i*****c 发帖数: 1322 | 1 How to do ttest in R if I only know the mean, sd and n of 2 groups? Thanks a
lot. |
n*****n 发帖数: 3123 | |
i*****c 发帖数: 1322 | 3 I can do it using formula. But how about anova? I think R is much easier.Can
anyone tell me the code? I'm learning R but didn't figure out how to do it.
Many thanks. |
x****n 发帖数: 43 | 4 sample1 <- rnorm(n1, mean=mean1, sd=sd1)
sample2 <- rnorm(n2, mean=mean2, sd=sd2)
t.test(sample1, sample2) |
h******3 发帖数: 387 | 5 I think you just translate your formula and use R as a calculator. This is a
fixed number and not sampling problem. just my 2 cents. |
i*****c 发帖数: 1322 | 6 Thanks a lot.
Another question: can R tell if the sample is normally distributed just by n
,mean and sd? |
b**********i 发帖数: 1059 | 7 R will return a message "Not enough information", I guess?
n
【在 i*****c 的大作中提到】 : Thanks a lot. : Another question: can R tell if the sample is normally distributed just by n : ,mean and sd?
|
f***a 发帖数: 329 | 8 光知道n,mean,sd这3个值,神也判断不了是不是normal吧。。。
有具体sample的话,一般就是qqplot, some normality tests, 或者不走寻常路
calculate
hellinger/kolmogorov distance between normal and empirical distribution of
your sample
by n
【在 i*****c 的大作中提到】 : Thanks a lot. : Another question: can R tell if the sample is normally distributed just by n : ,mean and sd?
|