Reputation
3,575
Next privilege 5,000 Rep.
Approve tag wiki edits
Badges
1 17 37
Impact
~300k people reached

Apr
7
awarded  Popular Question
Mar
9
awarded  Good Answer
Jan
28
awarded  Necromancer
Jan
13
comment Is it meaningful to test for normality with a very small sample size (e.g., n = 6)?
@Silverfish He can incorporate that in his answer. The OP clearly stated that he's interested in that normality test in the context of a T-test. In that context, you want to check whether your data doesn't deviate enough from normality to make the T-test lose power. Hence you're interested in the null hypothesis and not the alternative. Hence my focus on the power of these tests. And hence my answer that with N=6, even the most powerful tests can't distinguish between a normal and a heavily skewed distribution. In that context it does not make sense.
Nov
19
comment Binomial confidence interval estimation - why is it not symmetric?
@StephanKolassa You can find the Clopper Pearson formulae here as well : en.wikipedia.org/wiki/…
Nov
4
awarded  Taxonomist
Aug
31
awarded  Yearling
Jul
18
awarded  Popular Question
Jul
14
awarded  Nice Answer
Feb
2
awarded  Pundit
Jan
12
comment Variance on the sum of predicted values from a mixed effect model on a timeseries
@user52220 That's where you're wrong. E(Xi) is the expected value and hence a random variable, whereas mu_i is the mean of the population and hence a fixed number. Var(mu) = 0, but the same is not correct for E(Xi).
Dec
17
awarded  Great Answer
Dec
1
awarded  Enlightened
Dec
1
awarded  Nice Answer
Sep
30
awarded  Explainer
Aug
31
awarded  Yearling
Jul
2
awarded  Curious
May
22
awarded  Popular Question
May
12
awarded  Popular Question
Apr
30
awarded  Notable Question