ASSESSING THE ADEQUACY OF A SINGLE CONFIDENCE FACTOR IN ACCOUNTING FOR EPISTEMIC UNCERTAINTY