Two journals have dismissed allegations of research misconduct leveled against a political scientist at Harvard in an anonymous memo that labeled his work “pseudoscience.”
The 2018 memo signed by “Social Scientists for Research Integrity” – which does not have an internet presence that we could find – makes claims of academic misconduct against Ryan Enos, who denies any wrongdoing. The journals that published two of Enos’ papers singled out in the memo decided to let the articles stand after investigating the charges. A committee at Harvard University, where Enos is a professor of government and director of the Center for American Political Studies, also reviewed the claims and dismissed them.
The allegations primarily concerned purported manipulation of data in Enos’ 2015 article, “What the Demolition of Public Housing Teaches Us about the Impact of Racial Threat on Political Behavior,” published in the American Journal of Political Science (AJPS). The paper has been cited 120 times, according to Clarivate’s Web of Science.
The article explores the theory of racial threat in the context of public housing demolitions in Chicago that displaced more than 25,000 African Americans between 2000 and 2004. He found that before the demolitions, white voters in neighboring communities had turned out in higher numbers and voted more conservatively. The observed effect was related to their proximity to the public housing and the size of the African American population that lived in the developments.
The anonymous memo alleged the data had been manipulated, and claimed, “due to these distortions, every single analysis reported in the article… was impacted.”
Enos told Retraction Watch the allegations are untrue and were made in bad faith. “I’m perfectly happy to talk to anybody about it,” he said. “That’s my obligation as a researcher is to answer questions that anybody has about my research and to be fully transparent about it.”
The allegations were brought to Harvard’s chief research compliance officer in August 2018, according to correspondence shared by Rachael Dane, a Harvard spokesperson. The chair of the Committee on Professional Conduct, Peter Marsden, responded in October 2018, dismissing the case because the allegations “do not fall within the purview of the Standing Committee for Professional Conduct” according to Harvard’s research misconduct policy and procedures.
In an email to Retraction Watch, Dane referred to reports of the allegations on several Substacks, adding:
This has been reported in the media, and we do not consider any of this newsworthy. Further, we have nothing more to say and consider the matter closed, as we have, for some time.
She declined to share the identity of the author of the memo, citing limitations under The Family Educational Rights and Privacy Act (FERPA) – a law that protects the privacy of student education records.
According to Enos, the memo was written by somebody with a personal grievance who is a former student at Harvard.He said the same person brought the allegations before the Harvard Committee on Professional Conduct.
We asked Claudine Gay, Dean of Harvard’s Faculty of Arts and Sciences who will become the university’s president on July 1, about what steps had been taken to address the allegations but received no response.
The same memo appears to have resurfaced four years later after being sent to Chris Brunet, who wrote about it on his blog in March 2022, as did Jesse Singal several days later.
The editors at AJPS said they were first made aware of the allegations in March 2022, but received an “actual complaint” from a member of the political science community in June 2022. The complainant alleged Enos’ data was inconsistent with official election results.
The complaint triggered a formal investigation which found “no credible evidence to support the allegation of academic fraud”, co-editors-in-chief Jennifer Lawless and Kathleen Dolan told Retraction Watch in an email.
According to emails between the journal and Enos seen by Retraction Watch, the editors deemed his explanation of the data discrepancy “sufficient to address the complaint.” However, they requested that he draft a memo “addressing the potential implications of the slippage between your data set and the official voter file,” citing “the amount of attention and traction this discrepancy has received.”
A four page memo was eventually published on the journal’s dataverse, responding to the criticism that Enos’ replication files include fewer registered voters and a lower voter turnout than the official Chicago records. Enos wrote that the data contains measurement error.
In a call with Retraction Watch, Enos added he has no reason to believe the measurement error would lead to bias in the article’s inferences.
“It’s absolutely 100% not evidence of fraud and I don’t see how anybody could conclude that,” he said. “Anybody that’s familiar with the data would know there’s going to be discrepancies between voter files and official results.”
Enos later published another memo on his website responding to the criticism that the number of voter precincts don’t match official records. In short, he explained some of the precincts changed over time, rendering them incomparable in his analysis.
