The variation in sample estimates that would be obtained if many different samples were drawn from the same population.