Bad Science: Cause and Consequence

Scientific progress is dependent on accumulation of quality data with appropriate data analysis. Unfortunately, there are a troubling number of accounts describing an inability to replicate published work. Some explanations are lack of access to proprietary reagents and equipment, or lack of expertise and know how. However, it is clear that there are many publications that are fatally flawed, and it is difficult to ascertain which ones they are, but there are clues. Many articles are improperly controlled, resulting in false-positive or -negative results. Reagents and procedures are used without verifying their specificity. There is also confirmation bias, a tendency to seek and find conclusions that we like, which is exacerbated by faithful acceptance by readers of the publication record without assessment of merit. These and other issues have slowed progress, resulted in waste of scarce funds, and even put patients at risk when clinical decisions are made according to flawed data. Solving these and related problems requires recognition of the problem and better training. We also need to take personal responsibility for not only our own work, but also for the accuracy of information in the scientific domain.