The problem with many, if not all of these studies is that they are biased and flawed from the beginning. A study needs large samples, long terms of study, double blinds to off set any recorder bias if possible, clear review of study material by people not in the study who can offer a honest evaluation of the results. Too often reliance on correlations as facts flaw the results and conclusion reached.

If you go into a study with a fixed viewpoint, it is too easy to set the study up in a way that almost excluded all data outside your desired findings. I have seen it often in studies my wife and I have reviewed for publication. People who do these studies can be intent in proving their viewpoint to the point that they loose perspective. I saw one study that was done three times before it "came out right" and they just threw out the first two data sets a being "flawed". The only reason I knew about the first two data sets was I had been asked to massage the numbers by one of the people who ran the first study. They could not be massaged and they got dumped.

Sample size is often overlooked. One site, two sites or even three sites is too small. Give me 50 to 100 sites for a large sample. One year is too short for most studies. Give me decades to see real trends. How many studies get ruined by one person interpolating all the data with a minor or major bias? A ten percent error in data may be a extreme swing in results in small sample studies.

My pet peeve is the correlations that can become facts if repeated often enough. Understand that most studies cite others before them as proof that they are on the right tract. If one of these studied is off then many of the studies after them can be off if they do not go back and prove the early studies were wrong. So an accepted but wrong fact gets repeated over and over again and becomes an accepted fact when it is not a fact at all.

For example if I published the "correlation" that 81% lung cancer victims carried butane lighters for 20 plus years and draw the conclusions that butane lighters are bad for your health others will cite this "fact". Repeated often enough and that fact will get lost in the other facts and become a commonly held view based on long accepted studies. Look at the data first, look at the way the study was done, look at the number and quality of other studies cited. Then go one step further and look at those studies and see if any of them cite their own works or those of the first study you were reading. I have seen studies cite back and forth so often you that you get to feeling that all the studies were done a just a very few people and they seem to almost be working out of one office or with just one funding source. The money dictates the outcome. You would not pay for a study if it does not support your viewpoint.

Any study that cites their own previous publications is to be taken with a grain of salt. Any study who cites the same two or three previous authors over and over again is to be taken with a larger grain of salt. Any study done in an area the size of your backyard, in just one year, that does not include more than 30 test subjects is to be taken with a even larger grain of salt. Any study who does not tell you the person or group funding the study is not to be trusted at all. Follow the money, find the reason for the study.