You are here

Generating Random Sequences

Author(s): 
Mike Shor

This mathlet allows the user to input a sequence of 0's and 1's of length 20 and then tests the sequence for randomness using three tests: Kolmogorov-Smirnov, Wald-Wolfowitz, and Mann-Whitney. An overall assessment of how likely the sequence was generated randomly is given, as well as basic information on how the tests work.

Mike Shor is in the Owen Graduate School of Management at Vanderbilt University .

INTENDED USES:

  • Class demo
  • Student use

APPROPRIATE COURSES:

      Probability, Game Theory, Statistics

AUTHOR'S STATEMENT:

This applet was written for a game theory course to demonstrate how difficult it is for people to generate "random" sequences in the context of playing mixed strategies. It has also proved useful in showing the sensitivity of several commonly-used non-parametric and runs tests. Users simply enter a series of 0's and 1's, and the applet reports the results of three statistical tests, along with a more user-friendly (and subjective) determination of whether or not the sequence is likely to be random.

Open Generating Random Sequences in a new window.

AVAILABILITY OF CODE:

The mathlet page includes documented source code for the mathlet. This code is copyrighted, but may be adapted as desired and used freely for non-commercial purposes provided the original source is acknowledged and the author is informed.

Published December, 2001
© 2001 by Mike Shor

Mike Shor, "Generating Random Sequences," Convergence (December 2004)