Data Pollution
        I've got another great article to share on the TQM of
databases.
   The article comes from the 28 SEPT 92 edition of
'ComputerWorld'(CW).

   The main points of the article are:
        * Almost two-thirds of the 501 companies polled report    
          problems with poor-quality data.
        * Most companies remain closed-mouthed.
        * Fixes include more training, replacing problem equipment,
          using specialized equipment and forming task forces.

   CW states the problem:
        * Poor-quality data residing in corporate databases costs
          U.S. businesses and the government billions each year.

   Causes cited in the article are:
        * Data entry errors.
        * Data entered incorrectly because complete and accurate  
          information was not available.
        * Faulty data purchased from outside vendors.
        * Data 'mismatched' during merging process.
        * Poorly synchronized transaction processing.

   CW gives the following results:
        * Banks and brokerage houses violate SEC reporting.
        * Utilities, telephone and transportation companies bill
          incorrectly.
        * Hospitals cannot collect receivables because of invalid
          patient information.
        * Aerospace firms and large government contractors have   
          cost overruns and late deliveries because of unreliable 
          project management.
        * Manufacturers face product line re-work and recalls     
          caused by unreliable design and production management   
          data

   CW reports the following resources for information:
        * 'Cleanup efforts target dirty data' (CW, 28 OCT '91)
        * 'How clean is your data?' Data Base Advisor, FEB '92.
        * 'Database function: Today's DBMSs provide integrity
           checks, triggers and stored procedures,' DBMS, NOV '91.
        * 'The big picture: Proper data object identification can 
           help separate important information from the junk,'    
           Database Programming and Design, July 1992.

   CW also reports an arsenal of products to attack your data woes.
It also lists some other product and service providers.

   A sidebar on this piece is entitled 'Clean up your data act.' In 
   it Mark Hansen, President of QDB Solutions, Inc., suggests a   
   broad, five-step, Zero Defect Data approach for improving data 
   quality.
   Interesting to note, that after you complete all 5 steps of the
   Zero Defect Data approach, Mr. Hansen says to start again from
   Step 1.  Data quality requires continuous quality improvement  
   (CQI).

   If you asked for 'Devil in Your Data', I will send you a copy of
this article.  If you haven't asked of 'Devil in Your Data' and
would like it and this article, just drop a note here. I'll mail it
out to you.


Lori Barnhill