“To make that apples to apples comparison, to compare the same place at different points in time, I had to only include the precincts that hadn’t changed,” he said. “A number of precincts are eliminated because of that.”
Other allegations that weren’t explicitly addressed by Enos in his addenda were regarding the purported non-compliance with the AJPS replication policy, the inclusion of people below the age of 17 in the data, and differences with older versions of the paper.
“The replication files that were made available were ones that matched to the replication policy at the time,” Enos said. The current AJPS Replication and Verification Policy was announced in March 2015, according to the AJPS website – a month after the journal published Enos’ paper.
According to Enos, the inclusion of registered voters below the age of 17 boils down to “a lot of administrative error in these data sets.”
And the differences between the published paper and older iterations of the work were largely because how Enos included people in the dataset changed. He broadened the dataset from January 2000 to October 2000, to include up until the last day that voters could register. “Otherwise I’d be throwing people out of the dataset that could have voted in that election,Enos said.
“Contrary to what it says in the memo, that didn’t result in any great changes in the paper.”
The anonymous memo singled out another one of Enos’ papers, “Causal effect of intergroup contact on exclusionary attitudes,” which was published in 2014 by the Proceedings of the National Academy of Sciences (PNAS) and has been cited 146 times, according to Clarivate’s Web of Science. The paper examined how the views on immigration of mostly suburban whites changed after coming into contact with Spanish speakers on a train.
The anonymous memo republished an online comment that criticized the generalizability of Enos’ results and pointed out inconsistencies in the wording of a survey question.
As we reported in 2018, Enos attempted to set the record straight with the journal after becoming aware of the “mistakenly recorded” survey question in the paper. It eventually resulted in a correction.
Since then, Chris Brunet has alleged on his blog that the whole paper “stinks” because of irregular IP addresses, the presence of infants in the dataset and the removal of the IP addresses from the replication files.
The research was conducted near Boston but some of the IP addresses are from outside Massachusetts. Enos pointed to the inaccuracy of IP addresses to explain the inconsistencies, and also addressed why he removed the IP addresses from the dataset in this memo published on his website.
Responding to the point that there appear to be infants in the dataset, Enos said the age categories in Qualtrics are recoded into an ordinal variable, meaning a one in the dataset would represent an 18.
“If I was committing research fraud, why on earth would I have toddlers in the data?” Enos said. “The idea that anybody would be stupid enough to fake data by putting toddlers in, it is absolutely ridiculous.”
When asked about the allegations, a spokesperson for PNAS wrote in an email:
In March 2022, PNAS was notified of concerns regarding this article. After conducting an internal review, the journal concluded that no further action was required. We consider this matter resolved unless new information relating to the PNAS article is brought to our notice.
Like Retraction Watch? You can make a tax-deductible contribution to support our work, follow us on Twitter, like us on Facebook, add us to your RSS reader, or subscribe to our daily digest. If you find a retraction that’s not in our database, you can let us know here. For comments or feedback, email us at [email protected].
“anonymous memo” . . . “Social Scientists for Research Integrity”
If they had integrity, they wouldn’t be anonymous.
Also, FERPA would not protect this memo. Even if the author were a CURRENT student (which the article says he or she is not), it wouldn’t be an education record – administrative complaints are not education records.
Can someone clarify why a person would make an issue of this? Obviously the issue is not research integrity – because all that seems to be satisfactorily explained – but behind the campaign must be an issue of political importance that is gnawing at the complainant and that they want to see set right. What is it? I guess it is something to do with race and racial attitudes, but I could not quickly – reading on the fly, and I am afraid being a bit lazy – work out what it really was! I am very familiar with social science, public policy, and attempting to make strong inferences, so I am familiar with the methods and the disciplines, but not with the issue (yet!).
I read Singal’s post and I actually couldn’t figure it out either. The topic of race seems to have set people off of some persuasion of another, but naturally this is a topic of interest to academics and policy makers, and as with any other complex scientific topic, not all of the studies about it will be well designed.
