Often in a debate comparing two worldviews one side or the other will make a statement and back it up with some short of scientific study. Then they expect the other party in the debate to capitulate, because science is fact and you cannot argue with science. Often neither person capitulates and the debate continues anyways; sometimes, because the opposing party can produce a scientific study to support their side or they simply disregard the study. However, this raises a question: Are scientific studies absolute fact and is it illogical to disregard one?
To answer this question we can look at how often inaccuracies and fraud have been detected in scientific studies, then we have to ask ourselves if we are detecting most of it or only a small portion? Since humans participate, analyze, and draw conclusions as part of a scientific study, and human behavior is somewhat predictable, we can safely say that a lot of invalid scientific studies (due to accidental errors or fraud) go undetected.
Consider this excerpt and read the whole article for yourself:
"As a result, fraud in science is considered by many to be endemic. Biological research is one of the chief areas of concern. Some conclude that over 10% of all researchers in this area are less than honest. Indeed, probably most researchers have quoted data that are fraudulent, or at least inaccurate. Few extensive research investigations on fraud under the present system exist (and the cases unearthed probably represent only the tip of the proverbial iceberg)."
Source: Science fraud epidemic