SIMTOOLS 3.3 suggests using distINV(RAND()) to generate random numbers,
but the inverse distribution functions they provide are overall no
better (and often worse) than the native Excel inverse distribution
functions. A better choice would be to use Ian Smith's VBA code for
distribution functions http://members.aol.com/iandjmsmith/examples.xls
The accuracy of these are not only uniformly better than Excel and
SIMTOOLS 3.3, their accuracy and robustness are better than SAS, S-PLUS,
R, etc.
