Trust and Recognition Coming to Terms with Models

pecial issues serve many purposes. Commonlythey are used to bring attention to and focus ontopics whose importance is increasing or whoseimportance is unrecognized or insufficiently recog-nized, to reprise the state of the art, or provide aneducational opportunity for an interested commu-nity. In this special issue, all of these goals are ad-dressed. The articles by Enns and others, Jacksonand others, and Kong explore developing methodsused to calibrate models. O’Mahoney and othersand Goldhaber-Fiebert and Brandeau explore andelucidate time-related calibration issues. Alagozand others, Eismann and others, and Joranger andothers use calibration methods to establish andreport on foundational work for their modeling pro-grams. An article by Kimmel and another by van Cal-ster demonstrate the very different outcomes thatmight result when models are not calibrated appro-priately and the value of using local data to calibrate,validate, and refine models.The word calibrate appears to have originatedaround the end of the early industrial age and thebeginning of mass-produced artillery. It was impor-tant that the caliber of each cannon be uniform andconsistent with the standard. This idea was thenextended to other mechanical instruments. As onemight imagine, in the age of steam, calibration wascritical. Poorly calibrated instruments could literallyexplode in your face. Over time, calibration hasbecome understood as comparing the output of anyparticular instrument to a known standard andadjustingthisinstrumentsothatitsbehavior,usuallycharacterized by some output, is as close as possibleto the standard.As described in the recent SMDM-ISPOR consen-sus articles,