Retraction Watch

Tracking retractions as a window into the scientific process

Chemist sues University of Texas (again) to keep PhD

with 21 comments

Screen Shot 2016-02-08 at 4.34.25 PMA chemist is suing the University of Texas a second time in an effort to keep the PhD she earned in 2008.

In 2014, school officials revoked Suvi Orr‘s degree after finding it was based, in part, on falsified data. Some of the data were also included in a paper in Organic Letters that was retracted in 2011 after some steps in the chemical synthesis the authors described were not reproducible. Orr, currently working at Pfizer, sued UT, and the school reinstated her degree.

Now, the school is trying to remove it again, according to the lawsuit, filed last week. The lawsuit says the school has scheduled a “hearing” on March 4, during which three undergraduate students and two faculty members will deliberate — “none of whom are qualified to evaluate the scientific evidence being used against S.O.,” the suit says.

Orr has requested a temporary injunction to halt the proceedings, and a hearing has been scheduled for next week, according to the Austin-American Statesman.

The suit argues the school does not have the right to strip Orr’s degree from her:

…S.O. has a constitutionally protected property and liberty interest in her PhD. The University should not be permitted to so cavalierly attempt to take it away. S.O’s scientific career—and her livelihood—hinge on how a panel of undergraduate students review and interpret highly complex scientific data at the doctoral and postdoctoral level in organic chemistry…The Court must grant the requested relief so that S.O. is not stripped of her constitutionally protected property and liberty interests without the due course of law, due process and equal protection of the law.

Orr’s lawyer, David Sergi, told us,

Our focus is on accountability and transparency. We think there is a lack of accountability for the dissertation advisers.

The lawsuit (here’s a PDF) alleges that UT is using Orr as a “sacrificial lamb” for the mistakes of the retracted paper to protect her former adviser, UT chemist Stephen F. Martin:

The University has made it abundantly clear that it is willing to sacrifice the good name, reputation, and integrity of S.O. for the sake of Prof. Martin. Rather than have Prof. Martin admit his own errors and shortcomings as a graduate advisor, research chemist, and author, the University has embarked on years of secret investigations as it attempts to revoke S.O.’s PhD for what may be a subjective error in scientific judgment.

The lawsuit provides background on the retraction:

A paper was published in which Prof. Martin was the leading author, and the post-doc and S.O. were coauthors of the journal article. S.O. intended and expected that the post-doc would fully characterize the compounds as part of reproducing her work. Unbeknownst to S.O., the post-doc instead made markings, scanned, and submitted experimental data containing S.O.’s file names and some incorrect (and inaccurate) pages as part of the journal article. On relevant issues related to Compounds P, Q, and R, he reached the same or similar conclusions as S.O. that are discussed in his post-doctoral final report. Nonetheless, the article was ultimately retracted, and the decision to retract was apparently made by Prof. Martin without input from his co-authors (S.O. and the post-doc).

The suit alleges that Martin steered Orr wrong:

A test that can be used to fully confirm–not just propose—a molecule’s identity is a small molecule X-ray. It is considered the most reliable of the testing methods. Prof. Martin, however, considered the H-NMR data (and the other three tests) to be sufficient to identify molecules. Some other graduate advisors required an X-ray testing for complex molecules because H-NMR data and interpretation has more room for errors and can be inconclusive. Under Prof. Martin’s protocol, S.O. was required to identify molecules by the four tests described above. Prof. Martin recommended X-ray testing if the molecule could be crystallized but he gave little guidance on how to crystallize molecules.


After analysis and discussion, Prof. Martin agreed with S.O.’s conclusions as presented in her dissertation. Without his endorsement, she would not have been allowed to appear before the dissertation committee.

It argues that the problems with the paper were all “part of a scientific process:”

At worst, S.O.’s alleged misinterpretations demonstrate the inexperience of a graduate student who was still learning her craft. The alleged misinterpretations are nothing but part of a scientific process. In scientific process, there are only theories, not facts. A theory can of course be proven wrong. In this case, it is still up to debate what the actual identities of Compounds P, Q, and R are.

