There's a way to write a bias-free science paper and a way to determine whether it's biased or not. Generally speaking, scientific papers should be free of bias, but publication bias seemed very difficult to avoid especially when the researchers conducting it are paid for it or have been reaping rewards to arrive at a positive outcome.

Science' foundation is published evidence, ICSU reported. That's why students of science must conscientiously avoid bias by exercising careful judgment and maintaining the integrity of the science literature. The responsibility gathering and interpreting data and submitting papers to journals must be strived for.

While these pointers are common sense, even the brightest fail to follow the procedure. Young scientists who do not want their authored science papers to be discredited or make it indicate that they've done shoddy work must at all cost avoid bias.

According to Stanford University's Dr. Daniele Fanelli, there's a way for professors or editors to identify scientific papers with bias results or papers that announce large effects, IFL Science reported. Focusing on statistical analysis across 22 disciplines to compare the results of different studies that answered the same question, Fanelli said there's only a small bias or overstatement of effects.

Scientists can be rewarded for saying something wrong. There's the promise of career advancement once one's paper gets cited regardless if it reported on large effect sizes or not. That's the danger that scientific papers should avoid as per the Proceedings of the National Academy of Sciences in their paper entitled "Meta-assessment of bias in science."

Fanelli's paper found that early career researchers have published exaggerated results that's been cited by later researchers. Small teams of scientists are also prone to get their papers discredited than those who have large collaborations or have been funded by large corporations. But its relationship to the credibility of the author with the volume of publications he or she is involved is non existent per Fanelli's conclusion.