The Guide Book has been developed by Bob Giles (starting in 2005) to suggest things he might like to say or to which to point for employees, members, and cooperators of Rural System. Comments range from personal philosophy, statements with which he agrees, and things he calls "hints"
They may some day be organized, but for a start, they have just been added when and where he wanted to. Hopefully participants will use them, be taught their implications, and ask questions about them in training sessions and elsewhere.
What if you have complete faith in a relationship of
Z = A1.3
You want to know Z , so you measure very carefully and precisely A. Then you must estimate or guess at the exponent (as you might for the rate of change in growth or interest rates, etc.). Of course you include knowledge of limits and the possible range of values, but is it reasonable to be very precise (usually at very high costs - time, labor, storage, summary and analyses, equipment use, etc.) in one part of an equation when gross estimates are all that are available for the other? Balance is needed. We work with the expected value of Z, realizing that some numbers cannot likely ever be improved significantly and must always be estimated (e.g., the future value of the national average bank interest rate). Thus combining numbers in a model must be done in a balanced way. We need great precision (high probability), but it comes at great cost. It must be sought for reasonable price - dependent on the context, the probability associated with the other numbers in the model (or the other apples in the box).
Perhaps you will share ideas with me about some of the topic(s) above .
Robert H. Giles, Jr.
July 3, 2005