|
My apologies if this has been answered here. It's my first time here. I am a developer by trade. I am not really into this thing, but I was asked to do some data simulation using SPSS. But I am not sure what significance level to use. I was given a small data set of 24 cases. This is medical research so I assume I have to use .01. I just don't know if it is proper to change the default level of .05 in SPSS given I have only a small data set. Somebody please shed some light. |
||||
|
|
|
The significance level is your choice. It doesn't depend on sample size or the type of research you are doing. What is involved is how much risk of a type I error you wan tto take. The type I error is the probablility that you reject the null hypothesis when you shouldn't. The 0.01 and 0.05 levels are used only as a matter of tradition. There is no mandatory significance level. |
|||
|
|
|
What test are you using? A lot of tests have an assumption of normality somewhere that breaks down with small numbers. The significance level depends on how sure you want to be? |
|||
|
|
