Method for Specifying Business-oriented Data Quality Metrics
Type
working paper
Date Issued
2011-03-01
Author(s)
Research Team
CDQ, IWI2
Abstract
The quality of data (in particular master data) is becoming increasingly important in businesses. This is due to the increasing number of statutory and regulatory stipulations as well as the growing importance of information systems used to support decision-making and the data quality (e.g. timeliness, accuracy, completeness) these systems require. Effective data quality management calls for metrics that enable monitoring of such business-oriented data quality requirements and the identification of business-critical data defects at an early stage. The individual design of data quality metrics taking into account company-specific business processes, controlling systems and application landscapes ensures the business relevance of the metrics and enables data quality to be measured so that business-critical data defects can be identified in good time.This paper describes a method for identifying causalities between data defects, business problems and strategic corporate objectives, and for specifying data quality metrics to monitor the identified data defects. The method follows a top-down approach and only specifies metrics for data defects which can lead to critical business problems. Validation rules are proposes as the preferred measuring technique. By implementing these rules, a measuring system (not a focus of the method) can verify data characteristics and thus identify data defects (i.e. characteristics which can lead to known business problems).The method presented is described as the result of design-oriented research and explains various elements of the method using examples of its application in three companies. The examples illustrate the fact that it is not necessary to use all the elements of the method as the method can be configured for a specific application. To support further applications of the method, the appendix to this paper contains various examples of the requirements to be met in terms of data quality metrics, data quality dimensions, causal chains, cost types associated with defective data and validation rules which have been collected from case studies, focus group interviews and literature research.
Language
English
Keywords
Corporate Data Quality
Corporate Data Quality Scorecard
Data Quality Management
Master Data Management
Controlling
HSG Classification
not classified
Refereed
No
Publisher
University of St. Gallen, Institute of Information Management
Publisher place
St. Gallen
Number
BE HSG / CC CDQ / 28
Subject(s)
Division(s)
Eprints ID
84077
File(s)![Thumbnail Image]()
Name
DQmetrics.pdf
Size
2.56 MB
Format
Adobe PDF
Checksum (MD5)
e91aef26b8d47dec0663fc460a2911ef