Type = Logical

Default = F

The Analyst Drive program was originally designed to scale all data during the dynamic estimation process for which it was built. This has the effect of making the optimization algorithm treat all count values as equals regardless of their size, which in turn produces smaller relative errors. It may be the case, however, that one wishes to emphasize the large count values in the estimation, as this strategy should prove to be better at reducing the absolute errors. Use USESCALING=F if you do not want to use scaled data (recommended for most problems), use USESCALING=T if you wish to use scaled data. Note that this option is only available for static estimation, dynamic estimation always uses scaling.