How to perform a power analysis

This talk gives you the low-down on power analyses for research. I discuss what they are, why they're an integral part of study planning, and how to perform them. I also share ways to define your effect size of interest and why you avoid using past studies as benchmark for your power analyses. Talk slidedeck: https://osf.io/uk5bp/ This talk was given to the Student Initiative for Open Science (SIOS) community on February 12, 2021 --------- CORRECTION: In slide #7, it was originally written that "In other words are you happy with a false positive error rate of 20%?". This has should have been, "In other words are you happy with a false negative error rate of 20%?". It was also written that "If you’re trying to replicate a study, a false positive error rate of 20% is probably too low". This has should have been, "If you’re trying to replicate a study, a false negative error rate of 20% is probably too low". The slidedeck has been updated with these correct sentences Thanks to Cyril Forestier for pointing this out this typo!