SAT scores were originally scaled so that the scores in each section were approximately normally distributed with a mean of 500 and a standard deviation of 100. Use the empirical rule to estimate the probability that a randomly-selected student gets a section score higher than 600.
