11. 2 : VERIFICATION AND VALIDATION
SPECIFICATION STATEMENTS (AQA 5521)
Understand the distinction between accuracy of information and validity of data
Explain possible sources and types of error in data capture, transcription, transmission and processing
Describe methods of preventing and reducing such errors
Describe appropriate validation techniques for the various stages within an ICT system, from data capture to report generation
Remember the distinction between information and data...
No ?
Well it's this...
Data has no meaning or context. Information does....
Data entry can be DIRECT or INDIRECT.
Human errors: GIGO (Garbage in Garbage out) can be affected by VERIFICATION and VALIDATION.
VERIFICATION: Used to check that data has been entered correctly.
One method of verification is to enter data twice. Think about whenever you're asked for a PASSWORD for the first time. You usually have to enter the PASSWORD twice. This is called VERIFICATION. It is a check for ACCURACY. Data has to be checked when it is entered. My mum used to do ACCOUNTING and use a double entry system for each transaction, and this checking is part of data verification.
VALIDATION: Used to check that data is sensible. It has to be REASONABLE, and also COMPLETE.
Make sure that you can talk about (and understand) the following ideas:
the various ERRORS that can occur: we looked at various examples. Think about the UCAS form which you will all be filling in within the next year (unless you are heading for the 'real world')
the advantages and disadvantages of BATCH processing, and situations where this method of data processing would be used
various types of VALIDATION CHECKS: presence checks, format checks, check digits, headers etc.
TRANSMISSION errors - data can be corrupted in transit, even though it was entered correctly
ODD and EVEN PARITY
It's fairly common to have a series of CHECKS to suggest for different types of data. Make sure that you use a different check in each case - this is a common mistake.