different types of Power and examples

greenspun.com : LUSENET : History & Theory of Psychology : One Thread

I'm studying about the different types of Powers and examples any relevant information would be greatly appreciated

-- Kylie Dovaston (dovaston@alphalink.com), April 27, 2001

Answers

Response to I need lots of information on the different types of Power and examples

What sort of "power" did you have in mind? Statistical power? Strictly speaking power is the probablility of rejecting the null hypothesis when the null hypothesis is, in fact, false. Another way of thinking about it is:

1 - the probability of making a Type II error (beta).

(Type II error is failing to reject the null hypothesis when it is false, and so should be rejected.)

For more information, there are the famous books by Jacob Cohen, the "guru" of power in statistics. You might find those a little daunting, however. Easier to understand is his little paper, "A Power Primer" (Psychological Bulletin, 1992 Jul Vol 112(1), 155-159). One textbook I know of that has a very good chapter on statistical power is David C. Howell's _Statistical Methods for Psychology_. I believe the latest edition is the 4th (Wadsworth, 1997), but the earlier editions contain the chapter on power as well.

-- Christopher Green (christo@yorku.ca), April 28, 2001.


Response to I need lots of information on the different types of Power and examples

I echo Christopher's recommendations about sources for power. In addition, you might want to take a look at "Introduction to the Practice of Statistics" (Moore & McCabe, 1993?) for a section on power. It isn't nearly as thorough as Cohen's publications, but it presents power using z-calculations, and as the probability of detecting a "true" alternative given that the null is false. Also, although under a different name, that of "sensitivity" R.A. Fisher (1935, "Design of Experiment, early in the book) discusses how to increase the probability of rejecting a null hypothesis, mainly through larger samples and reducing within subject variability. Although I wouldn't call Fisher's "sensitivity" by the name of power in the strict sense, it is closely allied to it in principle, and might give you a useful perspective on the nature of power.

Dan.

-- Daniel J. Denis (dand@yorku.ca), May 05, 2001.


Response to I need lots of information on the different types of Power and examples

I would have interpreted this question entirely differently, thinking of power as an interpersonal/social phenomenon in psychology. There is a vast literature in family psychology and interpersonal psychology. If this is your interest, let me know and I can provide you with a few sources from my lecture notes.

-- Hendrika Vande Kemp (hendrika@fuller.edu), May 08, 2001.

Response to I need lots of information on the different types of Power and examples

It might be helpful to be able to mess around calculating values associated with power as you work with some of readings others have suggested. I know of two Windows programs that allow you to enter parameters that determine power (n, alpha, effect size, etc.) for some (not all) test statistics. One is called WDist95 (available at http://ourworld.compuserve.com/homepages/MSVonTress/wdist.htm) and the other, specifically for regression coefficients can be found at http://www.gsu.edu/~psyrab/BakemanPrograms.htm.

Trey Buchanan Department of Psychology Wheaton College

-- Trey Buchanan (Trey.Buchanan@wheaton.edu), July 18, 2001.


Moderation questions? read the FAQ