Subscribe to DMU

Search DMU Library

Categories

Information Quality Means More than Accuracy

Accuracy is not the only measurement that can or should determine quality for an organization’s data.  Data can be accurate in one context and inaccurate in another, and both versions can have value.

Many people think accuracy is the sole prerequisite for achieving high information quality.  Several authors have identified what can be done to increase an organization’s data accuracy.  However, typically this is not enough to guarantee highly accurate information within complex and / or geographically dispersed organizations.

Accurate – Not Accurate Debate Example

Take the example of an organization that has multiple divisions, each with its own information processing capabilities with a decentralized management structure.  In this situation, the general manager of each division is responsible for all business activity and for the business results that are communicated to corporate headquarters.  Each division is responsible for gathering and providing the information needed by the business participants to fulfill their designated tasks.

The people in each division are happy because the information they can access is deemed to be accurate and meets their individual needs.  They are satisfied and, if asked, would indicate they have access to  high quality information.  In fact, they would be correct in making that statement.

Now corporate headquarters wants to implement a new benchmark that will be gathered for all divisions.  For an example, use “gross sales” or “gross revenue” per customer/client.  Each division gathers the required data, and each division can provide the information to headquarters.  When the information is loaded into a data warehouse and analyzed at headquarters, the corporate analysts are surprised.  The results show that some of the divisions do not do as well on the new benchmark as expected.  The analysts begin to question the accuracy of the data that has been sent.  As stated before, each division is sure that the data it has provided is accurate.  They will state that any inaccuracies have been the result of actions at corporate headquarters.  The resulting discussions can be lengthy and exasperating.

Accuracy Differences Explained

There are several reasons that could explain the differences in the results, all of them valid based on viewpoint.  If the divisions are located in different countries, there can be different legal definitions (or algorithms) concerning what constitutes “gross sales” or “gross revenue.”  If corporate headquarters has not specified the definitive definition and calculation for the term, it is likely that each division sends the data to prepare the financial statements using the rules and regulations for their country.

The definition and calculation differences can be eliminated if the corporate analysts identify clearly what the term “gross revenue” means (and how it is calculated), along with stating that they understand that the new value may not be consistent with the value reported in the financial statements of the division.  Doing so means that the organization must provide clear communication to both senior corporate executives and division management that the values may be different and that the reports must use the corporate definition and calculation.  The need to have a consistent base for calculating the new benchmark requires the change in the process that results in a change in the value stated on the report.

Do not underestimate the amount of time and effort needed to explain the differences and to adopt the new process.  Be prepared to receive questions about the accuracy of the new values from senior executives and division managers.  In addition, make sure that any changes to the gross sales figure are formally accepted by division management.  If corporate analysts change the numbers unilaterally, then the resulting values will not match the values contained in the financial statements.

Senior corporate executives and division managers may hold the opinion that the new values are not accurate.  In this case, senior executives could favor the numbers in the financial statements since these figures are familiar to the executives.  Data governance professionals recommend that the value on the corporate statements receive a new name, so it is not confused with the existing term since that one is used in the divisional and other financial statements.  Although some effort is required to familiarize people with the new term, it is not as much as would be needed to explain the differences in the multiple uses of the same term.

Another common example of differences where straight accuracy is not enough can be found in the customer / client issue.  How divisions identify who is a customer or client may result in inconsistencies when the corporate entity compiles a common metric but the divisions continue to use different definitions and measurements.  There is a variety of actions that could be taken to eliminate these differences.

Conclusion

Information quality depends on more than the accuracy of the data; the context in which the data is captured and used is very important. Data that is accurate in one context may not be accurate in another.

Share on linkedin
LinkedIn
Share on facebook
Facebook
Share on twitter
Twitter

Richard Y. Wang, Ph.D.

Richard Y. Wang is Director of the MIT Chief Data Officer and Information Quality (CDOIQ) Program. He is a pioneer and leader in the research and practice into the role of Chief Data Officer (CDO). Dr. Wang has significant credentials across government, industry, and academia, focused on data and information quality. Dr. Wang was a professor at the MIT Sloan School of Management; he was appointed as a Visiting University Professor of Information Quality for the University of Arkansas at Little Rock. He is an Honorary Professor at Xi’An Jiao Tong University, China.
Wang has written several books on data / information quality and has published numerous professional and research articles on this topic. He received a Ph.D. in Information Technology from the MIT Sloan School of Management and is the recipient of the 2005 DAMA International Achievement Award.

© Since 1997 to the present – Enterprise Warehousing Solutions, Inc. (EWSolutions). All Rights Reserved

Subscribe To DMU

Be the first to hear about articles, tips, and opportunities for improving your data management career.