Select the appropriate statistical test without knowing how the data is distributed, normal distribution or not

#1
Hi,

I have two samples from two independent communities that have mean and standard deviation of time attribute on them, but I do not have access to the amount of time each participant in each sample so I cannot verify the normality of the data distribution. That is, for example, I know the mean and standard deviation of the time attribute for 20 participants are 10 and 1.5 respectively, but I do not know what the time attribute value was for each of the 10 participants, so I cannot understand whether the data distribution is normal or No to use the appropriate test. Is there a test to check the assumption of equality of an attribute in two samples from two independent populations without knowing the normal distribution of the attribute values in the two samples? Thank you so much.
It would be great if I can get little help on that.
 

staassis

Active Member
#2
You need to know at least something about the two distributions. If you can generate a probability of the distributions being this and a probability of the distributions being that, you can run Bayesian analysis to see if the characteristic of interest (mean, median, variance, etc) is the same.