Skewed dataset-multiple regression.

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

Newbie2SPSS
Posts: 1
Joined: Thu Mar 03, 2016 6:11 pm

Skewed dataset-multiple regression.

Postby Newbie2SPSS » Thu Mar 03, 2016 6:19 pm

Hi.

As you can tell, I'm pretty new at this. I've received from class a data set with multiple IV's. I'm supposed to examine the predictability of each variable when it comes to the OV. When I create a histogram the dataset for the OV is slightly skewed. Should I do a bootstrap on it because of the uneven distribution?

When I did a stepwise regression on it, SPSS excluded mots of the IV's expept from one, and the pp plots give an idea of very unevenly Distributed datas, and no linear relationship between IV and OV. Exept for the one that was found significant in the stepwise regression.

My point is; can someone help me with a checklist of what do do when receiving a bunch of data? In the correct order. I'm so insecure, I've been doing this paper all day, but I feel like I'm just rambling around. And the correlations between the numbers are negative, which makes me completely confused, beacause that would be like; the more confidence someone has, the less self-efficacy they have..? Right? That doesn't make sense. So maybe I need to use a non-parametric procedure? Because regressions are Parametric, right? *phew*

Thanks for all help!
GerineL
Moderator
Posts: 1477
Joined: Tue Jun 10, 2008 4:50 pm

Re: Skewed dataset-multiple regression.

Postby GerineL » Fri Mar 04, 2016 8:32 am

I would advice you to look into the andy field book for spss, he has exactly what you describe, and all checks (and what to do if you don't pass the check) .

Who is online

Users browsing this forum: No registered users and 2 guests

cron