Matching subgroups by two variables

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

Marcelo_PB
Posts: 3
Joined: Tue Jan 17, 2017 8:54 pm

Matching subgroups by two variables

Could anyone help me with an analysis using SPSS v22?
My problem is: I have a dataset of approximately 12000 people. In my analysis, I divided the entire sample into three different groups: Group A (n=6000), Group B (n=3000), and Group 3 (n=3000). I performed a multivariate analysis and we found 2 variables mediating the association. However, the reviewer have asked me to perform a confirmatory analysis creating three numerically identical groups of participants of different Groups but matched for major confounders (those 2 variables we found to mediate the association) and then test for differences in the dependent variable.
Is it possible to do this analysis using SPSS v22?
Thanks.
statman
Posts: 2757
Joined: Tue Jun 12, 2007 12:08 pm
Location: Florida, USA

Re: Matching subgroups by two variables

Not clear here

"three numerically identical groups of participants of different Groups .... "
Marcelo_PB
Posts: 3
Joined: Tue Jan 17, 2017 8:54 pm

Re: Matching subgroups by two variables

statman wrote:
Wed Jan 18, 2017 2:44 pm
Not clear here

"three numerically identical groups of participants of different Groups .... "
The reviewer asked me to make subgroups (one subgroup from Group A, one from Group B, and one from Group C). These three subgroups must have the same sample size, and also, matched (mean and standard deviation) by other two variables.
For exemple:
My three original groups are: Group A (n=6000), Group B (n=3000) and Group C (n=3000). The mean/SD for two variables are: Age (Group A: 50/8; Group B: 48/9; Group C: 51/9) and Body weight (Group A: 75/10; Group B: 73/9; Group C: 77/10). I need to make a subgroup from A, a subgroup from B, and a subgroup from C, but the subgroups must have the same sample size (does not matter the size, it probably will have a huge reduction in each subgroup), and present the same mean/DP for Age and for Body weight.
statman
Posts: 2757
Joined: Tue Jun 12, 2007 12:08 pm
Location: Florida, USA

Re: Matching subgroups by two variables

Might be able to get close by "Visual Binning" so check that
See the note below

Statman
Statistical Services
Marcelo_PB
Posts: 3
Joined: Tue Jan 17, 2017 8:54 pm

Re: Matching subgroups by two variables

statman wrote:
Wed Jan 18, 2017 9:54 pm
Might be able to get close by "Visual Binning" so check that
I tried the visual binning but unfortunately it did not work.
I think the "Matching Case-Control" function could be the better option, but I am still having problems. Even including AGE as the matching variable, the mean for age is different among groups.

Who is online

Users browsing this forum: No registered users and 4 guests