Dear Tipsters, To Trent: I was aware of the piece that you cite and that is indeed what started the thread. I was wondering whether there had been another report in a refereed journal where the studies might have been criticized. Chris and Linda draw our attention to the "obscure erratum" published by Rodin and Langer, where they report an error in their calculation for the data analysis of the death rate scores, and Chris presents a good case for the results being insignificant (see below).
However, the death rate data were only a small part of the outcomes of the two studies. It still seems to me that the design, execution and other data analyses are respectable. Sincerely, Stuart ______________________________ "Recti Cultus Pectora Roborant" Stuart J. McKelvie, Ph.D., Department of Psychology, Bishop's University, 2600 rue College, Sherbrooke (Borough of Lennoxville), QC J1M 1Z7, Canada. [email protected] (819)822-9600X2402 "Floreat Labore" ______________________________ From: Christopher Green [mailto:[email protected]] Sent: Tuesday, November 18, 2014 5:16 PM To: Teaching in the Psychological Sciences (TIPS) Subject: [tips] Rodin & Langer's obscure erratum Thank you Linda. If you follow that link, you only get a summary which says the z on mortality was re-calculated, whence if became "marginally significant." When you actually go to the original erratum, you find that, upon recalculation: z dropped from 3.14 to 1.74, (p=.0818). There is no reason to get overly-moralistic about the .05 level, but I think it is fair to say that p>.08 would not have been published, especially in JPSP. Going back to the original article, one sees that they calculated that z on the proportions, and did an arcsine transform on them first. Interesting because if you did a 2-way Chi-square on the raw frequencies (which would be the more common way to handle them): died lived Total plant 7 40 47 no-plant 13 31 44 Total 20 71 91 Chi-square (df=1) = 2.84, p > .05 (chi-square crit = 3.84) One can only guess why they went for the more exotic statistic. Interestingly, these frequencies generate a whopping Odds Ratio of 2.40 but, still, it is not significant (z=1.66)... and no one in psychology was using Odds Ratios back in 1977. The more common (in psych) effect size measure of phi (.177) looks even worse because of the imbalance in the table. Chris ..... Christopher D Green Department of Psychology York University Toronto, ON M3J 1P3 Canada [email protected]<mailto:[email protected]> http://www.yorku.ca/christo ....................................... On Nov 18, 2014, at 3:51 PM, Tollefsrud, Linda <[email protected]<mailto:[email protected]>> wrote: The "obscure erratum" link in Coyne's article leads to this: http://psycnet.apa.org/journals/psp/36/5/462/ L. Tollefsrud From: [email protected]<mailto:[email protected]> [mailto:[email protected]] On Behalf Of R. Trent Codd III Sent: Tuesday, November 18, 2014 2:37 PM To: Teaching in the Psychological Sciences (TIPS) Subject: Re: [tips] Psych science.? I don't see the beginning of the thread for some reason so I'm not entirely sure what you are referring to (although I have a good guess). Might you be referring to James Coyne's recent commentary about Ellen Langer's research? If not, you might find his thoughts on the matter interesting (although I suspect you may have already seen this & that this is what you are referring to): http://www.sciencebasedmedicine.org/eminent-harvard-psychologist-mother-of-positive-psychology-new-age-quack/ Trent On Tue, Nov 18, 2014 at 2:55 PM, Stuart McKelvie <[email protected]<mailto:[email protected]>> wrote: Dear Tipsters, I had always regarded the two nursing home studies (experiments, actually) as interesting and have regularly reported them in my classes. The significant finding of different death rates in the follow-up study was of particularly attention-grabbing and almost too good to be true, but I had not seen any reason to doubt it, even though the sample size was small. I went back to these two papers to see if I could detect serious errors in methodology and statistics. Of course, if there was something important that was not reported, we would not know that. Based on this re-reading I still do not see any serious errors, although the data on multiple measures could have been treated with MANOVA rather than ANOVA. The authors also report various attempt to keep extraneous variables constant - e.g. raters being blind as to the condition in which people were. In addition, they express their own surprise at the death rate data and admit that not everything was known about the patients. Of course, the matter of replication remains. If this has not occurred, we do not know what the outcome would be. As I mentioned in an earlier post, there has been a failure to replicate the results of the exercise-as-placebo experiment. So, overall, I still think that the original experiments, as reported, offer interesting results. Have there been serious criticisms of them that I have missed? Sincerely, Stuart ___________________________________________________________________________ "Floreat Labore" <image001.jpg> "Recti cultus pectora roborant" Stuart J. McKelvie, Ph.D., Phone: 819 822 9600 x 2402<tel:819%20822%209600%20x%202402> Department of Psychology, Fax: 819 822 9661<tel:819%20822%209661> Bishop's University, 2600 rue College, Sherbrooke, Québec J1M 1Z7, Canada. E-mail: [email protected]<mailto:[email protected]> (or [email protected]<mailto:[email protected]>) Bishop's University Psychology Department Web Page: http://www.ubishops.ca/ccc/div/soc/psy Floreat