BCSnob wrote:
The advantage of published studies is that they have been reviewed by others in the field for errors in methodology or interpretation of results.
Correlations are easy to establish and well thought-out surveys can find correlations; however, correlations do not prove cause and effect. This is a major mistake made by most people. Follow-up well controlled studies (no undocumented variables that are present in epidemiological studies/surveys) are REQUIRED to prove cause and effect.
I have absolutely no problem with controlled published studies. Makes my life easier to research. :B
BUT: published studies are done AFTER suspect clusters or cause is shown to be of interest to 'warrant' the controlled study.
Controlled studies are NOT conducted nor funded without probable suspect or cause, via the ever important studies/surveys/clusters.
To live by 'only' published studies is turning a blind eye to what is and could be very important information to a specific issue before presenting and/or funding is available for controlled studies to even begin.