Question about logistic regression.

#1
Hello guys, I have perforemd binary logistic regression with 400 people dataset. I have used as dependent variable default and as predictors gender and salary.
Do you know how would I be able to show probability distributions for the probability of defaulting, separately for actual nondefaulters and actual defaulters?
 

Attachments

hlsmith

Not a robit
#3
not a lot of SPSS users here. Split your dataset and score each subset using the predicted probabilities calculated using model coefficients.
 
#7
Hello my friend. It's my first year at college and I use spss for first time, so I am weak. I do not understand what this equation is. I suppose that it is the equation for my case study, right? What are e and z?
I have this data set (PRE_1, PRE_2, PGR_1, PGR_2 are sth I have not yet understood).

I would appreciate your help a lot.
 

Attachments

hlsmith

Not a robit
#8
I would look into studying logistic regression more before using it, so you dont misuse it or misinterpret the results.

e and z were just the monikers for two covariates in an example. You would replace them with your term and add more to have one for each model term.
 
#9
Thank you my friend. I have studied logistic regression, but theory is different than the case study.
I have a homework and I am really anxious since I do not know what is the equation for my case study.
The deadline for my homework is close and I try to figure the equation.
Do you know the equation for my case study?
 
#10
Hello, kluader,
As I understand, you have the regression equation already. If so, it is arithmetical question. Just make 2 identical sequences of salary, with the appropriate step, one for men and another for women, in excel sheet for example, then input to the next column a formulae of the equation with references to cells of predictors values and you receive the desired distribution.
 
#11
Hello my friend. the problem is that i dont have the regression equation. Im doing the case study for my homework just as the teacher told as in classroom (he did not finish the exercise).
I dont know what my equation is.
hlsmith helped a lot by providing an equation but it is the general equation and I do not know how my variables are represented at his equation.
Thank you.
 
#13
Thank you ask.biostat, I was thinking the same thing but not on a device that makes linking a webpage easy. kluader, just do an internet search for the topic with terms like SPSS, logistic, predicted probabilities.