I am data science beginner, and I have a question about methods that I could use to analyze the following data. It is a simple case, I am trying to check the influence of cohabitation before marriage and being a believer on divorce. So, I am asking what model could I build to test it? The only thing I did for now was checking the conditional probability of those cathegories. I know that the data is very simple, but I read about so many various tehcniques that I feel paralyzed and I don't really know what should be the best option. I am open for classical statistical methods, bayesian methods, and building ML models.

I think that chi square test would be suitable, but I am also interested in more sophisticated methods.

(This matrix is just a visualization, I have these data in df with IDs and stuff. Source of the data is NSFG)(I am using R)

enter image description here

  • $\begingroup$ Welcome to DataScienceSE. The first thing is to decide what kind of answer you're looking for: for example, to know whether there is any influence, a chi-square test makes sense. If you want to quantify the influence, I think that this can be calculated from conditional probs. You could also train a model which predicts the target 'is divorced'. $\endgroup$
    – Erwan
    Jul 2, 2022 at 14:43
  • $\begingroup$ Thanks for the answer! I did use conditional probabilities and it was fruitfull. What kind of model would be a good choice for such simple data? $\endgroup$
    – Karpi
    Jul 4, 2022 at 9:30
  • $\begingroup$ Well if you want to practice you can try to train a model, but it's not very interesting since you only have two boolean features. $\endgroup$
    – Erwan
    Jul 4, 2022 at 15:13


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy