Not OP. This question is being reposted to preserve technical content removed from elsewhere. Feel free to add your own answers/discussion.

Original question:

I'm being provided a dataset with several variables in it, and a success metric (1 or 0) at the end. I'm being asked to analyze the dataset and give insights on how to improve the success metric rate. To do this I intend to do a thorough data analysis to study correlations and relationships. However I'm also intending to run a logistic regression to confirm these correlations with the features coefficients.

My question is, if my sole interest is understanding the most important feature determining a metric, and not building a robust model, should I still split my datasets into 2 ? What benefits do I have splitting it ? Won't my exploratory analysis loose interest if I'm putting away - let's say- 20% ?

Thank you for your help