Reporting results of sensitivity analysis

[cmzyece]

I have just gotten a comment from a reviewer asking that I report quantitative results to support a figure in which I just showed the colour intensity to differentiate sensitive from less sensitive variables. Any suggestions on what would be meaningful values to report in support of the image? Maximum node sensitivity values? Derivatives? or any other?

Thank you in advance!

Chisha

[shooltz[BayesFusion]]

Please note that cooer representing the sensitivity of the node in GeNIe's graphical view is based on the max sensitivity over all node's CPT entries. I'd go for the max as the quantitive value. A node does not have a single derivative, there's one for each probability in node's definition.