You are currently viewing a sample of the Cram Kit. Click here to unlock everything.
A prediction interval gives a range in which we're confident a random observation will fall in if we "pluck" it from the population.
A confidence interval gives a range in which we're confident the true population parameter falls in.
When you're plucking an observation at random from a given population, there's way more room for variability / randomness than when you're trying to assess that population's mean.
