Overblog
Editer l'article Suivre ce blog Administration + Créer mon blog
Bioécologie

Analyse de puissance : logiciel G*Power

21 Février 2009 , Rédigé par JMB Publié dans #Statistique - modélisation - plans expérimentaux

Un logiciel gratuit pour réaliser des analyses de puissances : G*Power
G*Power 3 is a statistical power analysis program for MacOS X 10.4 and Windows XP/Vista. It is a major extension of, and improvement over, the previous version, covering many different statistical tests of the F, t, chi-square, and z test families as well as some exact tests. G*Power 3 provides improved effect size calculators and graphics options, it supports both a distribution-based and a design-based input mode, and it offers five different types of power analyses. Like its predecessors, G*Power 3 is free.
Download and register G*Power 3
Pour en savoir un peu plus sur les analyses de puissances, vous pouvez lire cet article : L. Thomas et C.J. Krebs (1997). "A review of statistical power analysis software".

How to Use G*Power

by Axel Buchner, Edgar Erdfelder, and Franz Faul.

First published on the WWW: August 28, 1997; last update: March 28, 2001.


This manual was designed for G*Power 2. The next version,
G*Power 3, is now available. The information provided here may still be helpful, however, because the help pages for G*Power 3 will remain incomplete for some time to come.

This is what you find here:

  1. a short tutorial
    for first time users contains an example which shows, step by step, how to do power analyses with G*Power;

  2. a user interface description
    explains how G*Power works (and contains lots of graphics, so be prepared to see slow downloads);

  3. a reference section for experienced users describes how different
    1. types of power analyses are done for a number of
    2. statistical tests and gives some details about the
    3. underlying theoretical concepts;
  4. a programmatic article
    provides a brief overview of the use of power analyses in behavioral research, explains why G*Power was developed, and informs about the algorithms used in the program;

  5. a list of questions and answers
    about how to perform power analyses with G*Power may help you with a specific problem;

  6. a list of the people who were very helpful in improving G*Power and this guide to G*Power.

Jorge Camacho-Sandoval, Costa Rica (jcamacho@ice.co.cr) has compiled a PDF version of the tutorial and the reference section which he kindly provided for download.


Where do I get the latest version of the program?

Check out our G*Power 2 web page to see if you do not have it already. Please note that the next version, G*Power 3, is now available. You may still use G*Power 2, and in some cases you have to: If you still use MS-DOS or MacOS 7 to 9 as operating systems, you are limited to G*Power 2. For all others, G*Power 3 is highly recommended.

 

Note

The information presented on these pages about statistical power analyses is designed to help users with different levels of experience in performing power analyses with G*Power. It must not be used for other purposes without explicit permission by the authors.

Although we have spent considerable time developing this web-based guide to G*Power, we can, of course, not guarantee that the text is free of errors. If you detect an error, or an omission which you think is crucial, then we would really appreciate if you could contact us.

If you want to cite this manual, this would be an appropriate format

Buchner, A., Erdfelder, E., & Faul, F. (1997). How to Use G*Power [WWW document]. URL http://www.psycho.uni-duesseldorf.de/aap/projects/gpower/how_to_use_gpower.html

 

About the terminology

Because HTML widely lacks features to display Greek and mathematical symbols, we used iso8859-1 (Latin-1) char set with fixed character width to display mathematical expressions.

Partager cet article
Repost0
Pour être informé des derniers articles, inscrivez vous :
Commenter cet article
W
Informative write-up on G*Power 3 program! I had read the programmatic article and it serves me the tips for using power analysis in behavioral research and also discusses the tips for developing the flowchart for the program!
Répondre