The “personal grievance” is likely not as much with the work as it is with how Enos treated Harvard graduate students during his time as DGS
The “personal grievance” is likely not as much with the work as it is with how Enos treated Harvard graduate students during his time as DGS
The anonymous innuendo is dumb and seems to be the standard method of whoever has been pushing this farce for over a year. Enos is a respected member of Harvard and the research community, except according to internet trolls.
“We asked Claudine Gay, Dean of Harvard’s Faculty of Arts and Sciences who will become the university’s president on July 1, about what steps had been taken to address the allegations but received no response. ”
Some questions for RW to consider:
1. Why would Gay lie that allegations of misconduct do not fall within the purview of the administrative entity in charge with investigating misconduct?
2. Why didn’t she at least ask for a statement from Enos, as AJPS did?
3. Why has Enos still not provided an explanation for the fact that around 1,000 precincts have turnout or Republican vote values different from official data, or have been entirely deleted? He can’t blame this on the quality of administrative data, since these are publicly available, and different from what he has in his replication archive. This is not addressed in his October 2022 memo.
4. Why was Enos not asked for the code he used to make corrections across the paper versions? No matter how bad the original data were, he made changes that resulted in effects becoming statistically significant out of the blue, or increasing in magnitude. All of these changes are in the direction of Enos’ pet theories. Coincidence?
The ‘honest error’ defense is an affirmative defense, meaning Enos has the burden of proof here. So far Enos has failed to meet that burden of proof.
Are you the person who originally wrote the dishonest hit job? Because you sure sound like it.
“And the differences between the published paper and older iterations of the work were largely because how Enos included people in the dataset changed. He broadened the dataset from January 2000 to October 2000, to include up until the last day that voters could register. “Otherwise I’d be throwing people out of the dataset that could have voted in that election,Enos said. “
The problem that Enos failed to address here is that he changed *more* than the time interval for his analyses. On p. 20 of the whistleblower report it is shown that on the dataset up to January 2000 (which is a subset of his enlarged dataset up to October 2000) effect sizes changed and number of voters within distance categories changed between the 2012 and the 2016 iterations of the paper. The left and middle panels in Fig. 7 should have been identical when running his code on the subset up to January 2000, but they are not. For example, in the 100 m distance category the effect size changed from -.05 to almost -.15 (that is, from 5% to almost 15% decrease in turnout), so the magnitude nearly tripled (in favor of his theory). The sample size on which the estimate is based also changed. He stated that he only deleted observations which he couldn’t geocode. If in 2012 he only had 16 voters in that category, as this is how many he was able to geocode, why does he have 20 voters in the same category in 2016 on what is supposed to be the same data subset? Where did the additional 4 voters come for? This is not the only example; in nearly all categories both sample sizes and effect estimates were changed, for no apparent reason.
This indicates that he performed other (undocumented) data transformations, in addition to the change of registration cutoff point from January to October 2000. And those changes cannot be attributable to the quality of the administrative data, as they were made *after* he acquired the data. These are changes made by Enos himself that he has yet to provide a justification for.
“Whistle-blower report”. That was not a whistle-blower report. It appears to be a poorly done hit job.
Speaking as someone who is very involved with local politics here in Illinois, this quote caught my interest:
“there’s going to be discrepancies between voter files and official results.”
What exactly are these discrepancies?
And this quote:
“the inclusion of registered voters below the age of 17 boils down to “a lot of administrative error in these data sets.”
Again, what is this “administrative error”? A registered voter below 17 is committing fraud, or someone is committing fraud on their behalf. The Attorney General needs to be informed if this is indeed widespread, which in my experience isn’t.
This state does change boundaries for local, statewide, and federal elections frequently, which does make it difficult to make comparisons or to extrapolate data from one election to the next. Anyone trying to make grand statements from election data in this area needs to be very careful they are seeing exactly what they see in the data.
If you check out the memos linked to the article, you’ll see that he goes into this in great detail.
Did anyone else attempt to read these memos linked to the article? I did, and they are incomprehensible word salads to me. He notes a widely-known problem with voter roles – that if/when people move away or die right before an election, the voter roles are inaccurate, and thus we will never have completely accurate voter roles. Fair enough. What follows are a dozen pages of transformations to his data. After all this flogging of the data I don’t know you can make any judgements on what you are seeing nor how you can even recall what the original data was. I know this is a non-technical thought on this, and the comment a few spots above does a much better job getting into the specific issues, but – wow – how much transformation of data can you do to fix a simple (although hard to solve) problem?
