Uses specially prepared (generally by the user) data that reflects each type of transaction that could occur in the application system and the calculated result.

Advantages

• Inexpensive
• Can be used for testing parts of the system
• Easy to use
• Takes little time
• Can verify answers

Disadvantages

• Transaction types forgotten; incomplete test
• Does not show how different transactions interact
• Shows nothing about how system will function in the real world

Historical data

Uses historical data for a full cycle and historical reports for comparison. History is re-run with the new system.

Advantages

• Relatively inexpensive
• Can be used for testing parts of the system
• A year can be re-run within a fairly short period
• Can verify answers
• Usually most inter-actions will show up within a full cycle

Disadvantages

• Shows nothing about how system will function in the real world
• Some rare inter-actions will still be missing
• Uses "clean" data, doesn’t test real data
• Generally not usable with online systems

Pilot test

Entire system is tried out in a real location in real time.

Advantages

• Can verify how system acts in real transactions
• Can see system function in real time
• Tests total system

Disadvantages

• Incomplete, not all transaction types tested
• Failure is very costly
• No answer to verify correctness against

   

© Copyright 2000 Raymond S. Kulzick. All rights reserved. 001108.