i have a rather methodological question. I have a theoretical model and to test it I run some simulations. Because some parametervalues are not entirely clear or are up for debate I want to do a sensitivity analysis.

For this I use a one-at-a-time sensitivity analysis and change only one variable at a time. My question is the following:

When running such a sensitivity analysis is it acceptable to show the "boundaries" of the model parameters, i. e. as the question "For which values are the results of the model in line with empirical research and for which values do I get unrealistic results?"

And my second question (I use an example to make things clear): In my base Model I use a interest elasticity of -0.25 which is based on an empirical study by Chrinko et al. 1999. I know that there are a lot of studies on interest elasticity with results ranging between 0 and -1. Should I use the entire range of this results for my sensitivity analysis and is it bad if I show that my model works for some data (i. e. for -0.25 and -1 but not for 0?)


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.