Postby Eva » Sat Aug 23, 2008 11:19 am
Hi,
Thank you for your response but I still seem to have a problem with this process. I managed to create multiple response sets but I can't use them in the regression analysis. Or at least I think that I can't.
I went to Define Multiple Response then I choose the variable that I want to have in the set, code them as a categories and so on but the moment I press "done" nothing is happening. What am I doing wrong?
I was also searching at forum and I found that Worf had the same problem as me, he wrote
"I've made a construction called "music involvement" and it's made out of 10 items (statements), of which the participants had to agree or disagree on a likert-scale of 1 (I totally agree) till 5 (I totally disagree). I get a Cronbachs Alpha of .81, which seems to be ok.
Now I'd like to do a Regression Analysis in order to find out, which factors influence music involvement? So here's my question:
- How do I condense those 10 items into a dependent variable with one value for each participant so that I can perform a regression in SPSS? The independent variables, which I’d like to use, are also made of several items each... "
He was advised as follow: "Factor Analysis is a data reduction model, perhaps it would get you there. With 10 attributes in the construct might let the FA run to see how many are created but do display the Scree as a cross-check. The factor scores could then, perhaps, become a DV indicator. Another option is via Discriminate Analysis's classification ability but you will need some baseline for that to work so you would have to set a scoring algorithm to create that baseline variable"
I am sorry but I am really confused…which test should I conduce to be able to create the one dependent variable and independent variable which I will be able to use in the regression analysis?