PARAMETRIC BOOTSTRAP TO THE RESCUE(!)

There are many parametric inferential problems where one doesn't have any clear-cut 
fixed sample optimal strategy, and must reply on either asymptotic or some approximate 
techniques. We propose a simple parametric bootstrap method which can come handy in 
such cases where one doesn't need to worry about the complicated sampling distribution(s) 
of the statistic(s) involved. We cite three examples where this parametric bootstrap method 
is found to be quite useful, and as good as the other existing methods (if not better): 
(i) Testing on the shape parameter of a Gamma distribution; (ii) Behrens-Fisher Problem 
(i.e., testing the equality of two normal means when the variances are unknown and possibly unequal); 
and (iii) Testing on the common mean of several normal distributions. The good thing about our 
proposed bootstrap method is that, though computationally intensive, it is easy to implement, 
and the algorithm works through a series of simple steps which can make it appealing to the 
applied researchers.