https://groups.google.com/forum/#!topic/comp.soft-sys.stat.spss/hMtXkcCIDYM
Testing two correlations coefficients within the same sample:
I need to compare two correlation coefficients within the same correlation matrix of
a single group. For example, is the correlation between beliefs and behavior
significantly different than the correlation between attitudes and behavior in
a single group of teenagers? Any help would be greatly appreciated.
In SPSS, standardize all of the variables using DESCRIPTIVES,
so that testing slope equalities will imply testing
correlations, then either use beliefs and attitudes as
dependents in a repeated measures ANCOVA with behavior as
the covariate and look at the interaction term, which will
test the null of equal correlations, or take a difference
variable using the standardized values of those two, and
regress it on the standardized behavior variable, which
will give the same test.
Multiple regression with 2 IVs and a moderator in SPSS
http://stats.stackexchange.com/questions/141325/multiple-regression-with-2-ivs-and-a-moderator-in-spss
I am running a multiple regression with 2 continuous independent variables and one continuous dependent variable and a categorical moderator. I am doing this in SPSS. I am not sure how to proceed. Please help!! My question is how do (1st continuous variable) and (2nd continuous variable) differentially predict (Continuous Dv) for (categorical variable split into it's 2 categories.
Likely what you want to do is to test whether the 'interaction term' is significant. This will test whether the slopes are significantly different from each other in the two groups.
In SPSS go to Analyze-->General Linear Model-->Univariate.
Put your DV in the DV slot. Put your continuous variables in the 'Covariates' box and put your categorical variable in the 'fixed factor' box. Click on the 'Model' tab and choose 'Custom'. Then move all of your IV over from the left to the right box. Then choose the 'interaction' option and move the continuous IV, together with the categorical IV over to the right hand box.
When you run your model the output will include the interaction term.
Here is a nice walkthrough that looks at various ways of checking if the relation between two continuous variables is the same or different for two different groups (it's all brought together on the last page so you might want to skim through that first):
Hiç yorum yok:
Yorum Gönder