Biology Forum › Community › General Discussion › Biostatistics questions
- AuthorPosts
- May 8, 2006 at 2:50 pm #4724
biobio3012ParticipantI am not sure whether violoation of the assumptions of an analysis will or will not have influence on p-value?
thanks - May 8, 2006 at 3:57 pm #48080canalonParticipant
Violations of assumptions means that you are applying a test that you shouldn’t be applying to your data. The p-value may still be computed, but you cannot interpret it.
For example if your population has two maxima, and you are applying a t-test, which assumes a normal distribution, you may still get a p-value, but whatever result it doesn"t mean anything. This mean that you should probably use another test, here a non parametric one.
- May 9, 2006 at 1:37 pm #48130biobio3012Participant
so you are saying that the p-value may be influenced by the violation of assumption? I am not sure what I think is right, but I think that you use P value to reject/accept the assumption(hypothesis), so if the hypothesis is violated, it doesnt change p-value because p-value is set.
sincerely, - May 9, 2006 at 2:59 pm #48136canalonParticipant
No I am saying that even though you may always be able to compute a p-value, if you violated the assumptions of your test in the beginning (if for example it assumes a normal distribution of your results and it is not true for your data) you should not draw any conclusion from your result.
The interpretation of the p-value depends on the set of assumption in the beginning and can only be interpreted if you respect them. Find yourself a good biostatitics book if you want more explantions.
- AuthorPosts
You must be logged in to reply to this topic.
