NowtAdminCase Studiesmodule1Module2Module3Module4Module5Module6Glossary

 
nowtCapturing DataVerification and ValidationOrganisation of DataCapabilities of SoftwareProcessing DataDisemination and distributionHardwareSecurity of dataNetwork enviromentsnowt
endErrorsVerificationValidationTransactionsBatch ProcessingInteractive ProcessingChoice of MethodQuestionsend
 

Data Processing Errors

Error can be introduced into a data processing system at any stage. Steps must be taken both to reduce the possibility of incorrect data being collected or processed and to detect incorrect data that has entered the system.

Data Capture / Collection Malfunction of data capture hardware or transmission error. In the case of data collected manually using a form or data recorder there is the possibility of data being entered incorrectly. Use of turnaround documents, automatic data capture and clearly set out and understandable data capture forms can reduce the chances of error at this stage.
Data preparation If data is converted to computer readable format by keyboard then there is the chance that data that is correct on the source document will be copied incorrectly. This is called a transcription error. Possibility of this error is removed if data capture used. Otherwise verification will be used to attempt to spot transcription errors.
Procedural Errors A source document may be lost from a batch of documents or data could be entered twice by mistake. Operations could be performed in the wrong order - e.g. new customer's account debited before the account has been created.
Data processing An error may occur during processing resulting from incorrect data that has entered the system earlier. It is also possible that faulty software could produce errors during processing - such as a final demand for payment of £0.

Note that whatever checks are carried out in an attempt to prevent erroneous data being processed it is impossible to ensure that data is error free.

The syllabus makes an important distinction between data that has been processed and checked either by verification or validation and data that is accurate. To be accurate the data must reflect the real world situation. For example a customer applying for a driving licence fills in a form with a date of birth of 4/6/88 The clerk typing this in miss-types 6/4/88 Two validation checks occur, one says the date is a valid date format (there are 4 months and their are 6 days in that month). The second says that the person is over 17 and thus allowed a licence. The data is thus valid but it is not accurate.

   

© LEV