On presenting Evidence and unboxing science

The first question we faced as we gathered to discuss the possibility of launching Evidence (then just called the new journal on evidencebased medicine, still no caps) was why (and if) we did need another medical journal in a highly competitive environment already full of those. Douglas Altman stated that we needed “less research, better research and research done for the right reasons” in his famous 1994 article “The Scandal of Poor Medical Research”1. Adding to that, the predictive value of research in general probably lies below 50%2, due to imprecision (low sample size), methodological bias (low quality of research design) and allegiance bias (conclusions being crafted in a manner that reconciles best with the investigator’s or researcher’s perspectives and preferences), not to mention P-hacking3, withholding datasets and publication bias, that is, publishing only or mostly papers that disprove the null hypothesis.