Problems: Low-precision measurements (large dispersion error) decreases denominator and drives up β, sometimes ridiculously high
If variance of order of dispersion error, we might assume all the variance in dispersion is caused by dispersion error
(if our scales are really bad, we start subtracting off from the intrinsic variance in weight)
Errors in Mbh not considered!
(uh, where did you say you got that ruler?)
Mbh & sig not symmetric
|