Labore" <image002.jpg> <image003.jpg> ___________________________________________________________________________ From: Michael Britt [mailto:[email protected]<mailto:[email protected]>] Sent: November 18, 2014 8:20 AM To: Teaching in the Psychological Sciences (TIPS) Subject: [tips] Psych science.? This is so discouraging. Eye opening perhaps, but discouraging. I remember well the nursing home study and I always thought positively of it. I have two parents in their 90s and I know they are frustrated by their lack of independence and the loss of control over their lives. But as I reflect on all this I had to ask myself, "Why would I think that the participants in Langer's study would lead healthier, longer lives simply because of their ability to take care of a plant?" Given how complex humans are, and how complex life is, why would I think that a simple "intervention" like giving people control over a plant would have such powerful effects? Maybe because I wanted to believe.... As for this counterclockwise "study"...oh boy..at least it is indeed an excellent point about how eminence doesn't necessarily mean credible. I am additionally discouraged because I recently finished reading a published article which appeared to have been carefully carried out (and which was filled with all manor of impressive advanced statistical techniques) but in the end all they really found were essentially correlations. I kept going back to my underlined sentences and I still couldn't figure out why this study was important enough to publish. The hypotheses and the conclusions were "tortured" into giving up some kind of "significance". I need some cheering up: can anyone point to a recently published article they think was interesting and credibly carried out? Michael Michael A. Britt, Ph.D. [email protected]<mailto:[email protected]> http://www.ThePsychFiles.com<http://www.thepsychfiles.com/> Twitter: @mbritt --- You are currently subscribed to tips as: [email protected]<mailto:[email protected]>. To unsubscribe click here: http://fsulist.frostburg.edu/u?id=13510.2cc18398df2e6692fffc29a610cb72e3&n=T&l=tips&o=40276 (It may be necessary to cut and paste the above URL if the line is broken) or send a blank email to leave-40276-13510.2cc18398df2e6692fffc29a610cb7...@fsulist.frostburg.edu<mailto:leave-40276-13510.2cc18398df2e6692fffc29a610cb7...@fsulist.frostburg.edu> --- You are currently subscribed to tips as: [email protected]<mailto:[email protected]>. To unsubscribe click here: http://fsulist.frostburg.edu/u?id=582499.3b862fcc285c6276efc1da9f00860b94&n=T&l=tips&o=40285 (It may be necessary to cut and paste the above URL if the line is broken) or send a blank email to leave-40285-582499.3b862fcc285c6276efc1da9f00860...@fsulist.frostburg.edu<mailto:leave-40285-582499.3b862fcc285c6276efc1da9f00860...@fsulist.frostburg.edu> -- R.Trent Codd, III, Ed.S., LPC Board Certified Behavior Analyst Diplomate & Fellow, Academy of Cognitive Therapy Academy of CT Certified Trainer/Consultant Clinical Faculty, Lenoir-Rhyne University Cognitive-Behavioral Therapy Center of WNC, P.A. 1085 Tunnel Road, 7A Asheville, NC 28805 828-350-1177 828-350-1188 (Fax) [email protected]<mailto:[email protected]> www.BehaviorTherapist.com<http://www.behaviortherapist.com/> Podcast: www.CBTRadio.org<http://www.cbtradio.org/> The information contained in this email message is intended only for the personal and confidential use of the designated recipient or entity named above. As such, this communication is privileged, confidential, and exempt from disclosure under applicable law. If the reader or actual recipient is not the intended recipient or an agent or employee responsible for delivering it to the intended recipient, you are hereby notified that you have received this document in error, and that any review, dissemination, distribution, or copying of this message is strictly prohibited by federal and state laws with civil and criminal penalties. If you have received this communication in error, we apologize for any inconvenience. We would appreciate your notifying us immediately by phone at 828-350-1177 or by email at rbaker@behaviortherapist. Thank you. --- You are currently subscribed to tips as: [email protected]<mailto:[email protected]>. To unsubscribe click here: http://fsulist.frostburg.edu/u?id=13355.5bffd68fb7c84ef12f478133e5791e9e&n=T&l=tips&o=40287 (It may be necessary to cut and paste the above URL if the line is broken) or send a blank email to leave-40287-13355.5bffd68fb7c84ef12f478133e5791...@fsulist.frostburg.edu<mailto:leave-40287-13355.5bffd68fb7c84ef12f478133e5791...@fsulist.frostburg.edu> --- You are currently subscribed to tips as: [email protected]<mailto:[email protected]>. To unsubscribe click here: http://fsulist.frostburg.edu/u?id=430248.781165b5ef80a3cd2b14721caf62bd92&n=T&l=tips&o=40290 (It may be necessary to cut and paste the above URL if the line is broken) or send a blank email to leave-40290-430248.781165b5ef80a3cd2b14721caf62b...@fsulist.frostburg.edu<mailto:leave-40290-430248.781165b5ef80a3cd2b14721caf62b...@fsulist.frostburg.edu> --- You are currently subscribed to tips as: [email protected]<mailto:[email protected]>. To unsubscribe click here: http://fsulist.frostburg.edu/u?id=13510.2cc18398df2e6692fffc29a610cb72e3&n=T&l=tips&o=40292 (It may be necessary to cut and paste the above URL if the line is broken) or send a blank email to leave-40292-13510.2cc18398df2e6692fffc29a610cb7...@fsulist.frostburg.edu<mailto:leave-40292-13510.2cc18398df2e6692fffc29a610cb7...@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=40296 or send a blank email to leave-40296-13090.68da6e6e5325aa33287ff385b70df...@fsulist.frostburg.edu