And I don’t see any explanation for his statement about below-17-year-old voters in the data. That simply doesn’t exist in the real world, even in Chicago which has faced allegations of vote fraud for a century. In which memo is this explained?
Sorry, but this simply isn’t true. The memos don’t discuss any transformations to fix inaccuracies in the voter roles because there were no transformations to do so. The inaccuracies are treated as measurement error, which is standard in literature. I’m not sure why you would claim these memos said something that they did not.
Honestly, this anonymous memo is, unto itself, a remarkable example of misconduct. Read it carefully and read Enos’ response to it. It’s no wonder it was anonymous. No serious social science researcher would put their name behind it.
The memo takes very common errors in administrative data, such as that the data contains people with errors in their recorded age, and claims that these errors are examples of research fraud. If this were research fraud, then nearly every paper that uses voter data would also be guilty of this fraud.
Then, even more remarkable, as noted in Enos’ response, it appears that the anonymous author of the memo didn’t understand that precinct boundaries change with time so they downloaded updated precinct GIS files, merged them with Enos’ data from a previous time period, and then created a map with the incorrectly merged data. They then used this incorrect map to accuse Enos of fraud. We will never know if this error by the anonymous author was intentional or not.
This is an excellent example of why science shouldn’t be anonymous.
Please read the memos carefully. Enos is working backwards from 2010 to 2004 precincts boundaries, deleting anything that had even a slight boundary change. Brute-force deletion of data (rather than using some form of imputation) is viewed as a highly questionable practice in academia. The data in the whistleblower report are merged correctly (merging 2010 GIS information to 2010 ward-precinct IDs).
The report shows that Enos deleted ~850 precincts (one third of the total), many of which are in close proximity to the housing demolition projects (that is, voters on which he claims to make inferences, his “treated” group). This introduces severe bias into his results, as deleted precincts are unfavorable to his preferred theories.
Enos did not disclose in the article that he had performed those deletions (much less explain the reasons and how it may have impacted his results), and he had not uploaded the relevant code for precinct deletions. This would have been enough to raise suspicions of fraud (there are many other irregularities as well).
In his most recent 14-page memo posted at his website (why not at the official AJPS dataverse?) he defends the deletions as part of the backwards boundary matching process, and has now posted the SQL code for what he did. Failure to disclose the deletion of a substantial portion of the data and provide adequate documentation for the deletions constitutes falsification according to the Harvard policies — a form of research misconduct — on Enos’ part. The Harvard definition of falsification is “manipulating research materials, equipment, or processes, or changing or omitting data or results such that the research is not accurately represented in the research record.”
It remains to be seen whether the datasets he uploaded to the dataverse in 2014 can be reconstructed using Enos’ newly posted SQL code, whether he reverse-engineered the deletions etc. This is all new information that Enos should have provided from the outset, in 2014, and journal editors should now do their job to run his code and check.
I do not understand the memo. That is OK, this is not my area of expertise. However, I am confident that Harvard, under whose aegis the research was conducted, assembled a group with appropriate expertise to evaluate the allegations in the memo. That is how responsible Universities handle such allegations. Having launched at least a dozen such investigations during my ten years as Editor-in-Chief of Anesthesia & Analgesia, I know how this process unfolds. I would advise the correspondents commenting on the merits of the memo to defer to Harvard’s determination.
As many have commented, unsigned memos are themselves a red flag that the author has an ulterior motive. When I arrived at Columbia University, I soon discovered anonymous allegations of academic misconduct being leveled at me. It took three years to discover the author: another member of the Department of Anesthesiology. The allegations started after I rejected one of his manuscripts. Over several years his allegations of my misconduct became increasingly bizarre. Just as the memo connects multiple disparate dots into a story of misconduct, the individual alleging my misconduct connected unrelated points (e.g.,
) into a tangled web that “proved” I was an “academic buccaneer” (his words).
One of the letters alleging my misconduct was sent to Adam Marcus, then editor of Anesthesiology News, and co-founder of Retraction Watch. After the author made a tangential reference to physical harm, I gave Adam my Google Drive password in case I was shot. (Adam, I’ve since changed it).
When I left Columbia and returned to Stanford the author turned his malice on another member of the Columbia faculty. I was glad to leave that chapter behind.
The experience taught me that those with a deeply seated grudge have belief structures that are not amenable to data. They will not approach the subject with equipoise, will not accept that there may be innocent explanations for their observations, will reject facts that dispute their conclusions, and will interpret institutional exoneration as a cover-up.
That would seem to be the case here.
@Steven: Harvard has a long and well-documented history of corruption and administrative malfeasance. You say you don’t understand the memos but trust that Harvard administrators must have done the right thing. I would point you to the Dominguez scandal, where the administrators dismissed 18 valid formal complaints and allowed the perpetrator to become department chair and Vice Provost so he could retaliate against complainants (years later they found him guilty, apologized to the victims, and stripped him of emeritus status). They also accepted donations from and gave Epstein an office on the Harvard campus even after his criminal conviction. To say you blindly trust Harvard administrators is, to be polite, a naive approach. They have repeatedly proved that their judgment cannot be trusted.
If you read this particular story carefully (and this is something you can understand regardless of your field, even if you can’t evaluate the memos yourself), the complainant(s) were NOT anonymous, as they disclosed their true identity to Harvard officials and thus took full responsibility for the findings. This is why Harvard officials know that a federal statute (FERPA) protects their rights. Since Harvard (and universities in general) also has a non-retaliation policy and a legal obligation to protect whistleblowers, they were presumably advised to choose a pseudonymous name so they would be protected from retaliatory action from Enos.
Retraction Watch contacted Claudine Gay, the administrator who stated she had “readily dismissed” the whistleblower report, to ask her which steps she had taken to verify the allegations, and she did not respond. The official reason given was that the allegations “did not fall within the purview” of the Committee for Professional Conduct (CPC), even though Harvard policies clearly state that this is the entity formally tasked with investigating allegations of misconduct. So she dismissed it with a blatant lie about the role of the CPC, and did not forward it to any other unit to investigate.
Enos acknowledges that all the irregularities do indeed exist. He justifies them as “a lot of administrative error”, personal anecdotes etc. It took 14 pages and an additional data folder containing 26 items to explain the irregularities, and there are still unanswered questions, which means it wasn’t something that could have been easily dismissed. In fact, as long as the allegations met the definition of research misconduct and there was sufficient information to verify them (the whistleblower report includes software code to document each allegation), Gay would have been obligated to ask for a statement in 2018. Enos did not produce a statement until a political science professor formally complained to the journal in June 2022. It’s worth emphasizing that Claudine Gay works in the same niche subfield as Enos; in her dissertation, she defends the “racial threat theory”, the exact same theory Enos claims to have found support for.
Now Enos brazenly brags that he was told who wrote the report and who submitted the complaint to the CPC to investigate (not realizing he is publicly throwing the corrupt administrator under the bus), and launches ad hominem against the complainant(s). Which means that Harvard committed violations of federal and university policies, in particular FERPA and whistleblower protection / non-retaliation policy. This is egregious.
I would advise students who identify irregularities in faculty’s work to remain completely anonymous and not trust university administrators to follow the correct procedures.
Well, I’d say that reply demonstrates Steven Shafer’s point rather perfectly.
OK, maybe “transformation” isn’t the right word, but it is evident from these memos that he did change data, evidently by deleting data (as noted in a response below).
To clarify, Illinois has another issue with voter data: it changes boundaries so often that it is difficult to impossible to compare election results from one year to the other, with the exception of citywide or countywide elections where the boundaries don’t change. (this issue also disenfranchises voters, who have no idea who their elected representatives are, but that is a topic for another day). From these memos, it appears this is the problem the author is trying to solve. Again, after reading these (rather dense and repetitive) memos, I can’t understand what was done with the data. I don’t allege fraud – I think the more you transform (or whatever word you want to use) the data the more real it appears, but I would be very careful of drawing conclusions from it.
I would love to see Harvard researchers look into the very real issue of the constant re-drawing of state, local, and federal boundaries in states such as Illinois and the effects that has on disenfranchising voters than chasing a problem that may or may not exist.
And – where is the explanation of under-17-year-olds showing up in voter data?
In response to “John A”, I note that you are hiding behind anonymity. I do not respect comments from people who are unwilling to state their names. By contrast, I’m easy to identify from my post: Steve Shafer, Professor Emeritus, Stanford University. I can be reached at [email protected].
Your anonymously stated allegations of misconduct by Harvard raise several questions:
1. are you the author of the original complaint against Professor Enos?
2. Are you the author of numerous anonymous posts on this thread alleging misconduct by Harvard and Dr. Enos?
3. Are you trying to weaponize Retraction Watch against Harvard and Dr. Enos?
4. Have you made these allegations elsewhere?
5. What is your relationship to both Harvard and Dr. Enos?
6. What is your agenda in posting your allegations?
All 6 questions would likely be answered if you remove the veil of anonymity. You can demonstrate your good intentions by giving your name and email address, just as I have done. If you refuse to reply, then I would encourage anyone perusing this thread to ignore your comments.
Be well,
Steve Shafer
Steven:
You tend to overdiagnose one potential problem simply because you went through the traumatizing experience of being falsely accused, and you have our fully sympathy for what happened. I’m glad you were exonerated. But please recognize that the opposite can happen as well: there are scholars who are rightfully accused of misconduct, and journals and universities do engage in cover-ups and retaliation. Retraction Watch provides ample testimony to that.
You can choose to believe whatever you want, but your negative personal experience gives you absolutely no right to ask people to ignore other posters’ comments and unconditionally accept your point of view, for any reason. Some people may have friends, collaborators, or family members at Harvard, and may fear they could become targeted and retaliated against by the same administrators who refused to investigate. Everyone is entitled to read and agree or disagree with whichever comment they choose. There are links to memos, data, and code, and anyone who understands data analysis can verify the facts for themselves and make their own determination.
Nobody asked people to ignore your comments because, as you admit yourself, you have no expertise whatsoever in the field, and you do not even understand the memos. If you don’t understand the issues being discussed, why do you jump to conclusions about who is right and who is wrong? Please address the allegations on their merits, instead of offering your personal uninformed speculations. We deal in ideas and arguments, not in personalities. As a journal editor who handles comments provided though an anonymous peer review system you should know this. You should also know that the Committee on Publication Ethics (COPE) recognizes that “there are many legitimate reasons for individuals to wish to remain anonymous.” For example, the complainants in the Eric Stewart case (Florida) were also anonymous, and the journals did the right thing and retracted 6 of his articles.
The first public reference to Enos’s work as ‘pseudoscience’ was made in a comment on Andrew Gelman’s statmodelling blog. The poster essentially accused Enos of p-hacking (doing one-tailed hypothesis test without correcting for multiple group comparisons), pointed out the inconsistent question wording (ask yourself under what conditions a researcher could accidentally type a question wording that doesn’t even exist in the survey they fielded?), and replied to Enos that “both you [Enos] and [your advisee] make conclusions unwarranted by the data”, and that his data “do not “support” or “show” or “prove”” his conclusions. The Gelman post is from May 29, 2018. Please refer to the comments section and read the full-length comment.
Furthermore, the political science professor from Boise State who lodged the official complaint with AJPS in June 2022 concluded, after Enos’ initial response to the inquiry, that “I’m definitely a lot more on board with ‘this is fraud’ rather than ‘this is incompetence’ after his response.”” He told investigative journalist Chris Brunet that he could quote him on record. You can find the quote at Brunet’s substack.
As you see, several people have brought allegations of misconduct against Enos over the years. We will see who turns out to be right once the new memo, data, and code are scrutinized. Until then, people are free to exercise their right to flag inconsistencies in the author’s response and ask questions that remained unanswered. Since you have no expertise and no understanding of the statistical irregularities investigated, you are in no position of being the arbitrator of whose comments should be accepted and whose should be ignored.
“Shawn”, I do have the expertise to evaluate this and I’d be happy to discuss the merits of your accusations. But here’s the problem: I work in this field and have watched this sorry spectacle unfold for over a year and seen you are unwilling to accept the responses of experts who disagree with you.
Look at what you said in your response to Steven, where you conveniently ignored or dismissed the experts who evaluated your complaints against Enos. As the Retraction Watch article indicates, when you brought the complaint to Harvard, Peter Marsden, a world-renowned sociologist and expert in statistical methods, reviewed your claim and dismissed them. Then you brought them to experts at two journals and both of them dismissed your complaints. Above, you fail to mention that the complaints from the political scientist at Boise State were also reviewed and dismissed by the journal. Then you fail to mention that the “investigative journalist” Chris Brunet was thoroughly debunked by Jesse Singal (in the post linked in the Retraction Watch article). This is notable because Singal has exposed fraudulent political science before and in his investigation of Enos he interviewed several experts in the field, including Andrew Gelman, all of whom said they found no cause for concern in Enos’ research.
I could ask you what you make of all these experts who have repeatedly told you that you were wrong, but I am guessing it won’t matter.
Don:
I never sent anything to Peter Marsden in my entire life, and I don’t believe I had even heard of him before. I don’t understand why you make such claims. Please note that Brunet reports that he summarily dismissed the allegations with the claim that they did not fall within the purview of that administrative entity. He never investigated them. He never engaged with the merit of the allegations. Enos pleaded ‘no contest’ to the allegations that there are discrepancies in the turnout numbers and precinct deletions. Not mentioning those issues in the article seems at best like a glaring omission on Enos’ part, and he should have been asked to publish an addendum in 2018.
Singal lost all credibility as a journalist after the whitewash of Alice Goffman and doesn’t have the required skills to produce an informed opinion.
To assess the irregularities, an expert with strong quantitative skills is needed. And even experts can be wrong sometimes. Science experts published Lacour’s article. Political science “experts” offered him a Princeton job.
My point in response to Steven was that everyone is entitled to analyze the memos, data, and code, and arrive at their own conclusions. His point of view is purely informed by his emotional response, not by rational arguments. He is still entitled to his opinion, of course, but he shouldn’t police this forum and tell others what to believe when he doesn’t even understand the data irregularities involved. We don’t have to reach a consensus. We can all agree to disagree.
I’m more of a number cruncher but I hope as many people as possible look into that Data Update folder so we can all reach the correct conclusion eventually. In any case, I wish Enos best of luck in the future. If he’s innocent, I hope those files exonerate him. If he did commit fraud, I hope he will have the dignity to apologize and resign.
Let’s hope the truth comes to light, whatever it may be.
This is an interesting dialog. Four of us have posted our names: Peter Davis, Christopher Jacks, David J. Edmond, and me. Our comments are fairly straightforward.
Most commentators have hidden behind a veil of anonymity: John A, Shwwn, Don, A, C, D, ld, A S, Corral, J.L.S., Tom K. (It is fun bit of irony, Tom K anonymously posted “If they had integrity, they wouldn’t be anonymous.”)
I think it would help the discussion if individuals stopped posting anonymously. I suggested that John A indicate his real name. I make the same suggestion to Shawn and Don. Who are you? Do you have real expertise in this field, or do you have another reason for closely following this arcane thread?
My interest here is not so much the merits of the accusations as seeing whether Retraction Watch is being weaponized by anonymous correspondents intent on settling personal grievances. As a member of the Retraction Watch Board of Directors, this is not just an academic curiousity but a question that merits consideration by the Board.
Be well,
Steve Shafer
[email protected]
Steven Shafer, your tale of woe about your experiences at Columbia has nothing to do with Enos’ alleged research misconduct.
The poster John A has provided details about why he thinks Enos’ paper is the result of data manipulation. The paper and the Dataverse memos are publicly available. Refute John A’s points, if you can. No one cares about him being anonymous and you not being anonymous. This is about research integrity, not about John A’s or your personal identities.
Harvard is a prestigious and wealthy university, but prestige and wealth are not necessarily consummate with integrity.