Bias Hurt Us

Bias is simply defined as “prejudice in favor of or against one thing….” In Academia, we are inherently guilty of it. We research the topics that we subjectively find interesting, for investigators who have spent a good portion of their lives in fields related to their interests. Is is it bad? Is it required to even survive in academia? To perform favorable experiments, for favorable results, to tell good, convincing, fundable stories? Can it be both objectively bad and culturally necessary? I don't know. When it applies to finding inspiration, motivation, and a will to exist in your respective field, I think it can be good. Can that have an overall impact on the validity of your results? As hopeful professionals, I hope not. I do believe there is a level of maximal objectivity we should strive for in any kind of research. 

Bias can also be represented by a lack of consideration, and/or misinterpretation, that often leads to overestimated conclusions. This can indeed have significant consequences for entire fields, and can be largely intertwined with the problem of irreproducibility. 

For example, in social neuroscience, Oxytocin is generally believed to have a variety of influences on social cognition, particularly in reward and reinforcement learning. However,  In his article (Statistical and Methodological Considerations for the Interpretation of Intranasal Oxytocin Studies), published in a 2015 issue of Biological Psychiatry, Dr. Hasse Walum, a post-doctoral researcher at Emory, highlights the major issue of effect size inflation due to low sample sizes and low statistical power, and also points to its negative influence on post study probability (the chances of finding a true effect) given the pre-study odds of a synthetic version of the peptide, of which’s mechanism of action was ironically criticized (Intranasal Oxytocin: Myths's and Delusions) in the same issue by a different group. 

Bias thus becomes a negative influence on a field when people continue to publish significant findings, in high impact journals, with findings tied to small samples sizes and effect size overestimates. Many can agree this is what happens. But I guess these tend to be the stories we read most, and the stories many will either blindly or intentionally tell to have a chance at a major impact.