implications of using nonparametric test when instead parametric test should be done

[byuu]

hi everyone, our teacher in our statistics class left us a question what are the possible implications of using the inappropriate test on a certain problem. i want to ask whether anyone has encountered a question such as this. i tried solving some problems given in class like, using a nonparametric test to a problem when supposedly a parametric test should be done and vice versa. and i always come up with the same result for each test (Reject the null hypothesis, fail to reject the null hypothesis, etc.) specifically, our lesson is all about one population tests (the parametric t-test and it's nonparametric counterpart, the sign test).

any help is appreciated. :D

[omega]

It's ok to use non-parametric test when a parametric test should be used. In non-parametric tests, you do not assume a certain distribution, so the test is more robust, and can be applied to most datasets.