Experimental uncertainty analysis is like doing a math problem that has more than one answer. You do the experiment, or problem, and when you get an answer you can have some uncertainty. This uncertainty means that each time you do the experiment, you could get a different answer, even though you are following the same steps. To measure the uncertainty, scientists look at all the different answers and figure out how much the answers vary. Then they figure out how far off their original answer could be, so they know how accurate their answer really is.