12:29 AM
Yeah, I agree. An asker is often unsure of what they need to ask and doesn't realize that what they haven't said is crucial to dealing with their real problem. An apparently simple regression question changes completely once it's revealed that the response is binary, or cumulated values, or that there's error in the predictors. Or the question will explicitly ask about testing (because that's what they know about) but the actual problem is one of prediction.
7 hours later…
7 hours later…
2:01 PM
Good exchange. It reminds me of a lesson this site continually teaches me: an important component of creativity in statistics is the ability to interpret a question in multiple distinct ways.
« first day (2888 days earlier) ← previous day next day → last day (2119 days later) »