 If using a p-value of 0.05 to suggest a discovery, it is likely that the result is incorrect at least 30% of the time due to underpowered experiments. This conclusion is supported by multiple sources including tree diagrams and repeated simulations of t-tests. Additionally, simulations demonstrate how effect sizes can be overestimated when experiments are underpowered. A script is provided to allow readers to conduct their own simulations with relevant data. It is recommended that researchers should aim for a false discovery rate, FDR, of less than 5%, or insist on a p-value of less than 0.001 to ensure accuracy. This article was authored by David Kahoum.