Bear in mind, this assertion is for software testing.
Validating the Process
If the software is, for example, part of a control process and you’re validating the process, you’ll need some statistically based approaches to ensure the software is properly controlling the process through the duration.
Consider a software application that controls temperature in an oven.
Temperature is not going to be held at an exact constant so any point in time could well be above or below the setpoint.
(So, of course, our PROCESS requirements need to provide a tolerance for the temperature and, hopefully, all points in time are within that tolerance! But that’s a separate training course.)
Our software, though, is going to trigger the heaters when the temperature reading drops below a threshold and turn off the heater when an upper threshold is met.
The software will do this every time (unless it gets into an error condition or something odd).
Our software testing only needs to show that the software performs the heater control at the thresholds.
We don’t need to do this repeatedly.