This demonstration shows how thin-tailed the normal distribution is.

Press a button to add one or ten data sets, each generated the same way as in Assignment 1. The points in each data set are generated from a line, plus or minus a random normal error that is rounded to the nearest even integer. Each error's mean is 0 and its variance is 40.5, making the standard deviation about 6.4.

**Colors**: The first time a point is drawn, its color is magenta. When
a subsequently added data set includes the same point, the color is made slightly
darker, by subtracting some of its color.
For the 2nd through the 16th hit on the same point, a small amount of red
is subtracted, until the point is pure blue.
For the 17th through 31st hits, a small amount of blue is subtracted, until the point
is black. After 31 hits, the color is not changed any further.

The true line runs through the centers of the collections of points for each X value. You should see that even as hundreds of points pile up in the middle, within about 13 of the true line, very few points are more than 25 from the center.

To restart from a blank graph, you may have to close and reopen your
browser.

© 2000 Samuel L. Baker

The author is solely responsible for this page. Its contents have not been reviewed or approved by the University of South Carolina.

http://hspm.sph.sc.edu/COURSES/J716/demos/NormalDistribution/NormalDistributionDemo.html

February 1, 2000