It is necessary to choose a specified number of maintenance tasks for the demonstration test. In general, there are two basic approaches for sample selection:
- Observe maintenance tasks as they occur naturally in an operational or simulated operational situation.
- Induce faults in the system and observe the maintenance actions to correct these faults.
For the fault-inducement approach, a decision must be made on the type of sampling procedure to be used. This is generally between stratified sampling and simple random sampling. Guidelines are presented here for evaluation of the applicability of the two basic approaches, obtaining maintenance-task samples, and choosing the appropriate sampling design and procedure.