Premium
Quantitative Evaluation of Data Quality in Regional Material Flow Analysis
Author(s) -
Schwab Oliver,
Laner David,
Rechberger Helmut
Publication year - 2017
Publication title -
journal of industrial ecology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.377
H-Index - 102
eISSN - 1530-9290
pISSN - 1088-1980
DOI - 10.1111/jiec.12490
Subject(s) - data quality , computer science , quality (philosophy) , a priori and a posteriori , reliability (semiconductor) , data mining , information quality , dimension (graph theory) , information flow , scale (ratio) , transparency (behavior) , information system , mathematics , engineering , operations management , metric (unit) , philosophy , power (physics) , physics , electrical engineering , linguistics , epistemology , computer security , quantum mechanics , pure mathematics
Summary A method for quantitative evaluation of data quality in regional material flow analysis (MFA) is presented. The principal idea is that data quality is a multidimensional problem that cannot be judged by individual characteristics such as the data source, given that data from official statistics may not be per se of good quality and expert estimations may not be per se of bad quality, respectively. It appears that MFA data are never totally accurate and may have certain defects that impair the quality of the data in more than one dimension. The concept of MFA information defects is introduced, and these information defects are mathematically formalized as functions of data characteristics. They are quantified on a scale from 0 (no information defect) to 1 (maximum information defect). The proposed method is illustrated in a case study on palladium flows in Austria. A quantitative evaluation of data quality provides opportunities for understanding and assessing MFA results, their a priori information basis, their reliability in decision making, and data uncertainties. It is a formal step toward better reproducibility and more transparency in MFA.