Questions To Ask

Shodor > CSERD > Help > Validation > Questions To Ask

There are a number of general questions you should answer about the validity of the model regardless of the subject:

  • Is the application of the model based on current scientific principles?
  • Is the representation proposed here acceptable for the purposes that are described? (e.g. are the method's accuracy, precision, and agreement with current practice sufficient for the model's stated use?)
  • What are the known errors associated with this representation of the system? Are they indicated in the documentation or known based on the literature? Are these error levels acceptable for the purposes to which the model will be applied?
  • Are the model inputs sensible given what you know about the system?
  • What happens in the extreme cases? E.g. data that may cause instability in the model or the results?
  • Are other required inputs for the model readily available or will they be so hard to obtain as make the model impractable to use?
  • Are the results of the model credible? (Are the outputs in the correct range? Do they fall within the bounds of the known limits of the system? Do they make sense?)
  • Does the model produce relevant results that could be compared to measurable quantities?
  • Do the model results agree with (show coherence with) results of experiment or accepted theory?
  • Are there indications of where in the scientific literature one would find related experimental data or accepted theory?

©1994-2014 Shodor