A very common type of this involves running many experiments, or conducting multiple analyses on the same data, and reporting only the results that are statistically significant, or those that are consistent with the predictions of the researcher.
type I error
p-hacking
type II error
alpha error
