When assessing a safety-critical system, which metrics are most appropriate for determining if the test objective has been met?

In assessing a safety-critical system, it's essential to focus on metrics that directly reflect the system's ability to operate safely and effectively under all expected conditions. Regression tests run and passed during systems testing are vital because they ensure that any modifications or updates to the system have not introduced new defects or compromised existing functionalities. Safety-critical systems must be thoroughly verified after any changes, and regression tests provide a level of confidence that the core functionalities are intact and remain safe to operate.

In safety-critical contexts, the reliability and integrity of the system are paramount, and running regression tests helps identify any issues related to recent changes. Hence, tracking the number of regression tests that were executed and successfully passed is a sound metric, as it indicates that critical functions have been validated after changes have been made.

Using metrics like incidents raised and closed or incidents closed in testing can provide insights into problem-solving but may not directly correlate with assessing safety in the same way that validating regression tests would. While planned tests run and passed at all levels can also be beneficial, they do not specifically address the crucial aspect of ensuring that recent changes haven’t introduced new issues, which is central to maintaining safety in critical systems.