Salvatore S. Mangiafico isn’t the MovieType variable within participants? as each participant will take part in each level of the IV (movietype). Which statistical test would you recommend?

I would suggest you do a repeated independent t-test or Mann-Whitney test for both dependent variables differently, if both DVs are continuously measured and if sex has two levels. If sex has more than two levels, a repeated measure of One-way ANOVA would go as well as a one-way MANOVA, depending on your objectives. MANOVA will be more robust in estimating gender differences in both DVs jointly and separately. Either way, you are fine, but your objective should guide you better.

Valentine Joseph Owan, Salvatore S. Mangiafico thank you for your suggestions. I performed the Shapiro-Wilk test for normality and it turned out my data is not normally distributed. I will stick to the Mann-Whitney u test. However as far as I know it won’t be any helpful in establishing the direction, it will tell me only if the difference is significant. How can I check the direction then?

Magda Mazur There is direction because Mann-Whitney is not omnibus. The mean ranks of the two groups can tell which group had a higher extent of the attribute or variable measured.

Valentine Joseph Owan, attitude is measured with 7-point likert scale, therefore is it possible to treat the data as continuous and compare means instead of mean ranks? Some articles indicate that likert scales that have >5 categories can be treated as continuous without any harm to the analysis