Retraction Watch

Tracking retractions as a window into the scientific process

“Positivity ratio” research now subject to an expression of concern

with 9 comments

An expression of concern has been issued for the second of three papers on the idea that, if you have three positive emotions for every negative one, you will be more successful in life.

Psychologist Barbara Fredrickson, of the University of North Carolina, Chapel Hill, has spent the last decade building a brand around this ratio, initially described by a mathematical equation based on fluid dynamics by mathematician Marcial Losada. You can read our coverage of the debunking of that equation, presented in a 1999 paper that has been cited nearly 1,700 times, here.

Nick Brown, co-author with Alan Sokal on the paper that discredited the Losada equation, has written a blog post on the current state of affairs. He also got in touch with us regarding the expression of concern for a 2004 article in American Behavioral Scientist that he had also questioned, “The Role of Positivity and Connectivity in the Performance of Business Teams: A Nonlinear Dynamic Model”:

The Editorial Board and Publisher of the American Behavioral Scientist hereby issue an Expression of Concern regarding the nonlinear dynamic model and sebsequent model-based predictions in the article entitled “The Role of Positivity and Connectivity in the Performance of Business Teams: A Nonlinear Dynamic Model” published in the February 2004 issue of the American Behavioral Scientist.

The non-linear dynamic model employed was first introduced in Losada (1999), expanded upon here and in Fredrickson & Losada (2005).

This set of three papers was critically examined in a 2013 American Psychologist article (Brown, Sokal & Freidman, 2013). Fredrickson and Losada (2005) was subsequently corrected by Fredrickson (with Losada’s assent). Specifically, the non-linear dynamic modeling element of the 2005 article was withdrawn as invalid and, along with it, the model-based predictions about the particular positivity ratios of 2.9 and 11.6. Other elements of the 2005 paper were upheld as valid, including the empirical evidence that linked higher positivity ratios to flourishing mental health (Fredrickson, 2013).

As with Fredrickson & Losada (2005), Losada has chosen to not defend or comment on his non-linear dynamic model despite repeated requests over several months from the American Behavioral Scientist editorial board or the special issue editor of the issue in which Losada and Heaphy (2004) appeared.

As such, the non-linear model used in the Losada and Heaphy (2004) may be invalid and the resulting predictions may be unfounded.

Heaphy has been available for comment, but was not involved in the data collection, analyses, or interpretation of the analyses, which are of central concern here. She became involved after the paper had going through one round of peer reviews, invited by the Special Issue Editors to help Losada link the findings to relevant organizational research and to make the paper understandable to an organizational audience. Thus, her contribution to the paper is not related to the expression of concern.

You can take a short quiz on Fredrickson’s website, if you’re interested to see where you stand.

Written by Cat Ferguson

July 29th, 2014 at 8:30 am

Comments
  • failuretoreplicant July 29, 2014 at 11:34 am

    I wonder if the publisher of Fredrickson’s self-help book will be offering refunds since the research behind it has been largely discredited.

  • mickeymusing July 29, 2014 at 12:32 pm

    Any chance this will help end the tyranny of the positive thinking brigade? So, so tired of having to explain to board members the difference between ‘negativity’ and ‘actual reality.’ IMO, forced positivity and constant (often inappropriate) cheer leading are just insidious ways to get out of taking responsibility while assuming moral superiority. Biggest offenders seem to be peddlers of woo (and I must say Dr. Fredrickson’s website set off a number of woo alarms) and MBA’s–both of whom use meaningless terms to sound authoritative and hide the fact that they haven’t a clue what they’re talking about. I mean, if you aren’t willing to synergize your core competencies to leverage the dynamic energy all around us, how do you expect to scale your optimum life experience and achieve transcendence? It’s clearly all in your attitude and outlook. And also sometimes a prayer circle helps (actually had a board member recommend this to solve a budget crisis). Sigh…

  • Neuroskeptic (@Neuro_Skeptic) July 29, 2014 at 12:36 pm

    The last paragraph is very odd. “Heaphy became involved after the paper had going through one round of peer reviews, invited by the Special Issue Editors to help Losada link the findings to relevant organizational research and to make the paper understandable to an organizational audience.”

    Such “matchmaking” by editors, at such a late stage, is quite unusual.

    • Nick Brown July 30, 2014 at 4:56 pm

      Since the start of this saga, I’ve always assumed that the Losada and Heaphy article represented an attempt by an unspecified collection of people at the UMich business school (where Fredrickson spent a year around this time, I believe) to take Losada’s “groundbreaking” ideas and give them a haircut, shave, and a clean shirt, to make them suitable to go out in public (i.e., in the genteel pages of a business journal).

      However, the revelation that Heaphy only got involved after Losada had submitted the article seems very odd. I doubt whether “why don’t you bring on a grad student to make your article intelligible” is a very common response from journal editors.

      I also note that the issue of American Behavioral Scientist in which the article appeared was a special edition, one of whose editors was a senior faculty member at the UMich business school. But doubtless this is entirely coincidental.

      • JATdS July 30, 2014 at 9:26 pm

        Nick, in your field of study, and from your experience, would you say that it is common for editors to suggest the inclusion of an “additional” author, perhaps a senior member of the peer pool (as opposed to a grad student), fater a paper has been submitted, with the sole purpose of making it “intelligible”?

        • Nick Brown August 1, 2014 at 4:47 pm

          No idea. I’m new here. When I say that the inclusion of a grad student seems very odd, that’s based on a feeling of how things ought to work, not a large sample of previous experiences. 🙂

  • herr doktor bimler July 30, 2014 at 7:40 am

    As such, the non-linear model used in the Losada and Heaphy (2004) may be invalid and the resulting predictions may be unfounded.

    What has Losada’s failure to respond got to do with the invalidity of his ‘model’? It would remain invalid even if he did bother to respond to the editors’ second thoughts.

  • Post a comment

    Threaded commenting powered by interconnect/it code.