Transcript: Another reason that science cannot make statements with absolute certainty is to do with sampling or the limitations of data. Induction as a tool of the scientific method is based on generalizing from a finite set of observations or situations to a broader conclusion. If I had ten people in front of me and I asked which hand each of them wrote with and all ten said that their right hand, would I be justified in concluding that all people are right-handed? You know that the answer is no. The incidence of left-handedness in the population is about 10%, so with ten people questioned I might just be unlucky and find none that are left-handed. However with a hundred people, if I question them, I would expect ten of them to be left-handed, and the odds of finding no left-handers would be very small. Thus our ability to draw a conclusion on the incidence of left-handedness depends sensitively on the number of people asked or sampled. Astronomy has exactly the same situation. If I inspected the nearest ten stars to the sun and found them to be all of a certain type, would I be justified in concluding that that was the only type of star that existed? No, I would probably need to sample space more fully and contain larger samples of stars to make a statistical statement with any certainty. Astronomy depends on the issues of sampling as many other fields do. For instance in politics the voting intensions of tens of millions of adult Americans are based on samples of only a few thousand people. So these techniques can be used with reliability.