A spokesperson for UT told the Austin-American Statesman:

The university is prohibited by federal privacy law from discussing an individual student’s academic performance or issues related to it. The university will respond to these claims through the proper legal channels in the courts.

We reached out to Martin for comment, and will update this post with anything else we learn.

Update 2/10/16 12:30 PM eastern: This morning, the defendants filed a response. It reads, in part:

If the disciplinary panel finds that S.O. committed academic misconduct, a sanction may be imposed, ranging from a written warning to the revocation of S.O.’s Ph.D. Any finding of wrongdoing and sanction of S.O. is subject to appeal by either party to the President of UT Austin, whose decision is final. UT Austin first informed S.O. of its disciplinary process and procedures on February 27, 2014.

As discussed below, S.O. fails to plead a valid constitutional claim of deprivation of due course of law and equal protection. Indeed, S.O.’s requested relief seeks to avoid due process. S.O.’s wish to avoid an academic misconduct hearing does not, itself, create a right to relief at this stage. Instead, S.O.’s only remedy is to raise any legal concerns once UT Austin has fully completed its administrative process concerning the allegations of misconduct against her.

Here is a PDF of the whole thing.

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 new daily digest. Click here to review our Comments Policy.

  • aceil February 9, 2016 at 1:02 pm

    She is absolutely right. Students should not pay the price for their advisors’ incompetence or negligence. Furthermore, students submit their dissertation in PARTIAL Fulfillment of the PhD requirement. Detecting honest errors afterwards does not justify a retraction of the degree.

  • MannyHM February 9, 2016 at 4:02 pm

    Too bad for her picking up the wrong mentor.

  • ColinP February 9, 2016 at 9:16 pm

    So what we do not know from this (and the prior story on Retraction Watch) is whether Ms. Orr herself falsified any data. If she played a role in data falsification, then it may be proper to revoke her degree. But new evidence would have to come to light, since UT already granted her degree. It is reasonable to give a graduate student a “pass” if data were falsified by others, and she was unaware. Her supervisor is ultimately responsible, as has been pointed out.

  • Kenrod February 10, 2016 at 8:36 am

    “… the school has scheduled a “hearing” on March 4, during which three undergraduate students and two faculty members will deliberate…”

    Three undergraduate students? Are there more details available re. the nature of this hearing or can anyone shed light on this sort of process? It strikes me as quite odd.

    • aceil February 10, 2016 at 11:07 am

      “The panel consists of a journalism student, a
      business/marketing student, a Plan II/liberal arts and mathematics major, a natural sciences faculty member, and a faculty member from the law school”
      It seems that the decision has already been made! No?

    • Just Me February 11, 2016 at 8:15 am

      Odd, maybe but no more odd than the reality described in the linked PDF. Per that court document, the plaintiff started at UT in 2003 and, after her (first) degree revocation, herself requested this review format, which was not codified in University rules until 2013. In 2003 that code described one option, a single faculty reviewer, with no students involved. The plaintiff rejected that option in favor of the later incarnation, and now protests the inclusion of student reviewers.

      I smell smoke.

  • MB Richardson February 11, 2016 at 5:07 am

    This whole thing strikes me as really weird. The retraction notice says that the reason for the retraction is problems in reproducing an olefin metathesis of a key intermediate – and that therefore the sequence from that point forward is also difficult to reproduce. This is actually pretty common, and I don’t think would be grounds for retraction. This is completely at ends with what Orr’s lawyer is saying, which is that the identities of the compounds were incorrectly assigned. This would definitely be grounds for retraction, but is not what the retraction notice says. In either case there is no accusation of wrongdoing or misconduct, but that doesn’t mean Orr’s status as a PhD graduate is safe.

    If we are to believe the retraction notice, I think Orr is in the clear. It’s not uncommon to have a reaction work once and never quite get it working again, and most chemists will have struggled with this at one time or another. However, if you get enough compound that one time it did work, and manage to complete the rest of the sequence, then you’re fine. Chemists go to a great deal of effort to prove what compounds they have made – this includes 1D and 2D NMR and hi-res MS as the bare minimum, and these spectra would have been submitted to the journal with the manuscript. We may never know why this reaction worked this one time – but we would know that it did work because these spectra would provide the evidence for the product. If it is difficult to reproduce – fine – someone can go back and troubleshoot it and probably get another paper out of it, but the article doesn’t need to be retracted, and Orr is entitled to the award of PhD.

