Granted I am a hack who teaches intro stats (occasionally intermediate level) and not an actual research scientist but I always like to joke (and keep my students awake) by exclaiming that "Size doesn't count...except in statistical sample size." and (of course) 'Bigger IS better." ^ - Nancy Melucci Long Beach City College Long Beach CA -----Original Message----- From: David Epstein <da...@neverdave.com> To: Teaching in the Psychological Sciences (TIPS) <tips@fsulist.frostburg.edu> Sent: Sun, Jun 15, 2014 10:20 am Subject: Re: [tips] Sample sizes
On Sun, 15 Jun 2014, Annette Taylor went: > Now I am reading one of several papers on newer method in stats and > it appears that much greater sample sizes are in order. A paper in > Perspectives on Psychological Sciences is even advocating samples > over 500 if I would want 90 percent power. Back in the day that > would have been denounced as the very worst kind of stats thinking. Quick answer: Overpowering a study would indeed be a problem if you were going to base your conclusions on whether a null hypothesis could be rejected at some particular alpha level. But recent criticism of studies with too-small sample sizes has gone hand in hand with criticism of that "null-hypothesis significance testing" approach. The recommendation I typically see in these critiques is that you should formulate your hypothesis in terms of an effect size (which might itself reflect the ABSENCE of an appreciable difference, as in, "The difference in IQ between men and women should be no more than d = Some-Minuscule-Number-With-No-Practical-Significance") and then power your study so the 95% confidence interval is nice and narrow. With that approach, bigger is generally better, because there's not really any such thing a false positive--just more and more precision. --David Epstein da...@neverdave.com --- You are currently subscribed to tips as: drna...@aol.com. To unsubscribe click here: http://fsulist.frostburg.edu/u?id=12993.aba36cc3760e0b1c6a655f019a68b878&n=T&l=tips&o=37151 or send a blank email to leave-37151-12993.aba36cc3760e0b1c6a655f019a68b...@fsulist.frostburg.edu --- You are currently subscribed to tips as: arch...@jab.org. To unsubscribe click here: http://fsulist.frostburg.edu/u?id=13090.68da6e6e5325aa33287ff385b70df5d5&n=T&l=tips&o=37152 or send a blank email to leave-37152-13090.68da6e6e5325aa33287ff385b70df...@fsulist.frostburg.edu
