Agree with Milos. You can also start the process backwards, that is, crosstab each pair, then run the test among the three variables as Milos has suggested by interchanging the layer variable each time. Check your objectives for the research to choose the appropriate scheme to answer the research questions.