Mann-Whitney u test

Moderators: statman, Analyst Techy, andris, Fierce, GerineL, Smash

knight123
Posts: 2
Joined: Tue Jun 16, 2015 2:43 pm

Mann-Whitney u test

Postby knight123 » Tue Jun 16, 2015 2:48 pm

Help! I'm doing a project at university and statistics certainly isn't my strong point. This is probably basic but was hoping someone could tell me if this is correct.

I've been recording falls amongst patients in two different hospitals. For certain parts its just a yes or no answer. Am I correct that I need to label 'yes' and 'no' as numbers in SPSS, say 1 and 2, in order to do this.

I've attached the results from me attempting to calculate this for whether patients had a history of falls before they came into hospital. At one hospital 35/50 had a history of falls and at the other 37/50 did.

Does this make sense?! Thanks!
knight123
Posts: 2
Joined: Tue Jun 16, 2015 2:43 pm

Re: Mann-Whitney u test

Postby knight123 » Tue Jun 16, 2015 2:49 pm

You do not have the required permissions to view the files attached to this post.
GerineL
Moderator
Posts: 1477
Joined: Tue Jun 10, 2008 4:50 pm

Re: Mann-Whitney u test

Postby GerineL » Tue Jun 16, 2015 3:39 pm

If I understand correctly, subjects are patients (i.e., every patient has a line in your datafile), and variables are "history" (1/0) and hospital (1/0). correct?
If this is the case, go for chi square in corsstabs instead of mann whitney u test. You use mann whitney u when you have a dependent variable at ordinal or interval level, in your case it is binary (nominal).

Who is online

Users browsing this forum: No registered users and 4 guests

cron