Comment by ahartmetz

When I did the mandatory lab exercises in physics, there was a more benign variant of that problem: the conventional value had to fall inside the error interval. However, it was allowed to add additional errors with a good explanation (...some creativity). I really didn't like to increase the estimated errors to make the result work, and I think the (unimportant) grades were reduced for doing it.

I remember being really consistent with the stopwatch in one exercise, so sadly the spread of measurements (implying a natural uncertainty) was small. That was bad!