From: Dan GrunerSubject: power analysis software: digest Hello again, Below you will find a digest of responses to my request for information on power analysis software and shareware. A couple of the links seem very useful as they provide links to the software providers. These you will find as the first two entries in this digest. A big thanks to all those who responded! Dan ------------------- A comprehensive list of power analysis software for microcomputers This page is an attempt to provide an up-to-date list of microcomputer software that can be used to calculate the power of statistical hypothesis tests. It is maintained by Alejandro Escalante, http://www.insp.mx/dinf/stat_list.htmlnts/s-archive/ ------------------ It has links to the following series of programs that do power analysis, so you can download many of the programs. Some of these are freeware/shareware, while others are commercial (some very expensive). It gives the hardare requirements, who to contact, and that sort of thing. Hope this helps. http://www.interchg.ubc.ca/cacb/power/nts/s-archive/ CRCSSIZ Datasim Design Power ECHIP Egret SIZ EpiInfo Ex-Sample and Ex-Sample + GAUSS GLIM G*POWER Hyperstat JMP MacAnova MONITOR Microsoft Excel MODSTAT MSUstat N NCSS NCSS-CALC NCSS-PASS NONCDIST and SMPLSIZE nQuery advisor NSurv PC-SIZE and SIZE Power Power & Effect Power and Precision POWER.FOR PowerPack PowerPlant POWPAL Precision of survival studies PS R2 Sample Power SAS SHAZAM Sigmastat SIMSTAT S-PLUS SPSS SOLO Power analysis SSIZE SST STATGRAPHICS Plus Statistica Statistical power analysis StatMate STAT-POWER Stats STPLAN Stata SYSTAT Design TRENDS True EPISTAT XLISP-STAT wdist --------------------- There is a good and free of charge program, G*Power available at http://www.psychologie.uni-trier.de:8000/projects/gpower.html --------------------- S-plus does power analyses for a number of different experimental designs. You'll have to check and see if it works for your design, but I suspect it will. ------------------- I have used Monitor which I got off the Web from a USFWS website quite simple and user friendly. ------------------- If your experimental design will fit the framework of ANOVA or general linear models, and if you use SAS for your data analysis, there are a couple of options in SAS. There are at least 2 sets of macros available from SAS Institute's web server (and probably included in the sample library on the cd-rom) that perform moderately sophisticated power analyses. There is also proc PLAN for generating complex designs, although I haven't used it. They also have code for setting up lattice designs, but that's not usually a question of statistical power. -------------------- JMP (a SAS product) conducts power analyses and is available for the MAC, and I believe for PCs. ------------------ See Thomas, L. and C.J. Krebs. 1997. A review of statistical power analysis software. Bulletin of the Ecological Society of America 78(2): 128-139. See also the following web address for a list of power analysis software (you can even download a reprint of the review) ----------------- SigmaStat (SPSS Inc.)provides power and sample size analysis for t-tests,proportions,ANOVA,chi-square,and correlation. --------------- This probably isn't what you had in mind, but JMP (JuMP), a windows-based package put out by the SAS institute can calculate power for a variety of experimental designs. You can vary all the parameters in any combination and get a complete list of power levels for all combinations entered. JMP is a general stats package so it may cost more than you had in mind for something just to do power analysis. I don't know the cost as I have only used a demo copy. ----------------------- i've used a very nice package called Power and Precision. it is available for windows, at least. they have a webpage, although i don't have the URL handy at the moment; try searching "power and precision". ------------------ Minitab v.11 or the newest is 12. Try, info@minitab.com & http://www.minitab.com/.ni-trier.de:8000/projects/gpower.html This is the MOST complete statistical software I have seen and I have worked as a consultant for three years and have recently attained a v.11 release for my home use. Not free, but nothing truely worth having is free of investment :). -------------------