## SPSS aggregate family variable and running a bivariate correlation

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

GuillaumeD
Posts: 1
Joined: Mon Sep 21, 2015 6:30 pm

### SPSS aggregate family variable and running a bivariate correlation

I am using a census for the year 1900 of one City as database and here's what I want to calculate. I will give a reduced sample of what it looks like with only the variables that interested me so my statements would be clear

John........54.......rural....... Worker...................1..................................100
Marie......34.......rural.........no job....................2 ..................................0
Gustav....10.......urban.......shoes.....................3..................................50
James......12.......urban.......shoes ...................3..................................50
Marc........14 .......urban...... shoes ..................3..................................60
Felix........35.......urban....... barber..................1..................................200
Nina........32.......urban .......factory..................2..................................100
Sam.........13.......urban.......no job...................3..................................0
Robert.....40.......urban....... Plumber................1..................................150
Mary.......34.......rural..........no job...................2..................................0
Joe..........11.......urban ........shoes...................3 ..................................60
Martha.....19....... urban....... factory.................3..................................60

The variable "relation to the head" gives me the information of the kin. 1=father, 2=mother, 3=child

So we see we have three family here.

The job variable is also listed according to number even if I named it

I want to see if the social class (the job), the urban or rural origin of the father, the father salary and maybe other variable influence the fact of having a child between 10 and 14 years old working. I want to use a bivariate correlation first and adding other variable in a linear regression.

The thing I did was to create a new variable in which is restricted age (from 10 to 14)&(the job: it can be any job, as long as it is one). So I get a variable for only 10-14 child having a job to which I give the name "working child"

I tried running a bivariate correlation with all of the job variables, my new created variable, child, 10-14 ... and looking at my results.. I don't think I did it well.

So I realized I am missing something.

I think I need to aggregate the 10-14 child to his father. The kin information is found in the census through the variable "relation to the head" as I sketched it above. But, I just don't know how to proceed. I never used SPSS aggregation function here even tough I don't know if it is the right thing to do here

What is the break variable, the summaries variables ? And once I have an aggregate variable, let say according to my need, the father linked to the 10-14 child working of not, how can I include it to a bivariate correlation?

Well you see, I never aggregate anything.

Notice that I know how to create a dummy variable and recode an ordinal and nominal one in a 1 and 0 binary thing.

Well, I just can't make it clear, because I don't know how to proceed...

Also, I do not really use SPSS syntax, so if it is possible to clarify with manipulation.

It would be very appreciate to have some help here.

Thanks !

### Who is online

Users browsing this forum: No registered users and 1 guest