3 ms·
I really like the idea, but a "±14.0% significance threshold" is meaningless here. The larger monthly scale should be the default, or you should get more sampl
by goldenarm 9mo ago
I really like the idea, but a "±14.0% significance threshold" is meaningless here.
The larger monthly scale should be the default, or you should get more samples.
- zacmps 9mo agoCould you elaborate what you think the problems are? I guess they should be using some form of multiple comparison correction?
- goldenarm 9mo agoThe daily scale is not statistically significant and is meaningless. You should lower the confidence interval by either increasing the scale or the evaluations.