income, etc. - in the same pattern as the general populace) actually holds up well. People who don't respond usually only do so because they don't feel like it at the time.

This may be true much of the time, but in this particular case, I don't think it does. What you've basically said is that they've shown that the probability of responding to these surveys is statistically independent of the response itself. In this case, however, they're using a web-based survey to measure how much time people "waste" at work. The problem is that a web-based survey is much more likely to turn up people who spend a lot of time surfing the web--likely when they should be working!