The transformation of parameter values, which
may range over several orders of magnitude, into
ratings presenting a much narrower range of values,
strongly affects the results produced (Dushoff et al.,
1994). These effects can be illustrated by an example.
In the approach by Kovach et al. (1992), a
pesticide having a toxicity to fish of 0.5 ppm will be
rated 5, whereas a pesticide having a toxicity to fish
of 20 ppm will be rated 1. In this case a 40-fold
difference is reduced to a five-fold difference. On the
other hand, the pesticide dose applied is directly
proportional to the overall index used to compare
pesticide applications. This type of approach inevitably
will favour the selection of relatively toxic
pesticides applied at low doses over less toxic substances
applied at higher doses.