High School Success

While trying to come up with an example of something that could not be accurately described using a Gaussian distribution, I was struck by how often normal curves are used to describe the academic performance of high school students.  Especially in my high school, I feel that academic performance could have been better described using a bi-modal rather than normal distribution. While a Gaussian distribution assumes that the majority of scores will fall near the mean with less and less percentages of scores falling one or two standard deviations from the mean, I believe that it is  far more common for high school test scores to fall on either the high or the low end of the spectrum. Because certain high schools require their students to have a certain amount of self initiative in order to be academically successful, I would expect the test scores to reflect this by producing a set of high test scores for the students who take the necessary self-initiative and a set of low scores reflecting students who do not take the initiative necessary to achieve academic success.