3 ms·

> After three years, 16 Finnish children and 14 Estonian children had these antibodies; only four Russian children did. I am not a statistician, but wouldn't t

• LionessLover 10y ago
  Depends on effect size and variance. https://www.ma.utexas.edu/users/mks/statmistakes/FactorsInfluencingPower.html https://www.ma.utexas.edu/users/mks/statmistakes/FactorsInfl... A study with low power is perfectly alright for an exploratory study that is not meant to prove anything but to check if it's worthwhile to spend a lot of resources on an idea in the first place. http://lynn-library.libguides.com/researchmethods/researchmethodsd8 http://lynn-library.libguides.com/researchmethods/researchme... > 1. Exploratory research generally utilizes small sample sizes and, thus, findings are typically not generalizable to the population at large. > 2. The exploratory nature of the research inhibits an ability to make definitive conclusions about the findings. > ... The problem is when exploratory studies make it to the media and people confuse them with studies meant to settle questions rather than just raise them.
• thaumasiotes 10y ago
  As an illustration of LionessLover's point, compare "under treatment, two patients regrew their lost arm, compared to zero in the control group". The statistical power of that study just isn't relevant.