Sometimes, not all that much. Google for Dr. John Ioannidis -- this guy has made a career of going around and informing the public when people do things like screen a disease against thousands of genes and, by sheer chance, hit a type 1 error in the process that they think is significant. No malicious intent, it's just honestly not knowing any better.
1. That's why most studies go through peer-review processes in order to get published in (respectable) publications and 2. After Googling Dr. John Ioannidis, I came across http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1182327/?tool=pmcentrez wonderful paper.
Not only is this paper a complete rehash of what any good statistician already knows, it's basically outlined like this:
"Math... basic stuff... math... generalization... math... money is evil... math... people are evil by repressing refutation of their studies... math..." (note: Obvious hyperbole is obvious.)
Basically his overall point is this: Large studies are good. Randomized studies are good. Multiple, independent studies that examine the same idea / hypothesis = great.
... so... what just about every researcher already knows.

(Edit: Though, as you mentioned, Dr. Ioannidis may be writing this for the layman... and if he is... well, it's not likely coming across well.)
Also, as promised, here is one study that talks about the amount of exposure children have to endotoxins and other such things and how it may reduce the level of allergic reactions in children:
Schram, D., Doekes, G., Boeve, M., Douwes, J., Riedler, J., Ublagger, E., & ... Brunekreef, B. (2005). Bacterial and fungal components in house dust of farm children, Rudolf Steiner school children and reference children--the PARSIFAL Study. Allergy, 60(5), 611-618. Retrieved from EBSCOhost.