0
$\begingroup$

I have a bank customer loan dataset with columns loan amount, funded amount, interest rate(high, medium, low), annual income of customer, loan status as (default and fully paid).

Could I use two sided P tail test to compare each columns with loan status, whether there is any significance difference between default and fully paid analysis? I don't know what would I infer after the results.

For e.g. tail test on annual income of customer for each loan status(default and fully paid). After analysis the result came out to be rejecting Null hyopthesis(H/u default= fully paid) i.e. There is a significance difference between the annual income mean of defaulters and fully paid.

$\endgroup$
  • $\begingroup$ Hello and welcome on StackOverflow!!! In order to get your a better answer, and your answer don't get deleted later on, you would have to read and follow some basic rules and conventions that will help everyone to communicate and help better. Please go to stackoverflow.com/help and read about the topic "What topics can I ask about here?", and "What types of questions should I avoid asking?". $\endgroup$ – Jonathan Gagne Oct 4 '18 at 19:49
  • $\begingroup$ Yes you could do that, provided all variables are numerical. The test would test whether the means are differente among the columns. For non numerical data you would have to use modifications of the t test. $\endgroup$ – user2974951 Oct 5 '18 at 10:32
0
$\begingroup$

Your purpose of this analysis is not entirely clear, can you try to give more detail/explain better, which question do you want to ask from the data? There is already a lot of information about credit scoring on this site, look through this list.

You would probably be better off with a logistic regression for default status, with your other variables as predictors, at least as a point of departure.

$\endgroup$

Your Answer

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