Data quality - more important than ever in the Internet age

Abstract

The CODATA Constitution begins with a list of goals in its mission, and first on the list is "The promotion of the evaluation and, in general, the quality control of data ...." Data quality was foremost in the minds of the early leaders of CODATA. The first President, Frederick Rossini, had developed a program for systematic evaluation of thermodynamic data and had established a continuing data center with a large staff of scientists for this purpose. Lewis Branscomb, who was on the CODATA Bureau in the early 1970s, wrote several articles such as "Support for Reviews and Data Evaluation" (Science 187, 603, 1975) and "The Misinformation Explosion: Is the Literature Worth Reviewing?" (Scientific Research 3, 49, 1968). I was invited by the Editor of Science to contribute an article "Critical Data for Critical Needs," published in 1981, which gave many examples of unreliable data appearing in the primary literature. Other active participants in CODATA, such as Yeram Touloukian, spread the doctrine of critical evaluation with an almost religious fervor. The programs of the early CODATA Conferences were dominated by papers from data evaluation centers that produced reliable, critically evaluated data sets in various scientific fields.