Next: Theorem on Asymptotic
Up: 3: Estimation of
Previous: Comparison of Two
- The correct statement is that your results are simply unreliable
for small n, and error is not calculable from likelihood function
alone.
- The error is defined for small n, but must be calculated as in
(a)---this is general as not possible to do with just the single
experiment with only a few
observations.
- Note that when math books discuss maximum likelihood estimate for
small n---it is only the value
they discuss and not the
shape of L and the method (b) of finding error.
Geoffrey Fox, Northeast Parallel Architectures Center at Syracuse University, gcf@npac.syr.edu