A language called
*Resampling Stats* and a program by the same name were developed
by Dr. Julian Simon and Peter Bruce as a new way to teach Statistics
to social science students. The history, description, and application
of the Resampling method to a vast range of statistical problems from
simple to most complex are described fully in Dr. Simon's book
__ Resampling:
The New Statistics__, which is available free
online.

The original Resampling
Stats program has been discontinued, but the language lives on in
*Statistics101*. *Statistics101* is a clone of the
*Resampling
Stats*__
program__, in that it implements most of *Resampling
Stats*' commands and adds many new commands. But *Statistics101*
is not a clone from the viewpoint of its user interface, which is an
independently created, very powerful program development system
complete with full debugging support. *Statistics101* is written
in 100% Java and will run on Microsoft Windows, UNIX, Linux, and
Apple Macintosh - any operating system that supports Java 1.4 or
better.

*Statistics101* executes programs written in
a greatly expanded version of the *Resampling
Stats*__
language__. *Resampling Stats* is a statistical
simulation language. The language is used to describe the process
behind a probability or statistics problem. That process description,
when executed by *Statistics101* computes probability and
statistics solutions without using arcane formulas or tables.
*Statistics101* takes advantage of the power of the personal
computer to quickly run your *Resampling Stats* model thousands
of times, each time with different random numbers or samples, keeping
track of the results. When the program completes, you have your
answer.

This document describes how to use *Statistics101*
and lists and describes in detail each of the commands supported by
*Statistics101*.

As a very simple example, say you wanted to know
the probability of getting exactly two heads in a toss of three
coins. You could toss three coins many times, counting the number of
times you got exactly two heads and dividing by the number of tosses.
That would take considerable effort and time. You could also
calculate it precisely if you knew the correct formula. Instead, with
*Statistics101*, you could model that process as follows (text
following a single quote to end of line is a comment):

URN (0 1) coin ' let 0=tails, 1=heads REPEAT 1000 ' repeat the following 1000 times SAMPLE 3 coin toss ' simulate toss of 3 coins COUNT toss =1 heads ' count number of heads SCORE heads results ' store heads count in "results" END ' end of repeat COUNT results =2 successes ' count how many results were exactly 2 DIVIDE successes 1000 probability ' probability = successes/1000 PRINT probability

The program simulates 1000 tosses of three coins and prints out the resulting probability. The output looks like this:

probability: 0.368

Here's what the above program is doing:

Put the numbers 0 and 1, representing tails and heads, into an "urn" named "coin".

Repeat the following three commands 1000 times:

Take three samples at random, with replacement, from the "urn". This is equivalent to three tosses of a coin.

Count how many of the tosses were equal to 1 (i.e., heads).

Record the number of heads in the "results" vector or list.binomialprob

Count how many of the 1000 results in the results vector were equal to two, i.e., two heads.

Calculate the probability by dividing the number of successes by the number of trials (1000).

Print the probability.

There are several example *Resampling Stats*
programs included in the package. Some you'll recognize from __Dr.
Simon's book__
(http://www.resample.com/content/text/index.shtml)
and/or the *Resampling Stats* __website__
(http://www.resample.com). You
can access them through the **Help>Example
Programs** menu of the *Statistics101* program.
Many more examples are available from
http://www.resample.com/content/examples/index.shtml.
All the examples will run in *Statistics101*.