Retraction Watch

Tracking retractions as a window into the scientific process

Study about words’ effect on mood to be retracted after investigation finds evidence of data manipulation

with 19 comments

A study examining whether the verb tense you use to describe a positive or negative experience influences your current mood will be retracted after a university investigation found the data had been manipulated.

By whom is the question — the notice cites an unnamed graduate student as the source of the manipulation, and says the only author, William Hart, was unaware of what had occurred.

We spoke with Hart, based at the University of Alabama, who declined to identify the student, nor say whether he or she was still working at the university. He did say the experience has been trying:

You feel devastated. You feel…yeah, I don’t know. Very uncertain. Worried, embarrassed. And honestly concerned for the student. I knew this individual quite well.

Hart said he and others first learned the data were problematic after posting the student’s data from another project online. A scientist outside the lab flagged the data as being inconsistent with what was reported, which led to additional questions, he told us. Those questions

led to the individual who posted and collected the data to come forward that the data posted online were altered strategically to yield a particular conclusion…the student was forthcoming about which papers were at stake.

Hart said his lab will be issuing additional retractions as a result, but declined to specify how many, citing the university’s investigation.

Regarding the specific paper that will be retracted from Psychological Science, “Unlocking past emotion: verb use affects mood and happiness,” Hart told us:

There has been an investigation, and it is known that the data were manipulated. Exactly how, or to what extent, is not entirely clear at this point.

The 2013 paper has been cited four times, according to Clarivate Analytics’ Web of Science, formerly part of Thomson Reuters. It found that study participants who talked about a previous happy time using verbs in the imperfect tense — meaning “I was laughing” — were in a better current mood than people who used verbs in the past tense (“I laughed”). The same trend applied to negative experiences — using imperfect tenses, which imply an ongoing experience, was associated with more of a negative current mood in study subjects than using past tense. It received some press coverage at the time it appeared, such as from The Huffington Post.

Hart explained the student was not listed as an author on the paper because he or she only collected the data, and didn’t participate in writing the manuscript.

He added that the experience has led him to change how data are stored in his lab. Before, lab members would save information on their computers, then download it to a common hard drive. Now, all the information is stored online, and if anyone tries to delete it, it can easily be restored.

All the data now is going to be in a place where I can check it, and it can’t go anywhere.

We want to offer kudos to the editor of the journal, Stephen Lindsay, who alerted us to this retraction before it appeared. He told us:

Psychological Science is committed to promoting openness and transparency.  We are also striving to increase vigilance to methodological and statistical issues during the review process.

Lindsay also sent us the text of the upcoming notice:

The retraction follows an investigation by the University of Alabama’s Office for Research Compliance. That investigation found that a former graduate student in William Hart’s lab altered the data in strategic ways. The investigation found that William Hart was unaware when the article was published that the data had been manipulated. William Hart cooperated in the investigation and agreed to this retraction.

Like Retraction Watch? Consider making a tax-deductible contribution to support our growth. You can also follow us on Twitter, like us on Facebook, add us to your RSS reader, sign up on our homepage for an email every time there’s a new post, or subscribe to our daily digest. Click here to review our Comments Policy. For a sneak peek at what we’re working on, click here.

Comments
  • Andy Patterson February 7, 2017 at 9:35 am

    The paper is strange — it does not have any “Acknowledgments” section indicating that anyone else was involved with the research. Instead, it is a single-author paper that says, e.g., “I tested…” “I examined” “I measured…” etc. Who is the “I” doing the work, if not the author?

    • Colleen McMichael, Ph.D. February 8, 2017 at 10:53 am

      Agreed!

  • TL February 7, 2017 at 9:38 am

    Is it technically possible to throw your co-authors under the bus and claim innocence when you are the sole author? How concerned is Prof Hart about publishing work using other people’s data without including them as co-authors?

    • anon February 8, 2017 at 10:39 am

      Many journals require an attestation to the effect that all authors claim responsibility for the integrity of the data

  • Lee Rudolph February 7, 2017 at 10:08 am

    Several direct quotations from RW’s conversation with Prof. Hart appear above. It would be interesting for the RW interviewer to compare her/his impression of Hart’s mood during the conversation with Hart’s verb use therein. Hey, if RT would post a recording of the conversation, we could all assess it and write a joint paper!

  • R. Grant Steen February 7, 2017 at 10:16 am

    Had Dr Hart included the student who collected the data as a co-author, the outcome might have been quite different. At very least, this represents a missed opportunity to act as a good mentor. At worst, it is unethical publishing behavior. It is not possible to throw your co-authors under the bus when you don’t have any co-authors; your name is your guarantee.

