when to use true or false in normal distribution