In my questionnaire I had the 5 scale likert type question “How whould you evaluate new technologies?” (1: entirely negative, 2: relatively negative, 3: neutral, 4: relatively positive, 5: entirely positive). My sample was 400 people (selected with a non-probability procedure) of which none responded the first option, 4 responded the second, 37 the third, 297 the fourth and 69 the last one. I want to check if the responses depend on gender. I thought I could do the x2 test but when I did, under the “Chi-Square Test” table, had a note that said that more than 20% of my cells had expected count less than 5. If I am right that goes to say that the results from x2 test are not valid, so I have to do another test. I did a little research on the internet and I saw I could do the fisher test, but I also read that you cannot do the Fisher test if your sample is more than N=200.

Please, if anyone knows what I should do, I would be deeply grateful!

Please, if anyone knows what I should do, I would be deeply grateful!

Last edited: