2
$\begingroup$

here are distributions of four samples:

and the following is the table of data:

enter image description here

I would like to test: (1) if the difference between means of "Give" and "Work-Give" samples is significant; and (2) whether the difference between means of "Take" and "Work-Take" samples is significant. As you can see, distribution of the data are non-standard.

$\endgroup$
0
$\begingroup$

Note even if the distributions of the variables are non standard you can still use a two sample t test. This test relies on the normal distribution of the arithmetic averages of data points, and this should be approximate if you have large enough sample size by centeral limit theorem. The distribution of data points doesn't matter.

Now if you dont think you have large enough sample size you could use other nonparametirc methods like bootstrapping possibly

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.