  • Anon February 7, 2017 at 10:18 am

    If the doctoral student was involved in collecting the data, why was such person not acknowledged? What is even more appalling is that the journal bought into this BS and agreed to cite the student in the notice! If the student’s name was on the paper it would entirely be a different story. In the next retraction, may be the author can cite the “ethereal elements in the online ‘cloud’ storage changed the data.”

    Then again, several papers with blatant errors in Psychological Science have been flagged on PubPeer, but the editor chooses to ignore these issues that can be verified just by looking at the reported numbers. Instead of grandstanding on “openness and transparency,” may be the editor needs to get his act together and check several of these papers on PubPeer.

    Please stop taking credit for doctoral student’s work when convenient, and when #@!% hits the fan blame these students again!

  • Rolf Degen February 7, 2017 at 11:59 am

    Interestingly, the findings of this study also failed to replicate in a thesis that was dedicated to just that point. http://diginole.lib.fsu.edu/islandora/object/fsu:204778/datastream/PDF/view

  • Paul A Thompson February 7, 2017 at 2:37 pm

    I am not convinced that participation in data collection is enough to get a publication credit. However, I would like to know what has happened with the person who committed the data manipulation. Have corrective teachings been given? I don’t necessarily believe that all persons who do this kind of thing are inherently condemned. It would be good if the University told us what they did, and if they believe that this person understands the errors of previous practice.

    • Andy Patterson February 7, 2017 at 5:19 pm

      Perhaps the point is that this “graduate student” was not even mentioned in the (non-existent!) Acknowledgments. And, the paper is written to make it appear that the sole author did the work himself.

  • Dave Fernig February 7, 2017 at 5:30 pm

    Well, a new variant on “I’m the PI, it’s those pesky underlings at fault”. My sympathy for the PI is zero. Either you do your job as a university Professor or you don’t. Manipulated data may on occasion sneak by, but at least you will have a reasonable degree of understanding and control of those data.

  • Rob Ward February 8, 2017 at 5:12 am

    I agree the author is having it both ways by not including a co-author, but also saying he is not responsible for the data.

    HOWEVER, how and why would a student who is not a co-author, manipulate the data to create a significant finding?

  • herr doktor bimler February 8, 2017 at 5:19 am

    Hart said he and others first learned the data were problematic after posting the student’s data from another project online. A scientist outside the lab flagged the data as being inconsistent with what was reported, which led to additional questions, he told us. Those questions

    led to the individual who posted and collected the data to come forward that the data posted online were altered strategically to yield a particular conclusion…the student was forthcoming about which papers were at stake.


    So the student has filled the data-collection role for a number of projects and publications. Was he or she being paid for this? Were they the student’s own honours / masters dissertations? I am trying to understand why he or she was so committed to provide data that fitted the hypotheses.

  • Andy Patterson February 8, 2017 at 8:29 am

    And why wasn’t this “student” named in the acknowledgments? Why weren’t there any acknowledgments at all?

  • Paul A Thompson February 8, 2017 at 9:45 am

    The more I think about this, the dodgier it sounds. We have an unnamed individual. This individual, although that person has no stake in the research, manipulates the data inappropriately, to the extent that the paper is retracted. We do not know at WHAT STAGE the manipulation took place. Presumably the University has examined the data. Depending on the manner in which the data were recorded and stored, manipulation could take place at several points. More shoes will drop on this one, I am thinking.

    • Anonymous February 8, 2017 at 10:09 am

      Above it is mentioned that “The retraction follows an investigation by the University of Alabama’s Office for Research Compliance. That investigation found that a former graduate student in William Hart’s lab altered the data in strategic ways. The investigation found that William Hart was unaware when the article was published that the data had been manipulated. William Hart cooperated in the investigation and agreed to this retraction.”

      I wonder if the Office for Research Compliance has to make the reports of this investigation available upon request.

  • another anon February 11, 2017 at 6:11 pm

    At best, we have a situation where a prof was asleep at the wheel, and showed himself a very poor judge of character: it’s doubly embarrassing that someone from outside his own group had the embarrassing task of telling him he’d failed to notice his own underling had manipulated his own data. At worst we have a situation where the prof has claimed work as his own unaided product, and when it’s turned out to be fraudulent, tried to disassociate himself from the blame. Who funded/was funding this group? Are they happy?

    And as for the University of Alabama’s office for research compliance as quoted above, it sounds like their statement basically translates to “the sole author had no idea where the data came from, so it’s OK that they were fraudulent.” I’m lost for words.

    • herr doktor bimler February 12, 2017 at 3:54 pm

      Just came here to post the same link. Morey explains why the results reported in the paper were preposterous.

  • Post a comment

    Threaded commenting powered by interconnect/it code.