Different Results based on Number of Variables in Regression Syntax

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

drc821
Posts: 1
Joined: Sat Aug 26, 2017 3:15 pm

Different Results based on Number of Variables in Regression Syntax

Postby drc821 » Sat Aug 26, 2017 3:26 pm

I am running SPSS version 22. While double-checking results for my thesis, I noticed that I get different results for the same univariate regression based on whether the dependent variable is entered alone or with a series of other variables. For example, the following syntax lines give me markedly different results, even though they involve the same dependent variable ("Ztotal_highest_pe") and the same predictor ("ZCapEud").

"regression/statistics COEFF CI anova R/dep = ZCESD_TOT ZSTAIT_SUM ZIIP_BPD Zemodiv ZMean_pe Ztotal_highest_pe/method = enter ZCapEud." <-- this gives me a B of .07, p = .46

"regression/statistics COEFF CI anova R/dep = Ztotal_highest_pe/method = enter ZCapEud." <-- this gives a B of .20, p = .02.

This is a little scary, as I thought these commands did the same thing (run a separate univariate regression for each variable in the "dep =" command). I noticed that for some other dependent variables, the results are identical, whereas others are not.

Thank you for any help you can offer!
David

Who is online

Users browsing this forum: No registered users and 1 guest

cron