| Index | Last Week | Contact DQ News | Latest Issue |
| Featured Articles | Data & Information | Science & Technology | Education |
| Economics | Business | Law | Medicine |
Enterprise Resource Planning Software - Unforseen Consequences
According to an article in the December 8th issue of The New York Times, so-called "enterprise resource planning software" is causing unforseen problems - including data and information quality problems - for manufacturers who install it. Enterprise resource planning software (E.R.P.) is spreading fast among manufacturers. The E.R.P. programs are huge, and are designed to coordinate a manufacturer's entire business from the time raw materials come in the door to the time the customer pays the invoice. E.R.P. software is not only big in terms of program length but expensive as well, selling for between $300,000 and hundreds of millions of dollars.
Unfortunately, E.R.P software is generally difficult to install. Departments must be retrained, jobs redefined, and procedures discarded or rebuilt from scratch. The core of most E.R.P. software is prepackaged, so off-the-shelf coding lets a company customize configurations and options. The installation process is usually lengthy. Eastman Kodak expects it will take 4 years to install E.R.P. software written by SAP.
How to deal with data has become a major problem for manufacturing corporations that install E.R.P. Selecting the kinds of data a manufacturing firm really "needs" has proved difficult. So has discerning obsolete data that a manufacturing corporation doesn't need. Data quality is a must, since data drives E.R.P. software. It appears from the Times article that E.R.P. software is vulnerable to many types of data quality and information quality problems, ranging from errors in surveying employees to manufacturing cost and inventory errors. The article was written by Claudia Deutsch and appears on the first page of the Business Section.
Gun Control Issues Involve Data Quality
According to an article in The Wall Street Journal on November 11th, the gun control issue is fraught with myths about gun ownership and crime. A related Journal article on November 12th examined the new system of instant background checks for firearms purchases.
The November 11th WSJ article explores gun control "myths." These include shooting incidents involving friends, relatives, acquaintances, and strangers. The article's author, John R. Lott. Jr., explored such incidents and concluded that the FBI's definition of "acquaintances" used in the FBI's Uniform Crime Reports includes drug buyers killing pushers, gang members killing other gang members, and prostitutes killed by their clients. Another gun control issue involves murder rates among nations with similar rates of gun ownership. According to Mr. Lott's research, citizens of nations like Switzerland, New Zealand, and Finland own firearms about as frequently as Americans, but have significantly lower homicide rates where firearms are involved. Finally, it appears that Americans who are granted permits to carry concealed handguns are almost never forced to used them in self-defense. The article appears on page A22.
A related article, by WSJ staff reporter David Cloud on November 12th, concerns the controversy between the U.S.National Rifle Association and the Clinton Administration about "instant background checks" for all firearms purchases in the United States (pistols, rifles, and shotguns). The instant check system is due to begin on November 30th. The FBI, which willl conduct the instant checks over the telephone, has hired over 500 employees and constructed a database of over 57 million records on criminals nationwide. Of these, 34 million are electronically available, but only 20 million of the criminal histories specify the disposition of cases. Without the dispositions available on-line, the FBI must track down individual names manually by calling local jurisdictions to determine the outcome of an arrest. If no case disposition is found within 3 days, the FBI policy will be to reject the sale. According to the Journal, case dispositions will pose a significant data quality problem for the system.
Obviously, data and information quality are essential for any instant background check system. It remains to be seen what the database error rate will be when the system has been operational for awhile. Other problems may also arise. For example, convicted felons who are ineligible to purchase firearms may resort to "identity theft" - impersonating other people using stolen documents. The article was written by Journal staff reporter David Cloud and appears on page A32.
Global Oil Supply Numbers Don't Add Up
An article in the November 9th issue of The Wall Street Journal examined serious deficiencies in global crude oil data. According to the Journal, oil analysts recently discovered that about 300 million barrels of crude oil is "missing." Either the oil never existed, or spreadsheet errors, secrecy by oil-producing nations, or some other cause or causes account for the discrepancy.
The "missing" crude comes at a time when crude oil supply significantly outstrips demand. The 300 million barrels would fill 135 tanker ships. If the crude oil exists and is being stored it would significantly depress crude oil prices for months, according to the Journal.
So, where is the "missing" crude oil? According to the Journal, the missing crude may be in any number of places. The missing crude oil may be stored aboard tankers held at sea and stored in Third World countries. (Demand for petroleum has decreased significantly in Asia over the past year).The quality of production data may also be flawed because many countries keep crude oil production secret. Most oil companies refuse to divulge how much crude oil they have in storage.
The target for much of the oil analysts' ire is the International Energy Association, the world's leading authority on energy statistics. Energy analysts are angry at the IEA for consistently overstating supply. In response to outcries from small oil producers in New Mexico, U.S. Senator Pete Domenici has asked the U.S. General Accounting Office to examine how the IEA collects and processes its data. The IEA has responded that collecting petroleum data is an inexact science. The article was written by Journal staff reporter Steve Liesman, and appears on page C1.
Nonfiction Books May Not Be Factual
Although it's big news when newspapers print false stories or serious factual errors, it's possible to correct the mistakes and fabrications the next day. That is not the case in nonfiction book publishing, where thousands or hundreds of thousands of copies are already in bookstores. According to a report in the November 12th issue of The New York Times, some publishers have a relaxed attitude when it comes to the truth of some of the nonfiction they publish.
Unlike news-gathering organizations, book publishers don't have much interest in fact-checking, especially if nonfiction events happened decades ago, the participants are dead, and there is little documentation to support or refute someone's version of what happened. If "corners are rounded" to make a better narrrative, the tendency is for editors and publishers to sit out any controversy that may occur. And, in most cases, publishers aren't by contract legally responsible for what an author writes. Both author and publisher are also covered by insurance.
There are two kinds of nonfiction fiction. The first is the hoax manuscript. No matter how marketable, editors and publishers are generally afraid to publish totally fictitious books which claim to be nonfiction. A second type of "nonfiction" involves blurring of facts to make the book's narrative a better read. This doesn't seem to bother editors and publishers much, according to the Times. One notable example of creative fact-gathering was Truman Capote's "In Cold Blood, " the account of the murders of a Kansas City family in 1959. The murders occurred, but the conversations, thoughts, and events leading up to the murders were fictitious. The report was written by Martin Arnold and appears on page B3.
Go to: Data Quality Home Page
Comments: dqemail@aol.com (11/8/1998)