Miró's Control Group Size Calculator Now Available
Rather surprisngly, the first component of Miró to be released is a control group size calculator. You can use it here. I'll blog a bit about what it does and why it exists over the coming weeks. There are help links for descriptions there too.
There's also slightly more information about Miró itself now online here.
Labels: controls, data errors, miro, targeting
posted by njr at 21:29 PERMALINK
3 Comments:
I dont think this is working. Entering for example:
pop size of 540 and T-C=4 and highest likely rate in control group =24 yields no answer. But if I switch T-C and control group I get answer. Are the two labels switched?
Hi Brian
I think it's OK. The reason it reports N/A is because 540 isn't a large enough population size to allow you to detect a 4 percentage point uplift against a control rate of 24% (i.e. treated 28%, control 24%). It gives a result when you swap the figures because you don't need much of a population to have confidence that a rate an observed rate in the treatment group of 28% is significantly different from an observed rate in the control group of 4%.
Hope that helps.
OK thanks. But what tipped me off is the label below (I increased the sizes to get a solution). Maybe just these labels are reversed? (I entered 4 in the box entitled "Smallest expected uplift (T – C)") and it appears as "max CG rate".
Fleiss sizes (pop = 25,000 min uplift = 25.00% max CG rate = 4.00%):
Post a Comment
<< Home