Closed MassD closed 1 year ago
That's a good idea. Age is almost certainly informative here (maybe birth year even more so).
But the smaller the group you select for, the less data you have, making the estimated probabilities less precise. This is what hierarchical modeling is good for!
Thanks for the reply.
If we forget about the real world, assume this is the dataset that contain full data (often assumed in university exams), i presume the age would be used? especially the age of 31 has been given
That's a good idea. Age is almost certainly informative here (maybe birth year even more so).
But the smaller the group you select for, the less data you have, making the estimated probabilities less precise. This is what hierarchical modeling is good for!