That is very common scientific utilization, Most likely simply because you always set out to disprove your speculation (In this instance, the model) rather than to demonstrate it (which You can't do). It is possible to only examination whether or not your design is consistent with the data, so a