0
$\begingroup$

I have a random sample of the following data on JMP. How do I find if there is an association between gender and favorite movie genre? I've been looking through different formulas and can't find anything that will work. Any help will be appreciated.

enter image description here

$\endgroup$
1
  • $\begingroup$ are these all different subjects? i.e. did you ask different genre questions to the same person in this study, and if you did is there any way to link the subjects? basically, imagine that in an extreme case there were only two people: one man and one woman. it's an important detail $\endgroup$ – Aksakal Apr 1 '14 at 18:16
1
$\begingroup$

To get a better overview over your data, you should first build a two-way contingency table. I believe this is called "Contingency Analysis" in JMP. You will get a $2 \times k$ frequency table with two rows (say) for men and women and a column for each genre. You can then look at the row frequencies to compare the distributions by gender.

To visualize the differences, one sometimes uses a mosaic plot, or you could just use (stacked) bar plots (stacking relative frequencies for the categories separately for the two genders).

To test whether the frequencies differ significantly among men and women, you could use a number of statistical tests, the most famous one being Pearson's Chi squared test (implemented in JMP as well as far as I see). Its validity depends on the distributions being "not too extreme", see http://en.wikipedia.org/wiki/Pearson%27s_chi-squared_test#Assumptions for example. If these conditions should not be fulfilled, there are alternatives such as Fisher's exact test.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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