correspondence analysis

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

JoXs
Posts: 1
Joined: Fri Jun 27, 2014 10:12 pm

correspondence analysis

Postby JoXs » Fri Jun 27, 2014 11:34 pm

Hey guys,

want to start with a thanks for the board, i am happy to have a place to find help.

For a project I have to compare education with the chance to get a job (for two different countries). Afterwards I check them with a third variable, with the sex.

Because all three of the variables are nominal values, I used CHISQ LAMBDA and PHI.

So that would be my synthax:

Code: Select all

CROSSTABS education BY country1
/CELLS COUNT ROW
/STATISTICS CHISQ LAMBDA PHI.
SORT CASES BY SEX.
SPLIT FILE BY SEX.
CROSSTABS education BY country1
/CELLS NONE
/STATISTICS CHISQ LAMBDA PHI.
SPLIT FILE OFF.

CROSSTABS education BY country2
/CELLS COUNT ROW
/STATISTICS CHISQ LAMBDA PHI.
SORT CASES BY SEX.
SPLIT FILE BY SEX.
CROSSTABS education BY country2
/CELLS NONE
/STATISTICS CHISQ LAMBDA PHI.
SPLIT FILE OFF.
I wanted to ask you, if that is enough to actually do a good interpretation?

Thank you very much!
Cheers!
GerineL
Moderator
Posts: 1477
Joined: Tue Jun 10, 2008 4:50 pm

Re: correspondence analysis

Postby GerineL » Mon Jun 30, 2014 8:45 am

It depends a bit on what you want, what you have now is the analyses separately for each sex.
You may want to consider entering both country and sex simultaneously in your analysis, as for instance, the effect of sex may depend on country.
In that case you could use logistic regression. That may be too complex however, I don't know what it is for...

Who is online

Users browsing this forum: No registered users and 1 guest

cron