| Hodges, J. S. and J. A. Dewar, Is It You or Your Model Talking? A Framework for Model Validation, Santa Monica, CA: RAND, R-4114-A/AF/OSD, 1992. |
....runs. Doing so efficiently benefits from advanced design of experiments. This approach contrasts strongly with the naive use of weakly predictive models to make pseudo predictions that could result in very misleading conclusions. For an expanded discussion of this see Dewar et al. 1996) and Hodges and Dewar (1992). 16 ARROYO CENTER ## : R Linking Decisions to Experiments Decision to be made Argument Hypotheses Experiments Experiments designed to affect decisions Decisions based on experimental outcomes The question is, How can we construct experiments with nonpredictive tools The outline of ....
Hodges, J. S. and J. A. Dewar, Is It You or Your Model Talking? A Framework for Model Validation, Santa Monica, CA: RAND, R-4114-A/AF/OSD, 1992.
....the behavior of these systems. A prediction is a statement about a (potentially) observable quantity or event. The usefulness of such a statement is questionable in the absence of information about accuracy. This is reflected by the following, more restrictive definition of the term prediction [Hodges and Dewar, 1992]: a statement about a potentially observable quantity or event, supplemented with a claim about the accuracy of the statement as well as with an argument that backs up the claim. This definition explicitly connects prediction with validation. DEE 3 Generic models should be distinguished from ....
....requires not a binary criterion (true or false, valid or invalid) but rather a continuous one [Hassanizadeh and Carrera, 1992] After all, there is a difference between a model which has passed extensive testing and a second, untested model. One might say that the first has accrued validity [Hodges and Dewar, 1992] whereas the second has not. Oreskes et al. 1994] advocate use of the term confirmation , which is a matter of degree, rather than validation . Model validation is a regulatory as well as a scientific issue [Anderson and Woessner, 1992] Those who are ultimately most concerned with the quality ....
[Article contains additional citation context not shown here]
Hodges, J. S., and J. A. Dewar, Is it you or your model talking? A framework for model validation, Tech. rept. R-4114-AF/A/OSD, RAND Corporation, P. O. Box 2138, Santa Monica, CA 90407-2138, United States, 1992.
No context found.
J. S. Hodges and J. A. Dewar, Is It You or Your Model Talking? A Framework for Model Validation, RAND, Santa Monica, CA, 1992.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC