Professionalism/Marc Hauser and Research Fraud

Introduction

Marc Hauser was an American evolutionary biologist and associate faculty of the Harvard University Edmond J. Safra Center for Ethics, whose research topics focused on the intersection of evolutionary biology and cognitive neuroscience. He is described as a "trailblazer" in the field of animal cognition and was one of its most prolific publishers, having 143 articles, several books, and 221 authored or co-authored papers to his name. As a professor he commanded respect among his students, winning several teaching awards and earning distinction as one of the most popular professors by undergraduate vote in 2001 and 2008. His research focused on analysis of primates such as cotton-top tamarins and rhesus macaques, demonstrating that they had advanced cognitive abilities previously thought exclusive to humans. Dr. Hauser's access to tamarins for experimentation was noted to be mostly exclusive, making him a pioneer in the field.

Dr. Hauser's academic feats suggested a sterling reputation for interesting work that crumbled in 2010 when Harvard University found him guilty of research misconduct.

Initial Concerns

Despite his reputation at the time, Dr.Hauser's record was not completely clean before investigations started. An anonymous co-worker shed some light on Dr. Hauser's tendency to jump to conclusions from results that were publishable, but premature. The co-worker recounts a specific instance where Dr. Hauser wanted to use a specific statistical test that would have supported a publishable conclusion, but the qualifications needed to validate the test were not met. Dr. Hauser only backed down from publishing the work after a level of discussion that the co-worker found "definitely concerning," suggesting Dr. Hauser's motivation to publish was clouding his objective professional judgment.

A more specific instance of questionable conduct can found in 1995, when Dr. Hauser published a paper contending that cotton-top tamarins could recognize themselves in a mirror. At the time, only humans and great apes were known to be capable of this feat. Gordon Gallup Jr., the inventor of the mirror test, questioned the results and asked to see the recordings of the experiment. Upon review of sample tapes, Gallup noted that there "wasn't a shred of evidence" supporting Dr. Hauser's conclusion. Dr. Gallup requested more data, only to be informed by Dr. Hauser that the remaining tapes were stolen. In 2001, Dr. Hauser finally admitted in another paper that he could not replicate the findings.

Events Leading up to the Investigation

During the last months of 2006, a grad student working under Dr. Hauser ran a sound-pattern recognition experiment with Rhesus monkeys; recording the trials and preforming the first set of coding. The experiment, labeled AXA in the official Harvard investigation report, explored the hypothesis that the monkeys would respond more to ungrammatical stimuli compared to grammatical stimuli . The grad student found that his results did not support this hypothesis at all.

When Dr. Hauser received the trial recordings, he preformed his own coding and found results that supported the hypothesis of the AXA experiment as indicated in an email he wrote to one of his research assitants:

“	so i reanalyzed all the trials from -RETRACTED- run with AXA on rhesus, using a 2 sec cutoff post-pback... it is successful... will have -RETRACTED- recode some of these to make sure!	”
—Hauser to anonymous research assistant

When he sent back the trials to the grad student for recoding and validation, Dr. Hauser passed along new instruction on how the coding should be preformed, claiming the old criteria was producing false results.

When the grad student finished his recoding, Dr. Hauser had the research assistant he emailed review a small subsection of the trial results, 26 out 201, to confirm that recoding validated his findings. During this review, the research assistant noticed a number of differences between the grad student's original coding set and Dr. Hauser's coding set that called into question the validity of the experiment. The assistant believed that the circumstance called for recoding all 201 trials with new coders and emailed Dr. Hauser his suggestions. Dr. Hauser rejected this suggestion, claiming that the grad student's original data set was taken during the trial runs, called "online" data which is considered unreliable. The research assistant emailed Dr. Hauser a week later saying they were going to go ahead with the full recoding as the inconsistencies in the data sets were too much to ignore. That email elicited the following response:

“	on AXA, i am getting a bit pissed here. there were no inconsistencies! let me repeat what happened. i coded everything. The -RETRACTED- coded all the trials highlighted in yellow. we only had one trial that didn't agree. i then mistakenly told -RETRACTED- to look at column B when he should have looked at column D. B is the online coding. D is the offline coding. so there were no were no inconsistencies, and this was done as we have always done...	”
—Hauser to anonymous research assistant

The research assistant rebutted Dr. Hauser's claims that the grad student's original data set was online code. Additionally, the assistant pointed out that having 25 of 26 trials match should be cause for concern as blind and careful coders, which lab protocol dictates for these experiments, never reach reliability rates that high. Shortly after the research assistant resigned from laboratory due to Dr. Hauser's increasingly hostile behavior and alerted Harvard's ombudsman to Dr. Hauser's actions.

