Sources of Error

 

Home
Project Information
Introduction
Purpose/Questions
Hypotheses
Procedures
Results
Statistical Analysis
Sources of Error
Conclusions
Bibliography
Graphs of Data
The Test Taker

 

Sources of Error

There was only one source of error in this investigation. However, this error was unpreventable. The source was the computer itself! In order to simulate the students guessing on the answers, several random number generation procedures were used in the "The Test Taker". The problem lays in the way these random numbers are generated. These numbers are not actually randomly generated; instead they are generated by long mathematical functions involving mod 232 and beyond! Since a mathematical function is predictable, one can predict the next number and thus these numbers are not random. Therefore, all numbers generated this way are classified as "pseudorandom" numbers.

One cannot avoid this problem as all computers are algorithmic devices. They follow certain algorithms (procedures) at all times. To put it bluntly, they are machines carrying out sets of instructions, making them impossible to come up with a random number as they have to follow a mathematical function (when generating a "random" number).

The only way to avoid such a problem is to use a physical random number generator device such as a coin or a die. However, this means carrying out 800 000 trials by hand! Clearly, this is impossible... making this source of error inevitable.

 

  Home | Project Information | Introduction | Purpose/Questions | Hypotheses | Procedures | Results | Statistical Analysis | Sources of Error | Conclusions | Bibliography | Graphs of Data | The Test Taker

This site was last updated 04/28/07

© Copyright 2007 by Ryan Peng. All rights reserved.