i need help

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

Taini
Posts: 1
Joined: Mon Mar 17, 2014 5:04 pm

i need help

Postby Taini » Mon Mar 17, 2014 5:25 pm

Hi. I am student from Slovakia. Sorry, my english is not good.

I need help with my research. I am doing research, where i compare symptoms (in scale 1-5) in 3 settings. The problem is, that i dont know what to do, becouse each of my 3 settings has unequal number of items (my first setting have 22 items, second - 16 and third 12). When i used paired sample t-test (which autors of my questionaire used) the results are not right... What can i do? Is there a test that can calculate with these unequal numbers of items? Or i just have to do direct correlation to fill up the number of items by average counts?

And my second question is:
If i have scale with 5 points for symptoms: 1 is much worse, 2 worse, 3 same as usual, 4 better, 5 much better and then i have column: "i dont know". What can i do with this columns in sppss? i cant give it the number, becouse it will be bad (if i will give it the number 3 for example, becouse 3 is same as usual) what can i do with this answer?

Thank u for ur help, and i hope you will understand my english and my problem.
costaSPSS
Posts: 3
Joined: Sat Mar 29, 2014 9:54 pm

Re: i need help

Postby costaSPSS » Sat Mar 29, 2014 10:05 pm

I don't know about your first question.
About the second one: just delete this column/response. It will help you more if you don't have it rather than taking it into account.
statman
Administrator
Posts: 2699
Joined: Tue Jun 12, 2007 12:08 pm
Location: Florida, USA

Re: i need help

Postby statman » Sun Mar 30, 2014 11:56 am

Your English is fine and welcome to the forum

#1: Right, the paired t-test is right if you are testing each pair separately but your case counts are at issue so look into the nonparametric version. Do you have the need to say test by gender. If so look into an independence test (again most likely the nonparametric version).

#2: Would not support deleting a column but code DNK as, e.g., a 9 then set it to Missing

Also, see the Note below
See the note below

NOTE: Please read the Posting Guidelines and always tell us your OS, the SPSS version and information about your study and data!

Statman
Statistical Services

Who is online

Users browsing this forum: No registered users and 5 guests

cron