-
One lesson I think is the hardest for students to learn. There usually isn't a "correct" answer in an analysis, only more defensible ones. And you need to be flexible, and not dogmatic, in how you do your analyses.Permalink On twitter.com
♻️ 52 Retweets ❤️ 300 Favorites Mood 0
-
Ultimately an analysis is a culmination of many many decisions: how the data was collected, how the data was explored and cleaned, not just which model you pick.Permalink On twitter.com
-
We talk a little about this in terms of assumptions for a statistical modeling technique, but the discussion is totally disjoint from the other portions of the process.Permalink On twitter.com
-
You're not getting paid as an analyst/data scientist for running a model on data. You're getting paid as because you are taking responsibility for the outputs of that model, that you believe the results are real.Permalink On twitter.com
-
You are absorbing some of the uncertainty to help them make a costly decision.Permalink On twitter.com
-
So you better check, and double check that everything is correct, because your currency as an analyst/data scientist in any organization is your reputation.On twitter.com