|Index||Last Week||Contact DQ News||Latest Issue|
|Featured Articles||Data & Information||Science & Technology||Education|
FAA Data Quality Deficiences Compromise Safety
A front-page investigative report in the July 27th issue of The Wall Street Journal criticizes the Federal Aviation Administration for serious data quality deficiencies. Although the FAA has mountains of data, powerful computers, and hundreds of inspectors, the agency gives the airlines and its own regional offices wide latitude to collect or not collect various types of safety-related data.
The Journal investigation resulted from a passenger complaint about how the FAA handled an in-flight engine failure. The Journal found that the FAA increasingly relies on "incident report" databases furnished by air carriers. But many of the records in the databases have missing or incorrect data. The Journal followed how a number of in-flight "incidents" - explosions or shut-downs of JT8D engines - were tracked and recorded by Continental Airlines and the FAA. Of 27 such incidents during 1997, the FAA's databases showed 19 incident reports.
According to the Journal, problems reported by FAA inspectors varies significantly by region; FAA Inspectors often fail to report problems because of the amount of paperwork involved; airlines want to minimize problems to avoid alarming the public.
The U.S. General Accounting Office recently surveyed FAA inspectors and concluded that problems were being underreported. The Clinton Administration has promised to help cut the rate of fatal accidents by 80% by 2010. The Administration's plans include heavier reliance on data, added safety equipment in cockpits, and more-rigorous engine inspections. The GAO has cautioned that if safety systems data quality remains poor, the safety systems will not be reliable.
Journal staff reporter Scott McCartney wrote the article.
Cartilege Implant Surgery Ahead of FDA Efficacy Data
According to an article in the July 27th issue of The Wall Street Journal, cartilege implant surgery and other cartilege therapies are being being performed without sufficient data from controlled studies that would show if, and how well, the techniques work. Moreover, several cartilege replacement therapies are done concomitant with other therapies like ligament repair - which makes it difficult or impossible to determine cause-and-effect. And no reliable data presently exist that would enable patients to choose the "best" cartilege replacement therapy.
At least one pharmaceutical company marketed a cartilege replacement therapy ahead of data. The U.S. Food and Drug Administration approved clinical use last year on an accelerated basis and is only now demanding clinical trials data. The report was written by Marilyn Chase and appears on page B1.
Better Data for Oceanographers
A report in the July 28th issue of The New York Times, indicates that a global network of deep underwater microphones built by the U.S. Navy during the Cold War will be reconfigured for civilian use. In its reincarnation, the listening system is to be known as the National Oceanographic Environmental Monitoring System.
The Navy has given a consortium of civilan and defense organizations permission to reactivate seven installations where cables from undersea listening networks came ashore. The Navy is also taking steps to degrade the quality of data from the undersea hydrophone so its submarines and ships can't be located. The surveillance microphones can hear natural sounds like whales singing, seals barking, ice cracking, and volcanoes erupting over distances of hundreds of miles. The sensitivity of the arrays is said to be such that a whale could be tracked for weeks as it sang. The article appears on page C4 and was written by Willam Broad.
Computer Resume 'Data' Confound Job Applicants
A front page article in the July 30th issue of The Wall Street Journal, examines the effect that digital resumes are having on job seekers. No longer are resumes elegant typescript and expensive paper. Plain paper resumes with standard fonts are now scanned into computer databases. Then text searchers and artificial intelligence-based extraction engines analyze key words and phrases, and display or print out a list of "best" candidates.
According to the Journal, the process leaves job seekers and career counselors confused about how to make a good impression with an electronic resume. Some feel that putting descriptive "key words" at the top of the resume is the best approach. Others think that "traditional" resumes are better. The resume scanning systems manufactured by the two top corporations in the field produce significantly different results from the same batch of resumes. And scanned resumes that don't contain up-to-date key words and technical jargon may never be matched to job requirements.
On the other hand, resumes scanned into a database offer much higher flexibility than paper resumes. Data about job seekers are readily available throughout the organization and employer searches can be highly refined. Unfortunately, the Journal article doesn't discuss data quality issues like falsification of skills and achievements, misspellings, and errors in scanning. The article was written by Journal staff reporter Ellen Pollock.