| Index | Last Week | Contact DQ News | Latest Issue |
| Featured Articles | Data & Information | Science & Technology | Education |
| Economics | Business | Law | Medicine |
Train Location System: Data Quality Essential
A satellite/computer system for locating trains has been tested for three years by the Burlington Northern Santa Fe and Union Pacific railroads, according to a front page report in the June 29th issue of The Wall Street Journal. The tests were prompted by concerns about railroad safety after several recent acidents, and in an effort to increase track utilization, improve on-time performance, and lower fuel bills and track costs. Tests by other railroads are also underway.
The systems being tested are complex and costly. Trains need to be located precisely - within a few feet - due to parallel tracks and sidings. But the data from U.S. global positioning satellites is deliberately degraded to prevent its use by unfriendly nations for military targeting. Trains also lose satellite signals when they cross steel bridges and enter tunnels. Therefore, BNSF and UP are using LORAN C, inertial navigation platforms, and precision odometers to improve positioning data quality.. So far, the tests are proving successful, but the railroads are having problems integrating the data sources.
If accurate and reliable train location systems can be developed, the railroads stand to gain significant safety and economic benefits. Human error causes about half the accidents involving trains. The economic benefits of improved train location data would be substantial. Because of bottlenecks in rail corridors and waits for trains in yards, locomotives sit idling as much as 40% of the time. The article was written by Journal staff reporter William Carley.
Publishing Pressures Rise. So Do Errors.
According to an article in the June 29th issue of The New York Times, the quality of editing fiction and non-fiction has significantly deteriorated during the past decade. This, in turn, has caused more data quality problems in works of fiction and non-fiction.
Today, an editor's sharp pencil can be a luxury in the consolidating industry of large trade publishing houses, which have recently undergone a significant contraction in their editing staffs. These are the editors who shape, tighten, and polish American literature and the copy editors who detect grammatical errors and rescue writers from creative spelling. According to data from the U.S. EEOC, the number of publishing professionals in New York who are largely editors has declined by 16 percent since 1990. But the number of books published in the United States has surged by 42 percent during the same period according to the publishing firm R.R. Bowker.
According to the Times, editing has gradually deteriorated over the past decade. Many books have one or two percent errors - whether grammatical, typographical, or factual. (The article doesn't mention scientific journals, textbooks, or newspapers.) Editors are expected to do more than edit. They are increasingly taking on other corporate responsibilities, such as acquiring new titles. Manuscripts that were formerly published nine months after submission are now published in as little as four months. Consequently, writers and publishers are hiring freelance editors to edit books both before and after submission. Several people, of varying capabilities, may edit a book before it is published. The book's editor may not have the time to effectively coordinate the editing. The article was written by Doreen Carjaval and appears on page B1.
WSJ Survey of 'Hot' Stock Analysts: Was IQ Ignored?
A special section of The Wall Street Journal, published on June 30th, chose America's "All-Star Analysts" for the sixth consecutive year. The Journal's survey is, purportedly, a comprehensive look at corporate securities analysts. According to the Journal, the information used in the "All-Star Analysts Survey" is self-reported by securities firms and their analysts. Securities firms were contacted by the Journal and by Zachs Investment Research Inc., of Chicago. Securities firms were asked for data or corrections to data they supplied about firm and analyst performance.
Analysts were ranked according to how an investor would have fared buying and selling stocks on the basis of their recommendations and according to how accurately they predicted the earnings of the companies they follow. But ranking the analysts is complicated, relies on subjective judgements, and involves several inspect/repair cycles for data and information.
According to the Journal, to be eligible, an analyst had to have followed at least five stocks in an industry during 1997. Coverage had to include at least two stocks in an industry with market capitalizations of more than $1 billion as of January 1, 1997. This requirement was relaxed this year to two stocks with market capitalizations of $500 million in which there weren't 10 stocks with market capitalizations of over $1 billion.
Analysts (with a few exceptions) were required to have covered their industries for the full calendar year and to be senior analysts. Analysts who actively changed employers during 1997 weren't eligible. Analysts had to be senior analysts to be eligible - they couldn't be subordinate to another analyst. Eligible stocks were selected from about 9,500 in the Zachs database. IPOs were included in the survey for the first time. The eligible stocks were those issued by companies in 50 industries that had market capitalizations or initial market capitalizations exceeding $50 million.
The calculation for the All Star Analysts Survey were performed by Zacks Investment Research, which specializes in disseminating and analyzing brokerage house research. A spokesperson for Zachs told Data Quality News that the survey methodology used by Zachs and the Journal is proprietary and couldn't be divulged without the permission of Dow Jones & Company.