There is a rising chorus of calls for science to become more reliable, replicable, and reproducible. Many of these demands focus on increasing statistical power by collecting larger sample sizes. However, science is not a heterogenous pursuit, and larger sample sizes may not make sense for many types of research problems. In this talk, I describe why improved statistical power solves many research problems, and what other tools are available to you when a larger sample size does not make sense.
Data Scientist (2018-2020)
Events in September 2019
Part of CMN Presentation Series