Erratum: Correction: Sensitivity Analysis and the Expected Value of Perfect Information

I t has come to our attention that inadvertent errors in an earlier article of ours have contributed to a small controversy 1 on the proper procedure for estimating the expected value of perfect information using Monte Carlo simulation. The errors occur in our procedure MC1 2(p101) for forming a Monte Carlo simulation estimate of the information value EVPI (ξ I) of the parameter set ξ I. The correct version of our procedure MC1 should be as follows.

[1]  Gordon B. Hazen,et al.  Sensitivity Analysis and the Expected Value of Perfect Information , 1998, Medical decision making : an international journal of the Society for Medical Decision Making.