Frequently Asked Questions

Your group needs data with at least 3,000 rows you actually analyze and enough columns for 10+ predictors. One dataset is enough if it clears both bars. Join more only when they add information you need.

Your group must analyze the data using 3 models, each a specific instance of an analysis, with 10+ predictors. The point is to see meaningful variable selection applied: start broad, then let the data guide you toward a parsimonious final model.

A "model" is a specific instance of an analysis — not just a model family. Your 3 models could be MLR + Poisson + Logistic, or three flavors of MLR with different predictors and/or outcomes. Either is fine.

You may change the outcome variable between models. Picking a different response is often the cleanest route to a different model family — a count outcome for Poisson, a binary one for Logistic — even when your first model was MLR on something else entirely. The models should still speak to the same research question.

Use Python, R, or both — whatever fits your analysis. Your code must be well-commented and runnable.

Not on its own — there is no separate code grade. But it is checked against the deck you submitted, and gaps cost points.

No code in the repository at the deadline costs 25% of that deliverable. Code that is there but does not cover the analysis you presented costs 10%. This applies to Deliverable 1 and Deliverable 3.