The ones most suspect are those relating to food habits and health. I am tired of reading these results; and the frequency with which these studies have been held. The vast differences in the conclusions confuse rather than enlighten any reader.
Take consumption of water for example. We have always been told that we should drink lots of water, ideally it seams at least 8 glasses every day. (Source: Parent Jazz.) Day before yesterday, there was this bit of news that quoted a study by scientists saying drinking lots of water may not of any use. (Source: BBC and Rediff).
Another is sex surveys, and these come out at very frequent intervals. And, quite understandably command high readership. Again day before yesterday, AP put out this story quoting sex therapists, which was picked up a number of media outlets, saying "the optimal amount of time for sexual intercourse was 3 to 13 minutes". (Source: AP feed on Yahoo News). Here is another one: Italians prefer football to sex. (Source: PTI story in DNA.)
The problem is that all these surveys and studies sound very conclusive, there is a ring of certainty at least for a casual reader - very much like the astrological predictions column. What is missed by most of us, is that a number of parameters determine the conclusions. The findings might be right only for the sample examined.
The validity of a survey largely depends on the how well the sampling has been done. Contrary to popular notion it's not the size of the sample that matters, it's how representative the sample is what matters. And, most of these surveys hardly elaborate on the sampling details.
Matters of health are very individualistic. What science give us is broad guidelines in idealistic situations. What is okay for one person need not necessarily be okay for another; and vice-versa. So, let's not jump to conclusions reading these random surveys, mistaking them for universal application. We should take a call on how well or not the findings are applicable to each one of us.