Complex software requires rigorous testing to prevent errors in the field.
The study looked at how complex software affects testing and reliability. They tested different metrics to measure software complexity and found that more complex software needs more thorough testing. However, once the bugs are fixed, software complexity doesn't affect reliability in the real world. They suggest measuring software complexity at a low level before putting everything together.