This editorial says that no inferential statistical procedures are required. Then it goes on to say that "strong" descriptive statistics including effect sizes are required. Hmmmm, are these not point estimates of parameters, and as such inferential statistics? If not, are we to assume that the sample is the entire population of interest? Then he goes on to write "Finally, we encourage the use of larger sample sizes than is typical in much psychology research, because as the sample size increases, descriptive statistics become increasingly stable and sampling error is less of a problem. "Stable" and "sampling error" do not make any sense, to me, in the absence of sampling distributions and thus inferential statistics. What am I missing here?
Cheers, Karl L. Wuensch -----Original Message----- From: Teaching and Learning Statistics [mailto:[email protected]] On Behalf Of dmr Sent: Wednesday, February 25, 2015 9:39 AM To: [email protected] Subject: Basic and Applied Social Psychology It's interesting if you read David Trafimow's editorial in 2014 about his becoming the new editor of the Journal ... that HE has a personal agenda that he wants to inculcate to the Journal. However, in the first policy statement in that editorial, he says that the Journal will NOT REQUIRE null hypothesis testing and that authors are free to decide if and when they might use such methods. Then of course, in the latest issue of BASP ... he says the "grace period" is over ... and now such use is banned. Not only NHST ... but CIs ... and that any author would have to strip such references to p values and the like prior to publication. --- You are currently subscribed to tips as: [email protected]. To unsubscribe click here: http://fsulist.frostburg.edu/u?id=13090.68da6e6e5325aa33287ff385b70df5d5&n=T&l=tips&o=42305 or send a blank email to leave-42305-13090.68da6e6e5325aa33287ff385b70df...@fsulist.frostburg.edu