    The lawsuit says something entirely different, that these compounds never existed, but were thought to because Orr’s interpretation of the data was poor. It then goes on to blame Orr’s supervisor, Prof. Martin, for not catching it. While I agree that Prof. Martin must shoulder some of the blame for this, Orr and her colleague (who is the postdoc?) are ultimately the ones responsible for misinterpreting the data. This should not happen, and if they were worth their salt as chemists, it would not have happened. If the lawsuit’s version of why the article was retracted is accurate, then Orr has failed to demonstrate the competence expected of a PhD organic chemistry graduate, and it is worth revising whether Orr’s other contributions are strong enough on their own to justify the awarding of a PhD.

    There’s more. Orr’s degree was originally revoked because a university investigation determined that Orr had fabricated the data in her thesis and in this article. Three different stories, and I think I know which one I believe.

  • MB Richardson February 11, 2016 at 6:51 am

    If there are any chemists reading this, please look at the supp info for the retracted article. It looks like these compounds probably were successfully synthesized, although it’s hard to be sure; the proton NMR for compound 24 is pretty convincing, although the carbon NMR is very poor. Compound 25 looks filthy, and the yields should probably be revised, but otherwise it looks okay. The carbon NMR of 25 is basically just noise, and I wonder if this is the “fabrication” allegation comes from. Orr’s thesis lists full 13C NMR assignments for 25, but if it is based on the published spectrum, most of the resonances have been pulled from thin air. Compound 27 has a completely unacceptable proton spectrum, with bad phasing and probably some shimming issues as well, and there is no 13C spectrum for it either.

    If people are going to attack Prof. Martin for overlooking the spectacularly bad quality of the data for these compounds, then surely the editors and reviewers that gave the green light for publication are equally accountable. Even so, I think a certain level of critical data evaluation is expected of someone aspiring to a PhD in organic synthesis, and I see evidence to the contrary here.

    • NH February 12, 2016 at 4:00 pm

      I agree with these comments, and would include the 13C NMR spectrum of compound 26 in the list of poorly obtained/useless spectra. The signal assignments seem quite subjective based on the spectrum. It is concerning that the 13C spectrum included in the SI does not feature a number of signals described in the text. Compare the reported data on page S11 with the actual spectrum on page S37 – some of the claimed signals do not appear in the actual spectrum.

      It is curious as to why most of the spectra for compounds 24-27 do not display the integration and/or chemical shift values for each signal; the only one that does display these data is the poorly shimmed and phased 1H spectrum of 27. This spectrum also offers no definitive evidence that compound 27 contains a carboxylic acid group because the chemical shift range ends before the CO2H signal would be expected to appear (>10 ppm). The OMe signal from the ester group of 26 appears to be absent, but this does not seem sufficient to confirm the presence of CO2H group in 27. The lack of 13C data for this compound is problematic, as is the absence of infrared spectra for the carbonyl compounds.

  • JS February 12, 2016 at 12:55 pm

    Hi there,

    To answer to MB Richardson, I’m an organic chemist and obtained my PhD few years ago and had a look at the data as suggested. I am not sure what to think and I think there is few information missing to really judge this. The point seems to be mostly turning around crystal data to prove the structure what I feel is a little bizarre. Crystal data is not always necessary to prove the identity of one compound because all compounds do not crystallize. We can just think about lipids.

    “he gave little guidance on how to crystallize molecules”
    This is very weird. There is plenty of papers out there on the art of crystallization and there is plenty out there to try. Advisers usually let PhD students learn by themselves and that’s part of learning to be an independent researcher.

    Some don’t even provide NMR spectra in their supporting information and provide only the assignments, heck some even only refer to previous paper for the assignments. So I think she was mostly transparent showing her data even if they didn’t look that great. Most chemists use a HSQC for characterization purposes and they are usually of much better quality compared to the carbon. HSQCs are not usually shown in papers but it is good practice to show the carbon and it often depends on the reviewers if there is proof enough for the paper to be accepted or not. So I think we don’t have all the information to judge this.

