# Logistic Regression Interpretation help

#### morepil

##### Member
Hi. I have two variables which I want to find the correlation coefficient of. One of them is ordinal the other interval, so I used logistic regression.

I want to present the strength of this relationship next to another relationship(basic pearson correlation) so I just want the r reported.
Does the Pseudo R-Square do the job? The Model Fitting Information and Goodness-of-Fit results both portray significance properly (the first one <0.05, the second one >0.05).

Thanks in advance.

#### GretaGarbo

##### Human
I have two variables which I want to find the correlation coefficient of. One of them is ordinal the other interval,
What about Spearman's correlation?

#### morepil

##### Member
Can I do that? According to this website, you can use a one-way ANOVA, spearman's rho, or an ordinal logistic regression to test the strength of the relationship between an ordinal variable with more than two levels and an interval/ratio variable. My ordinal variable has 8 or 9 levels.

It says on the link above that I can do a spearman's rho but it would treat the interval/ratio as an ordinal variable.
Whereas my ordinal variable is ranged from 0-9, the interval/ratio variable is ranged from 20-150. Given the huge difference of spread I thought it would be erroneous. I don't know. Would it be?

#### GretaGarbo

##### Human
Whereas my ordinal variable is ranged from 0-9, the interval/ratio variable is ranged from 20-150.
I can't see any problem with Spearman's rho.