Hi everyone

Great site just what i have been looking for! :-)

I have a problem!

I am conducting a series of questionnaires each with sample size of 200+. The majority of the questions asks the respondent to score different things (ie what activities they do on holiday) on a scale of 0-10. Each question has roughly 6 categories.

As the data is categorized and is not normally distributed I am using the Krushel-Wallis test to assess if the activities are all similarly used.

First of all is this the most appropriate test to use?

Secondly, if the results are significant what posthoc comparison would be suitable?

However, I am using SPSS and I have heard through the grape vine there are no non parametric posthoc analysis options available on SPSS, is this true?

If so, would it be ok to run individual Mann-Whitney tests for each possible pair of activities?

I would really really appreciate any advice!

Take care

Dan

Hi,

Mann-Whitney U tests can be used for post-hoc comparisons with Bonferroni correction of the alpha in order to maintain the overall probability of a type I error at 0.05

Regards,
Re.Vinaitheerthan

Thanks Vinaitheerthan

Your a star!

Take care and

what a great site!