I know at a sample size of 31 or larger, and from a random sample selection, you are able to assume normality.

Is this in general always true? If no, how do you discern?

Printable View

- September 21st 2009, 09:39 AMallyourbass2212Always able to assume normality if sample size >= 31
I know at a sample size of 31 or larger, and from a random sample selection, you are able to assume normality.

Is this in general always true? If no, how do you discern? - September 22nd 2009, 06:00 PMDrew
You can't assume normality unless you are told it is a normal distribution. At least that's what I was taught.

- September 24th 2009, 09:02 AMallyourbass2212
Well if both sample size x >= 31 and the raw data distribution mimics the central limit theorem then I would say go ahead and use parametric tests (that is assume normality)