| Index | Last Week | Contact DQ News | Latest Issue |
| Featured Articles | Data & Information | Science & Technology | Education |
| Economics | Business | Law | Medicine |
ACM Communications Focuses on Data Quality
A special section in the February issue of Communications of the ACM focuses on Data Quality. Guest editors Giri Tayi and Donald Ballou wrote the introductory article, "Examining Data Quality." Tayi and Ballou discuss what "data quality" means to the people and organizations that work with data. Their article defines "data quality," explores several dimensions of data quality (accuracy, completness, consistency, and timeliness) and discusses in length the difficulties people and organizations encounter in ensuring data quality. Their article also examines what the approriate level of data quality should be for an organization. They emphasize that an organization's major use of data should determine the overall level of quality.
Dr. Richard Wang's article "A Product Perspective on Total Data Quality Management," emphasizes the fundamental importance of having an overall plan or blueprint for ensuring information quality. Dr. Wang's information processsing (IP) system model (define, measure, analyze, improve) is analogous to Deming's manufacturing model of a decade ago (plan, do, check, act). Like Deming's manufacturing model, Dr. Wang's information processing system model relies upon a complex system of information measurements, metrics, and inspections to ensure product conformity and improvement. But recent state-of-the-art manufacturing quality systems are design-based. Today, manfacturing quality requirements are designed-in while manufacturing variability is designed-out.
Ken Orr's article, "Data Quality and Systems Theory," emphasizes the need for continuous user feedback to ensure data quality is maintained. Orr's data quality paradigm describes a system whereby the data quality of an organization's information systems is accurate enough, timely enough, and consistent enough for the organization to survive and make reasonable decisions. Orr promulgates six general data quality rules that one can deduce from a feedback control system (FCS) view of information systems. Unfortunately, Orr doesn't include error components in the various feedback loops he succinctly describes! Moreover, like Wang, Orr devises a complex system of metrics, measurements, and inspections to "solve" data quality problems.
A third article, "Assessing Data Quality in Accounting Information Systems," written by Kaplan, Krishnan, Padman, and Peters is a hodge-podge of complex techniques the authors have devised to bring rationality to accounting information systems (i.e., systems that organizations use to plan, evaluate, and diagnose the dynamics of operations and financial circumstances). Unfortunately, the authors of the article admit that their data quality model doesn't necessarily provide the "guidance" that accountants and auditors are seeking.
Finally, it seems Tom Redman's article, "The Impact of Poor Data Quality on the Typical Enterprise," is more of the same "ain't it awful?" stuff Redman has been writing about DQ and IQ for almost a decade. Yes, it IS awful. What are DQ/IQ Gurus, Seminar Givers, and Consultants doing to make data and information quality problems a lot less awful? Those who are in the DQ/IQ trenches would like to know.
Net Privacy Precautions vs. Data Quality
According to a February 6th article in The Wall Street Journal, privacy on the Internet is an illusion. The illusion of privacy in cyberspace stems from how personal computers developed. Before the Internet became popular, people typed documents that were stored on their personal computer's hard or floppy disk drives. Electronic bulletin boards were local or topical and seldom shared or divulged personal information.
But in cyberspace any number of people could be watching personal messages posted in a chat room or a topical discussion area. Sellers of products over the Internet, proprieters of X-rated Internet sites, marketers, government and law-enforcement agencies, and Internet service providers themselves, have an organizational or finanacial interest in amassing personal information. (America Online recently divulged a subscriber's identity to U.S. Navy investigators. And last year America Online proposed selling its subscribers' home telephone numbers to telemarketing firms). Various types of criminal activities also take place on the Net, ranging from stock fraud to pedophilia.
According to the Journal article, while more than 80 bills are pending in the U.S. Congress that would put restrictions on the Internet, none of them is expected to come to a vote soon. The Clinton Administration generally opposes restrictions on the Internet, and urges the industry to police itself.
According to the Journal, the most obvious step consumers can take to protect themselves is to be wary of what they put on the Internet. The Journal advises using post office boxes for mailing addresses, nicknames rather than real names, giving out as little personal information as possible, and avoiding e-mail altogether. [Editor's note: If widely adopted, these precautions would obviously affect Internet data quality.] The Journal report was written by staff reporter Rebecca Quick and appears on page B5.