    What is the most interesting, I believe, is the reproductive part of it. Sometimes an experiment will work in a lab but not in another and will work in the hands of someone but not someone else. It might depend on solvent purity/dryness, skill, and a little bit of luck. Sometimes the quality of the purchased reagents might also be in cause…

    Not sure what to think about that one,

    • MB Richardson February 12, 2016 at 6:17 pm

      I think it is fine to use the 13C spectrum observed through the proton channel in an HSQC – but the HSQC spectrum should be provided. Having said that, the paper reports that >80 mg of this compound was obtained, which is plenty to get a decent 13C NMR even on old spectrometers.

      I also thought the crystallization thing was strange, as Orr’s lawyer seems to be asserting that this is the obvious choice that would have circumvented the need for other characterization data. But even with a crystal structure the proton and carbon NMR still have to be provided. More to the point, not everything can be crystallized, and it is just stupid to blame Prof. Martin for not bestowing on his students a supernatural ability to obtain crystals.

      Part of me wants to think that this is a lost-in-translation problem between Orr and her lawyer, but I can’t help but feel that this is really what Orr thinks.

  • JE ROSEN February 23, 2016 at 4:19 pm

    JE ROSEN February 23, 2016 at 3:20 pm
    I spent the greater part of my Ph.D. research doing organic synthesis. When it came to my defense I had prepared extensive characterization and analysis of all the compounds that I had synthesized that reflected their structure and purity. Organic characterization is a tricky business but any proficient organic chemist would be able to verify the characterization results. Any competent organic chemist knows that one method of characterization should confirm another…and when added together there is not much wiggle wroom for error. In the end it is the responsibility of the mentor and the student’s committee to extensively review the characterization data and verify it just as reviewers analyze data for peer reviewed journals. If there are questions as to the veracity of ANY of the data reported, both the mentor and the committee have a contractual fiduciary to the University to insure that what is being reviewed is correct and if not that the student be asked to reproduce the data or redesign the experimental protocol to compensate before the authorization of a doctoral degree for the work. The university retains faculty to insure that the dissertations and published work reflect the highest standards of integrity in the scientific process. It appears that the committee and the mentor failed in this regard and allowed a possible compromised thesis to be approved, well then they breached their fiduciary to the university. The student did nothing wrong…other than possibly commit honest error that was not caught by her alleged proficient mentor and committee.

    What I find amazing is that the mentor has not been able to correct for the alleged errors in all this time and establish exactly where he and the committee erred. The university is repsonsible for this debacle, not the student, because her thesis was approved….and her degree was granted and apparently a paper was published. The university, the mentor and the committee are liable to the scientific community to correct the manuscript with an erratum….produce the correct/amended results with supporting documentation and data but in all this time they have not done so. The student was granted her degree under the authority of the university, By taking action to revoke a degree the university is placing the spotlight on their own incompetence, not on that of the student, while at the same time ruining the reputation of the student and irrevocably harming her future career prospects. The university was responsible for guidance and they failed. To preserve their reputation, the university should make every effort to correct for the alleged errors and publish them, apologize to the scientific community for their own failings in their fiduciary to guide and produce a credible Ph.D. dissertation and move on – and leave the student alone. The time to prevent the granting of a degree is at the time the thesis is presented to the committee and to the mentor….not years later.

    This is a case of clear incompetence on the part of the mentor and the advisory committee…..and the student should not be used as a scapegoat. If I were a student seeking a Ph.D. in science I would stay clear of U of T in the face of what they are doing. Sometimes we have to eat our errors. In the end if the university is permitted to revoke this degree without due process for Orr in a court of law, what is to stop any university for any reason from revoking degrees from other students for all sorts of unproven nefarious reasons. For example….what if a mentor realizes years later that his student has become his competitor in his field and wishes to ruin that student by falsely claiming misconduct. This is a very dangerous situation for EVERY graduate student in the USA. I pray this case is permitted due process in a court of law….as no university should have the authority to revoke a degree years after it has been conferred without the benefit of due process in a court of law. Ruining a career and a reputation without due process is akin to a form of legal murder as this student may never be able to earn a living in their field….and this has to be stopped. We all need to wake up to the repercussions …..and it really surprises me that the university would risk its own reputation when their legal counsel has to know that the mentor and the committee breached their own fiduciary relationship to the university. The student is being used as a scapegoat for the errors of the U of T and a court of law will hopefully recognize this as such…and put an end to this once and for all. U of T has seriously damaged Orr….no question.

