Binary Logistic Regression Questions

#1
I have a few questions about binary logistic regression.

A-I have 1 categorical DV (2 level).
B-I have 1 categorical covariate (2 level) programmed as a categorical variable in the "Categorical" options in he regression window. That "indicator" and "Reference Category" stuff, which I'm not exactly sure how to work. I left it as "indicator" and "last." Correct?
C- I have 1 continuous fixed factor, which is really what I want to present. I want to present the relationship between this variable and my DV. So my Q is how to report this? Some articles I am following reports Adjusted Odds Ratios only. and probably p-values. Where do I get the p-value for this relationship? From the associated Sig, Exp(B) and 95% CI for Exp(B) Value boxes in the bottommost table, Variables in the Equation table?
D- I have 4 continuous covariates.

To make things more complicated, the 5 continuous factors above in sections C and D, I also have them as categorical. I would actually rather to use them as continuous. But I'm thinking SPSS doesn't take them as continuous because the dF is 1. I had a similar problem before but couldn't get my mind around it so I'm asking again. Does it look like SPSS treats these variables as continuous?

And is the result this sentence: "as the continuous fixed factor goes/up down, the chances the DV will go up/down is 0,862 controlling for the covariates mentioned above?" Since the AOR is actually lower than 1, it is in a negative relationship, right? I got that 0,852 from the fixed factor (IV of interest, the second from the top).

Thanks a lot.
 

Attachments

Last edited:
#3
Please define what you mean by "fixed factor". Do you mean fixed effect, meaning it is not a random effect?
I just realized I made a mistake in writing this out.
What I meant to say was, my IV of interest is the fixed factor, in section C.
I also have 4 other IV's (covariates) (sections B and D) that I am not interested in. I'm merely trying to control their effect on the DV.
Does this answer your Q?
 

hlsmith

Not a robit
#4
No. What do you mean by "fixed factor"?

Also, if the covariates you are controlling for are continuous, don't let SPSS call them categorical. Figure out what your issue is, it might be that they are categorical. Also, provide a snippet of the dataframe, so we can see what you are working with.

Lastly, you don't need to report p-values. Just report adjusted OR's with 95% CI's. If someone wants p-values, refer them to the 95% CI's.
 
#5
No. What do you mean by "fixed factor"?

Also, if the covariates you are controlling for are continuous, don't let SPSS call them categorical. Figure out what your issue is, it might be that they are categorical. Also, provide a snippet of the dataframe, so we can see what you are working with.

Lastly, you don't need to report p-values. Just report adjusted OR's with 95% CI's. If someone wants p-values, refer them to the 95% CI's.
The reason why I differentiate between fixed factor and covariate is just interest. I'm interested in the effect of the fixed factor on the DV whereas I'm not interested on the effect of the covariates on the DV. I thought I could name these as such b/c in ANCOVA SPSS lets you differentiate between fixed factors and covariates. They're treated as "independent variable of interest" and "covariates" to my knowledge and that's what I meant here.

There are two types of covariates in my data: 1 categorical covariate, and 4 continuous covariates. I tried to tell SPSS to treat the categorical covariate as categorical using the "categorical" menu, which I also had questions about (see above). One of my questions is whether or not SPSS treats the categorical and continuous covariates properly. That's why I attached the screenshot.

Does the attached screenshot help with the snippet of dataframe?
The first one is the categorical covariate. The second one is the DV. The rest are continuous.

Oh! I just realized why you wanted me to clarify what I meany by fixed factor.
What I want to have is this:
The effect of 1 continuous variable (fixed factor) on a categorical DV controlling for 1 categorical variable (covariate) and 4 continuous variables (covariate). I can also work with continuous variables transformed into categorical variables if need be.
 

Attachments