Cramer's V & Phi or Fisher's exact test?

[micdhack]

Hello everybody,
I am trying to process data and make correlation between gender and 4 likert scale variable(extremely polite, polite, rude, extremely rude).

I reduced the likert scale into 2 values (polite, rude) and since i have both independent and dependent variables as nominal i followed a Cramer's V & Phi test. The results that i got were interesting BUT i wanted to try also to compare the ordinal original likert scale. (I have no idea if i am allowed to use this test for ordinal data but i have no alternative since gender is nominal)
I got the results but the chi-square had expected cells above 25% from a value of 5. I have no idea if chi-square has any relevance with the Cramer's V & Phi result. If that is true then i have to follow the fisher's exact test. Which brings me to my question and the reason that i posted this under the spss subforum.

Where exactly is the Fisher's exact tests in SPSS? I cant find it in the menus

Also, if you have any comments about if i used the right tests etc please do tell :-)

[micdhack]

Just so that others that have similar problems can find it.
The fisher's exact test appears automatically when you do a chi square and the chi square fails because of the cell count but also you need to have a 2x2 situation on the variables that you are comparing.