0
$\begingroup$

I want to compare the percentage of a species in 3 different types of regions. Which statistical test should I use??

$\endgroup$

marked as duplicate by Nick Stauner, Stephan Kolassa, Peter Flom Mar 26 '14 at 10:22

This question was marked as an exact duplicate of an existing question.

  • $\begingroup$ That depends on what property of the groups you wish to investigate, and you said nothing about that other than that you wanted to "compare" them. Specifically, do you want to test something like $H_0 : p_1 = p_2 = p_3$ versus $H_a : \exists \, i \ne j \; s.t. \; p_i \ne p_j$? Or do you want to do contrasts? It's up to you. Then, depending on the nature of your hypothesis, consider what distributional assumptions you are making. That will influence the type of test to use. $\endgroup$ – heropup Mar 21 '14 at 8:01
  • $\begingroup$ I want to do contrasts. to show that the one region is different from the other. I guess I have to do one-way ANOVA but I don't know much about statistics, so I'm not sure $\endgroup$ – ninon Mar 21 '14 at 8:04
0
$\begingroup$

The Marascuillo test enables us to simultaneously test the differences of all pairs of proportions when there are several populations under investigation.

$\endgroup$

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