As per Dr Blaine,

ordinal logistic regression might work.

ordinal logistic regression would work.

Blaine Tompkins is likely right but it might depend on exactly what your research question is. Given what you said in your question I' d say Blaine has only a 5 per cent chance of being wrong. David Booth

Outcome variable decides the test. Here in you case, outcome variable seems ordinal therefore, ordinal/multinomial logistic regression would be appropriate test.

Multinomial logistic regression will work. You can also employ the Kruskal-Wallis One Way ANOVA for a single multilevel variable - https://en.wikipedia.org/wiki/Kruskal%E2%80%93Wallis_one-way_analysis_of_variance. If you have a 2-way ANOVA, the Friedman test might work. https://en.wikipedia.org/wiki/Friedman_test. Other than that, Douglas Montgomery (Design and Analysis of Experiments, 8th ed.) in Chapter 3.11 shows that classical tests on rank responses (which can include Likert scales) are equivalent to the corresponding nonparametric procedures. That opens the classical M-Way ANOVA as an analysis technique for ranked data, for which there is no nonparametric substitute.

As per Dr Blaine,

ordinal logistic regression might work.

Hi Tiolo Dimalibot Alvarez ,

I agree with Imran Anwar . I am not sure if I am correct that your outcome variable is absenteeism? - which I further suppose that if either measured in hours or days or days plus hours, then the measurement level of continuous. Are the factors either categorical (if categorical, how many groups?) or ordinal or a mix of both?

5 or 7 point Likert scale is there, not 6 points. My published article is attached to the data collection plan, data matrix & then data analysis. Have a look & give your feedback. Please explain 6 points scale.