Hello,
I have a question about confidence intervals. I have a variable with a sample of 30 numbers. When I do a stem-and-leaf display or a box and whiskers display, it shows that my data is right skewed. Can I still do the confidence interval? I know from the central limit theorem that usually n=30 is sufficient for normal distribution, but this isnt case this time.
I did the confidence interval on the computer, and got numbers, but I am thinking I should make some type of discloser about the right skewed data, or the other alternative is that I simply can not use that information at all.
Thanks,
Hudson
I have a question about confidence intervals. I have a variable with a sample of 30 numbers. When I do a stem-and-leaf display or a box and whiskers display, it shows that my data is right skewed. Can I still do the confidence interval? I know from the central limit theorem that usually n=30 is sufficient for normal distribution, but this isnt case this time.
I did the confidence interval on the computer, and got numbers, but I am thinking I should make some type of discloser about the right skewed data, or the other alternative is that I simply can not use that information at all.
Thanks,
Hudson
