Timeline for If a credible interval has a flat prior, is a 95% confidence interval equal to a 95% credible interval?
Current License: CC BY-SA 4.0
3 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jul 12, 2018 at 12:14 | comment | added | Frank Harrell | See this. Think of forecasting the winner of a soccer match as the game progresses. Your current probability that team x wins the game can completely ignore the past forecasts you made. But if operating in a frequentist mode you'd have to envision all the times your team lost the game and consider extremes of the scores at all the points during the game that you tend to make forecasts. Multiplicities come from the chances you give data to be extreme, and this factors only into frequentist calculations. | |
| Jul 11, 2018 at 21:13 | comment | added | badmax | What does it mean to be in "forward-time predictive mode" and why don't we need to consider selection or multiplicity effects? | |
| Jul 9, 2018 at 11:47 | history | answered | Frank Harrell | CC BY-SA 4.0 |