I am trying to analyze a survey that asked students to respond to questions e.g. “In my opinion, the economic system in the US leads to an unfair distribution of income” by indicating their opinion using a Likert scale. The number of respondents are 280. The answers are broken down by gender. Looking at the answers for a question, when using vassarstat online, the mean rank for sample A and B are also returned (130.9 for A; and 156 for B) when the result is significant at p < .05. Can I use this to indicate that in group B higher % of students answered agree and strongly agreed ( agree=4; strongly agree=5) with the statement posed?

Looking at the more detailed statistics calculated by http://www.socscistatistics.com/tests/mannwhitney/Default2.aspx

Result Details

Sample 1

Sum of ranks: 22648.5

Mean of ranks: 130.92

Expected sum of ranks: 24306.5

Expected mean of ranks: 140.5

U-value: 10913.5

Expected U-value: 9255.5

Sample 2

Sum of ranks: 16691.5

Mean of ranks: 156

Expected sum of ranks: 15033.5

Expected mean of ranks: 140.5

U-value: 7597.5

Expected U-value: 9255.5

Do I need to report the U values?

Looking at the more detailed statistics calculated by http://www.socscistatistics.com/tests/mannwhitney/Default2.aspx

Result Details

Sample 1

Sum of ranks: 22648.5

Mean of ranks: 130.92

Expected sum of ranks: 24306.5

Expected mean of ranks: 140.5

U-value: 10913.5

Expected U-value: 9255.5

Sample 2

Sum of ranks: 16691.5

Mean of ranks: 156

Expected sum of ranks: 15033.5

Expected mean of ranks: 140.5

U-value: 7597.5

Expected U-value: 9255.5

Do I need to report the U values?

Last edited: