the data you have is way way too weak to draw any meaningful conclusion. You really should just stick with your priors here until you have much more data.
This. (Stressing it as an epistemic principle that’s under-applied within EA.)
This. (Stressing it as an epistemic principle that’s under-applied within EA.)