What's new in IMa2 10/13/2010
Nov 1, 2010
- The problem is that a bias is introduced to the estimated posterior distributions:
- 1) for 3 or more populations
- 2) when the prior on migration is high
- By 'high' I mean that if you knew about what the estimated value is for 4Nu, (call it q) then an upper bound on migration should ideally not be much higher than 2/q
- This is quite similar to the guidlines given on p48 of the IMa2 manual, under 'Getting Started Checklist', but for now those guidelines are more emphatic.
- Of course this is tricky, since you don't know the true values of
- 4Nu.