Skip to main content
. 2020 Jan 17;11:358. doi: 10.1038/s41467-019-14203-0

Fig. 9. Effects of perturbations of the idealized scientific community.

Fig. 9

a and b show the effects of p-hacking on the number of experiments conducted and the number of effects discovered. p-hacking is operationalized by raising the type I error rate, α=.2. c and d show the effects of keeping α=.05 but adding up to five batches of additional participants (each batch containing 1, 5, or 10 participants), if an effect failed to reach significance. Results are shown averaged across levels of temperature.