I came across a quote from biostatistician Andrew Vickers that I really like:
A mistake in the operating room can threaten the life of one patient; a mistake in statistical analysis or interpretation can lead to hundreds of early deaths. So it is perhaps odd that, while we allow a doctor to conduct surgery only after years of training, we give [the SPSS statistical software] to almost anyone.
I can think of a number of caveats. For example, if you’re an amateur trying to conduct a statistical analysis of some phenomenon where no statistical analysis currently exists, maybe you should not let the perfect be the enemy of the good by suppressing your analysis altogether, if it means people will continue thinking about the phenomenon in an intuitive and non-data-driven way.
But I do think it’d be valuable for the EA community to connect with or create more EAs with a deep understanding of statistics. Improving my own statistical skills has been a big project of mine recently, and the knowledge I’ve gained feels very generally useful.
Curious what you’ve been doing to tool up on statistics.
Here are some resources I’ve been consuming: