Is Chi Square the right test to use?

Dear all,

I am having a bit of trouble figuring out which test to perform in the following case:

In my eye-tracking study participants are viewing short video clips of people's faces and I count the number of times each of five face regions are fixated by each participant. I have one between-subject independent variable: group (patients, n=6 and healthy controls, n=6) and one categorical dependent variable (visual attendance/fixation of facial features).

PROBLEM: I would like to test whether the two groups differ with respect to the amount of fixations to any of the five facial regions (have patients looked at the eye region, mouth region etc. as often as controls have?).

Since my data are frequencies of categorical data, I assume I couldn't be perfroming an ANOVA (although this is the principle I am looking to apply), but a non-parametric test. Thus, I was thinking of performing a Chi Square test. An internet search has revealed that Fisher's exact test might be more appropriate for small sample sizes (n=6 in each of the two groups) but on the forum I have read that a Chi square test may still be used if "no more than 20% of the cells in the table should have expected frequencies below 5". However I am not sure how to interpret this, because I don't know how to calculate "expected frequencies".

I would be very grateful for any help or input on how to proceed with my data. Thank you very much in advance!



TS Contributor
I believe that there is no need to calculate the exp frequencies, since every statpack does all the needed steps by itself.
I do not know what program you use (if any). If you want, you could try the Excel template I have built to run chi-square test. Just see the thread in the "Other software" section of this Forum.

Best Regards