Hello everyone
I have run into some STATA problems during my exam project, and I would greatly appreciate if anyone could help me.
The problem is that i am struggling with my multivariate analysis. I started out with a bivariate analyse with two variable, lets call them X and Y... X (the independent variable) is an interval variable, but Y (the dependent variable) is a nominal variable, therefore I have been inclined to use lambda as my correlation coefficient.
Now my problem is, that I want to test this correlation between X and Y by controlling a third variable, Z.
In class I have been taught that it is done with:
bysort "Z": tabulate "X" "Y", gamma nofreq
But as it also says in the code... this gives me a gamma correlation coefficient. As far as I believe, I cannot compare a lambda coefficient and a gamma coefficient. What should I do? Is there another code? Is it even possible to make a multivariate analysis when I used lambda beforehand? Please help!!
Best regards
Related Posts with Trouble with multivariate analysis
Warning when Running CSDID and Many Estimates Not Computing and Hard to Understand WhyHi all, I am working with some classmates to try to re-estimate some first-stage results from a rec…
Interaction terms and testing for significance in StataHi all! I have a mix of a Stata and statistics question that I've been struggling with. Basically, I…
Way to standardize table size using esttab?Hello, I am having a problem with two regression output tables I've created not being the same size.…
xtsur StataHi, I have a question about command xtsur command in Stata. I saw that the instruction as below: x…
multiple line refcat with labelsHi, I am running two regressions and report coefficients by using esttab. I am trying to add two ro…
Subscribe to:
Post Comments (Atom)
0 Response to Trouble with multivariate analysis
Post a Comment