Does the model include enough information to directly ask
questions about the consistency, justification, lawfulness, and reason of the
model?
Are the assumptions reasonable?
Is the model based on known physical laws?
Are the derivations used in the model correct?
Are there any impossible statements in the declaration of the
model?
Symptomatic questions regarding verification:
Does the model run on a given operating system or with a specific
version of the modeling package?
Given an example dataset or exercise, does the model accept all
of the inputs and give outputs without errors?
Does the model result appear to be consistent with the documented
equations - does it appear to give a credible answer?
Is there sufficient information in the documentation for you as a
user to understand the basic operations of the model, the required inputs, and
what the outputs represent?
Are there anomalous results - things you would not expect from
this model? For example, the documentation may indicate that you should get
results that simulate a particular pattern over model time increments. Do you
get that pattern or are there significant outliers? Do these appear to be
consistent with the documentation?
Do the results of the model show features that appear obviously
incorrect, such as discontinuous or non-differentiable results where well
behaved results were expected?
If you use a browser to run the model, does it run correctly in
the browser version you are using? Does it require specific browser add-ins or
options to be in place in order to work?