There certainly were plenty of remarkable discoveries this past year, my favorite demonstrating that running promotes the generation of new neurons in the aging brain. That prompted me to get back on the treadmill after a hiatus of more than a year. But the big scientific news story for me was not any single event. Rather I have been struck by the emergence over the past several years of two seemingly disparate yet related trends in the scientific world. Neither of these has made front-page news, although both are well known to those of us in the science business.
The first of these is the apparent increase in the reported incidence of research findings that cannot be replicated. The causes for this are myriad. In some cases, this is simply because some vital piece of information, required to repeat a given experiment, has been inadvertently (or at times intentionally) omitted. More often, it is the result of sloppy work, such as poor experimental design, inappropriate statistical analysis, or lack of appropriate controls. But there is also evidence that scientific fraud is on the increase. A number of sensational exposes have come to light this past year, and the incidence of retracted, withdrawn and corrected scientific papers has increased steadily over the past decade. Indeed, some pharmaceutical companies have made the decision to no longer rely on the results of published studies because they feel that in many cases these are not trustworthy. Some have argued that the increase in the incidence of scientific misconduct reflects the use of new technologies for uncovering scientific misconduct, such as programs that check for plagiarism. Technological advancements have certainly played an important role, but in my opinion are unlikely to explain entirely the failure to replicate deluge.
A more compelling explanation is the fiercely competitive nature of science that has accelerated tremendously in recent years. Grants are much harder to get funded, so that even applications ranked by peer review as “very good” are no longer above the pay line. This means that faculty at research universities must spend the bulk of their time writing grant proposals with the probability of funding often being less than ten percent. At the same time, the most highly ranked scientific journals have increased their rejection rates considerably with acceptance rates in the neighborhood of five percent being not uncommon. Moreover, to get a final acceptance for publication in the top journals, the editors (based on the recommendations of anonymous referees) often request additional experiments requiring considerable more time, expense and effort without any guarantee that the final product will be accepted for publication. For practitioners of science the level of stress has never been greater. Is it any wonder that some will succumb to a short-cut method to success by “fudging” their results to get a competitive edge.
I am not suggesting that simply increasing the amount of available funding for research would abate this problem. The solution will require a multi-faceted approach. One step that should be seriously considered is a substantial increase in the penalty for scientific fraud, with jail time for those that have wasted precious research dollars. Today, those that are caught often get away relatively unscathed, in some instances receiving a substantial buy-out package to leave their place of employment due to the nature of the university tenure system. This trend must be reversed to ensure the viability of the scientific enterprise.