The University then appointed an Investigating Committee to look into the Scientific Misconduct allegations. The committee of 3 met numerous times over the next few months and interviewed individuals who had worked with Dr. Hauser. By August 2007, they felt that they had come up with enough evidence of misconduct to justify raiding Dr. Hauser's laboratory to look for further evidence.

This committee would go on to quietly investigate Dr. Hauser for the next 3 years.

Aftermath

Harvard University concluded its investigation in 2010, finding Dr. Hauser guilty of eight counts of research misconduct, three of which were in published studies. As a result, Harvard researchers issued corrections to two of Dr. Hauser's studies and retracted a major study entitled "Rule learning by cotton-top tamarins" published in the 2002 edition of Cognition. In this study, Dr. Hauser fabricated a graph that overestimated the ability of tamarins to distinguish rules in sound patterns. The retracted study had been cited 86 times by other works. After a year of academic leave following the investigation, Dr. Hauser resigned his professorship at Harvard University in 2011, acknowledging "mistakes" in his research conduct but never explicitly accepting blame.

In 2012, the U.S. Office of Research Integrity (ORI) published a follow-up investigation that corroborated Harvard University's findings. The ORI prohibited Dr. Hauser from engaging in research funded by the U.S. Public Health Service (PHS) for three years without close supervision from an institution, as well as barring Dr. Hauser from serving as a peer reviewer for any future publications concerning PHS-funded research. Dr. Hauser voluntarily accepted the punishments.

The full extent of how Dr. Hauser's misconduct affects other works in the field of animal cognition remains unclear, and "hundreds" of articles cite or build upon concepts or methods established in his earlier work. Harvard Psychology Department Chair Susan E. Carey noted that a full review of the decades of research conducted by Dr. Hauser could "unfold over years." Other anonymous professors in the department expressed concern that Dr. Hauser's misdeeds created undue "uncertainty" for animal cognition researchers and harmed overall trust in the lab. Princeton Psychology Professor Charles Gross writes in The Nation that "irreversible damage" has been done to the field of animal cognition. However, as of 2017 the Cognition paper remains the only one of Dr. Hauser's works that has been formally retracted.

Criticism Against Harvard

Some of Dr. Hauser's colleagues believed that Harvard University acted unprofessionally during its investigation. Robert Seyfarth, an evolutionary psychologist at the University of Pennsylvania in Philadelphia and one of Dr. Hauser's former advisors, criticized Harvard for the lack of transparency during the investigation, pointing out that it "simply fuels rumors." In contrast, other academics at Harvard praised the university for maintaining professional confidentiality of the proceedings. The findings eventually were made public in the ORI report, though it was released two years after the original investigation concluded.

Several prominent academics, including Dr. Hauser's former mentor and MIT Linguistics Professor Noam Chomsky, have expressed displeasure in a letter to Harvard with the scope of the university's investigation procedures. The academics complained that Harvard's seizure of materials from Dr. Hauser's lab was excessive, including personal emails and financial documents. In particular, they decried the inclusion of the unpublished studies in the investigation as a "violation of free scientific inquiry," contending no harm was done by recognizing errors and choosing not to publish tainted works. Psychologist Bennett Galef of McMaster University concurred, pointing out that punishing a researcher for an unpublished work is nonsensical. Other scientists in the field disagree: Cognition editor Gerry Altmann notes that even if the paper retracted from his journal constituted the only infraction, that alone would be considered "unacceptable and reprehensible behavior." Jenny Saffran, a co-author in one of the papers implicated by the ORI, points out in the same report that it was "not the case that [Hauser] corrected all errors before submitting papers for publication."

Motivations and Consequences of Research Fraud

"Publish or Perish"

There is a push in academia for consistent publication in journals. This is not just for field advancement, most institutions use this as a criteria for earning tenure promotions. This culture has put pressure on professors/academics to regularly be doing research worthy of publishing. Although Dr. Hauser did not seem to fear "perishing" due to his high publication record, his tendency to jump to conclusions noted by his co-worker demonstrates a possible over-eagerness to publish.

Publication Bias

Publication bias is a phenomenon that suggests academic journals are more inclined to publish articles that return "positive" results. Experiments that reach an inconclusive or null hypothesis have less chance of getting published. Because positive results are more likely to be published, researchers may try to skew data in order to secure competitive opportunities for funding and career advancement. Examination of Dr. Hauser's misconduct suggests he could have desired to conform to the publication bias of top journals by intentionally fabricating data to indicate positive correlations that were actually negative.

Pride

Many researchers take pride in having their works published by academic journals or cited in other papers. Furthermore, they take pride in postulating accurate hypotheses, so there is an incentive to falsify data. The context of Dr. Hauser's story and accounts from his colleagues point to the notion that his personal pride played a role in the scandal. In particular, his high publication record and motivation to keep publishing suggested that he took pride in his publication count. The Harvard Report implicates the idea that Dr. Hauser falsified data to make it match his preconceived hypothesis.

Who Is Affected?

Research fraud affects a number of people and institutions. Even after a researcher has served their sentence for research fraud, it is much more difficult to get any sort of funding for new experiments. A significant number of people who get caught committing fraud never publish another paper. The source of the grant money is upset, and often times the image of the institution that sponsors the researcher is marred. Harvard received much backlash for Dr. Hauser's actions and eventually parted ways with him.

Postdocs and graduate students are often some of the biggest victims of research fraud. They are commonly used as scapegoats and blamed for the incident. Dr. Hauser tried to blame his graduate students multiple times for the inconsistent data, and at one point even claimed they were plotting against him. Postdocs and grad students can have their career paths uprooted if they are working for a professor who is committing fraud, and they are generally the ones who have sunk time and effort into follow-up research when it comes out that there has been fraud.

Related Cases

Research fraud is a serious problem that damages the credibility of academic institutions and scientific fields. Daniele Fanelli of the University of Edinburgh, in a 2009 study examining research practices by other scientists, found that 2% of sampled researchers admitted under anonymity that that they had "fabricated, falsified or modified data or results at least once – a serious form of misconduct by any standard - and up to 33.7% admitted other questionable research practices." While Marc Hauser's punishment is mostly typical of wrongdoers, in some cases research fraud invites punishment beyond academic and professional repercussions.

Eric Poehlman

Eric Poehlman is a former professor of the University of Vermont whose research focused on the physiology of obesity and aging. He had earned fame for showing that some metabolic changes followed from menopause instead of aging; however, the menopause study and much of his other research conducted over a decade long career was exposed as fraudulent in a 2005 investigation that began after Walter DeNino, one of his top lab technicians, caught him manipulating data. Poehlman had used the fraudlent research to secure millions of dollars in grant funding from the National Institute of Health. After a protracted battle in which Poehlman fought to discredit DeNino and other whistleblowers - even accusing them of homophobia - he pleaded guilty to misconduct in ten published papers and falsifying data in federal grant applications. In addition to losing his job and being barred permanently from receiving federal research grants, Poehlman was sentenced to one year in federal prison. The ten papers for which he admitted guilt were either retracted or corrected.

Dong-Pyou Han

Dong-Pyou Han, formerly a celebrated Iowa State University scientist, resigned from his position in disgrace in 2013 after confessing that he faked results that suggested a non-existent breakthrough in AIDS vaccine research. Han had spiked rabbit blood samples with human antibodies to give the illusion that the vaccine was working, and his breakthrough secured millions of dollars in grant funding for his research lab. Upon his resignation, Iowa State University had to repay $496,000 worth of federal grant money. According to Han, the misconduct began as an accidental contamination, but due to catching the mistake late he continued to contaminate samples instead of admitting fault. In 2015, Han pleaded guilty to two counts of making false statements and faking data used in federal grant applications. He was sentenced to four-and-a-half years of prison and fined $7.2 million. According to The Washington Post, a punishment of this magnitude for research fraud is "rare".

Future Work

Research fraud is perhaps the greatest unprofessional sin in the academic world, with the potential to ruin careers, damage the credibility of academic institutions, and even throw an entire field of study into question. That said, more research is necessary to determine the full extent of its detrimental effects. Future work could focus on how incidents of fraud perpetuated by researchers as prolific as Dr. Hauser could influence public mistrust in science. Parallels can be drawn to the anti-vaccination movement, where the link between vaccines and autism is still accepted on faith despite the backing study being proven fraudulent and retracted. Examining similar occurrences in other disciplines could be a useful area of future study.