    • JS February 23, 2016 at 7:54 pm

      I am pretty sure that UT don’t like doing this. What do they gain from it? It’s bad advertisement for them. If she still works at Pfizer, most of her work must be confidential anyway, if she is still in the lab she probably cannot publish because of IP so I don’t see why a prof would want her not to compete with him. I was told many times that my PhD was my responsibility and my advisor is there to advise me and not control my work as well as the committee. Let’s face it, most PhD students see their advisor once per month max and ask questions for help on the course of their research ,they’ve got to be independent. So again we don’t know enough to judge I think.

    • MB Richardson February 23, 2016 at 8:10 pm

      I understand your outrage, but I think you are absolving Orr of responsibility prematurely. Synthesis projects generate thousands of NMR spectra, and specific assignments can take hours. Asking a PhD committee or supervisor to check each characterization themselves is just absurd; even a small synthesis group would generate sufficient data to completely inundate their supervisor with spectra. Not only is there nothing wrong with allowing PhD students to take responsibility for their own data, it is also necessary.

      I am actually relieved that U of T is taking the matter so seriously – the article was retracted (although the reason why it was retracted may still be open to speculation), they are calling into question the validity of a thesis which may contain fraudulent data, and are attempting to hold the student accountable. Why else would U of T take Orr to court if it wasn’t trying to correct itself? The term “scapegoat” is only appropriate if U of T required one to protect itself against an external inquiry, which is certainly not the case here.

      It’s too early to assume guilt, but don’t make the mistake of holding a teacher responsible for his student cheating on an exam.

      • First of all you have it wrong. Orr is taking U of T to court ….not the other way around. If a thesis is to be called into question the time to do it is BEFORE the defense NOT years later. The external inquiry as you call it ….is the lawsuit filed by Orr. U of T is on the defense because they are being sued..they are the defendants. Not sure how much case law you have read but there is clear evidence that faculty have a fiduciary relationship to their employer…the university, to protect and insure the integrity of dissertations and published manuscripts from students before a degree is conferred. They are the filter and the university relies on faculty and faculty committees for this purpose. Students rely on their mentors for oversight and guidance. If a falsified or fabricated dissertation slipped through in spite of all the peer review…then shame on the committees, mentors and reviewers. It is clear that the U of T system is broken – because to date they have not corrected the alleged errors. Furthermore even if the law was on their side they cannot prove honest error not only on their part but also on the part of Orr vs. deliberate manipulation of data on the part of Orr. Also an opinion written by the Texas AG demonstrates (opinion number M-466 dated 9/11/69……that universities have the right to confer degrees but not revoke them without due process in a court of law. Orr’s attorneys are set to present their agruments in support of the right of Orr to due process and remove the ability of U of T to have a group of graduate students decide her fate. U of T set up some group of grad students as their excuse for due process….they did not sue Orr in court to revoke the degree as you believe. Orr sued U of T for the alleged denial of due process that she is arguing is in violation of the Texas AG decision cited above. If you wish to have the supportive documentation on this I will be glad to share it. You can email me at and title the email….REQUEST FOR DOCUMENTATION IN THE ORR CASE.

        • MB Richardson February 25, 2016 at 7:18 pm

          Dear J.E.
          I think you misunderstood me. Nobody made U of T revoke Orr’s thesis in the first place, the university decided to do this on its own – this is what I mean when I say they are not acting in response to an external inquiry. There is no reason to draw attention to themselves in this way unless they are trying to protect the integrity of their higher degrees. I have pointed out in earlier posts that the Org. Lett. paper contained spectra that are simply not acceptable, and I speculate that the carbon resonances that are listed for those compounds in Orr’s thesis could be where the fabrication of data allegation comes from. If this is the case, the university has a good reason to argue that the award of PhD was given under false pretenses.

