Reminds me of the old literature on personality traits. Number of people argued against traits and in favour of the situation on the basis of failure to find substantial correlations between measures (e.g., of honesty) across situations. Failure to replicate? Much of the problem turned out to be due to unreliable measures (e.g., single items).
Jim Jim Clark Professor & Chair of Psychology University of Winnipeg 204-786-9757 Room 4L41A (4th Floor Lockhart) www.uwinnipeg.ca/~clark -----Original Message----- From: Mike Wiliams [mailto:[email protected]] Sent: September-03-15 12:39 AM To: Teaching in the Psychological Sciences (TIPS) Subject: Re:[tips] NY Times Article on Reproducibility Psychologists have been using poor research methods for so long that we think our current methods are valid. We have been wise enough to detect these problems and often comment on them, and even study them, but we don't change them because there is essentially no correction. It's like the old quote about the weather: everybody talks about it but no one does anything about it. A good example I use in stats classes is reliability. Psychologists have actually made contributions to the study of measurement because our measures are so unreliable. I wonder how many studies will not replicate or have stable effect sizes if the dependent measures only have reliabilities of .8? If the dependent measures can't be improved, we still forge on using them as if they were perfectly valid and reliable. Of course, one consequence of this is a poor rate of replication. Mike Williams Drexel University On 9/3/15 1:00 AM, Teaching in the Psychological Sciences (TIPS) digest wrote: > Subject: RE: NY Times Article on Reproducibility > From: "Mike Palij"<[email protected]> > Date: Wed, 2 Sep 2015 08:54:34 -0400 > X-Message-Number: 1 > > On Tue, 01 Sep 2015 08:02:05 -0700, Jim Clark wrote: >> >Hi >> > >> >Piece in NY Times by psychologist defending the discipline. >> >http://www.nytimes.com/2015/09/01/opinion/psychology-is-not-in-crisi >> >s.html?emc=edit_th_20150901&nl=todaysheadlines&nlid=26933398&_r=0 >> > >> >Judging by comments, readers aren't buying the argument. > Maybe Scott Lilienfeld should write an Op-Ed piece because of his > background on reviewing psychology as a science vs being a > pseudoscience. He hasn't commented on the reproducibility project but > one imagines that he may have some useful insights as well as > explanations that go beyond "this is just an example of the > self-correcting nature of science". > > -Mike Palij > New York University > [email protected] --- You are currently subscribed to tips as: [email protected]. To unsubscribe click here: http://fsulist.frostburg.edu/u?id=13251.645f86b5cec4da0a56ffea7a891720c9&n=T&l=tips&o=46630 or send a blank email to leave-46630-13251.645f86b5cec4da0a56ffea7a89172...@fsulist.frostburg.edu --- 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=46632 or send a blank email to leave-46632-13090.68da6e6e5325aa33287ff385b70df...@fsulist.frostburg.edu
