Does-It-Work? Metrology for functionality and efficacy

Victor Tang*, Kevin Otto, Warren Seering

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingChapterScientificpeer-review

Abstract

We will unpack this question very systematically and rigorously. First, we make clear that this question cannot not be addressed as if discussing a light bulb, a used car, or a radio. To answer “does it work” in a meaningful and thoughtful way, we adopt the approach used by pharmaceutical companies. Demonstrating that a drug works is a stringent process, which is also regulated by unforgiving laws. A drug works if its developers can verify that it is functional. The science must be valid. Then its efficacy must be verified with people. The science and the statistics must be valid. A drug works if and only if both functionality and its efficacy are verified. This the standard we seek. Second, verification requires instruments, a measurement system, and processes that specify how the instruments are to be used and how measurement data are to be analyzed and interpreted. Simply stated, a metrology must exist. We need a metrology for our paradigm. Regrettably, in spite of our best efforts, we are unable to find a metrology for prescriptive decision paradigms. As a result, we developed a metrology and measurement instrument. To our knowledge this is a first in this field and very meaningful contribution. We invite scholars to research this subject and add to the body of knowledge of metrology in the praxis of decision theory.

Original languageEnglish
Title of host publicationContributions to Management Science
Pages173-201
Number of pages29
DOIs
Publication statusPublished - 1 Sep 2018
MoE publication typeA3 Part of a book or another research book

Publication series

NameContributions to Management Science
ISSN (Print)1431-1941
ISSN (Electronic)2197-716X

Fingerprint Dive into the research topics of 'Does-It-Work? Metrology for functionality and efficacy'. Together they form a unique fingerprint.

Cite this