Dear TalkStatsers,

I would be deeply grateful if you could help me with a simple question I have!!!

From what I have read about Mann Whitney U, it is crucial for the reporting of the results to know whether your two distributions (i.e., the distribution of scores for both groups of the independent variable; in my case “males” and “females”) have the same shape. If the distributions have the same shape, in the reporting of the results you mention the medians of the two groups. If they have different shapes you mention mean ranks.
My question is this; I have a 5-scale ordinal variable and after generating a population pyramid in spss I reached the conclusion that the distributions have the similar shape. However, the medians in my case do not show clearly the differences in the scores of males and females (almost every time being 4,00). Is it acceptable to report both medians AND mean ranks so that it is shown clearer the different scores between my groups, even though I know that my distributions have a similar shape?

e.g., “Distributions of the participation scores for males and females were similar, as assessed by visual inspection. Participation score was statistically significantly higher in males (Mdn = 4,00, mean rank = 220,74) than in females (Mdn = 4,00, mean rank = 192,00), U = …., z = …., p = ….

And one more question; In my case “means” show the real picture of the responses even better that medians and mean ranks. Is it by any chance acceptable to report means in a Mann Whitney U test instead of or along with medians and mean ranks?

Thank you so much for your time!
Kind regards,