          I also don’t think you have any appreciation for how long it would take for a committee and mentors to confirm all of the data their students collect. It is absurd to blame them for not catching Orr’s alleged dishonesty – no one would support Lance Armstrong if he took the Tour de France officials to court to get his medals reinstated.

          I’ll happily look over whatever info you have.

          • Ken February 26, 2016 at 12:57 am

            Agreed. In any collaborative scientific research there is always a measure of trust and faith that everyone else will do the correct thing, and won’t do anything dishonest or stupid. Obviously we should and do check the results of others but there are limits. Someone told me of a simulation that was supposedly derived from first principles that had a fudge factor hidden away in the code that made the results match reality.

            It may well be that previously that nobody knew quite what went wrong, but there was something wrong and that resulted in the retraction. Now they think that there was a deliberate attempt to falsify data and that is serious. A bad interpretation or other error and it is reasonable for her to keep the PhD, it would be far from the first.

          • Yes during the scientific process honest error is common. If there was honest error that was not caught….just as journals publish erratums so should U of T….as it appears that Martin and his Postdoc published a peer reviewed manuscript on this work that they claim has errors – so they should be obliged to correct and resubmit the paper. If there is an accusation of deliberate manipulation of data to support the revocation of the degree then due process must be offered in a court of law….otherwise anyone without proof of misconduct can create a potential case against an innocent person…and then we would all be at the mercy of any vindictive soul, university or group who wishes to create harm for their own purposes. I believe we have evolved beyond the days of the Salem Witch trials. Universities are not set up to be platforms that enable due process….but in a court of law the university can offer into evidence, data, experts and witnesses….whereby a Judge and Jury would be responsible for a final decision…..not a group of inexperienced graduate students.One huge issue is that when complex scientific data is on the table it may be impossible for a Judge and Jury to comprehend it sufficiently to make proper judgement. In any case without proof of deliberate manipulation of data (that I believe they do not have) this case needs to be totally dropped by U of T…..that has now already seriously damaged the reputation of Orr for the rest of her life.

  • MB Richardson February 26, 2016 at 8:43 pm

    JE ROSEN in response to JS’s most recent comment to my post
    In any case without proof of deliberate manipulation of data (that I believe they do not have) this case needs to be totally dropped by U of T…..that has now already seriously damaged the reputation of Orr for the rest of her life.

    I would expect they do have evidence of research misconduct, or none of this would ever have happened. I have been trying to reverse-engineer what that evidence is, and suspect it has a lot to do with the assigned NMR resonances that appear in Orr’s thesis and in the Org. Lett. article for which there does not appear to be any supporting original data for.

    I want to stress that I am totally ignorant of the details – I became involved in this discussion because the early comments were attacks directed at Prof. Martin and the University of Texas, which I thought were in poor taste. If Orr has made an honest mistake, no big deal; if she was unable to reproduce the sequence, no big deal. However, if Orr subverted her PhD panel by lying about her data, I don’t think she is entitled to keep her PhD. I am as interested as anyone to know what is actually going on. I would also hate to see an innocent person stripped of their degree, but I honestly cannot see a university doing that for no reason.

    • je rosen February 27, 2016 at 1:54 am

      Again….all the more reason to support and allow due process in a court of law. Again under the law there is a contractual obligation between faculty and the university to uphold standards outlined in the university faculty handbook and by virtue of the actual employment contract between faculty and the university. The university by awarding a degree imparts and implied warranty of its value….it is property. By definition the relationship between faculty and the university is that of a fiduciary. Faculty owe the university a duty of care and they are hired to maintain and expect high standards of academic excellence for their students. Case law supports this. It was the obligation of the faculty and Orr’s committee to carefully review the data for its veracity prior to authorizing the defense and prior to awarding her the Ph.D. If there was or is evidence of deliberate manipulation of data by Orr, U of T can present it in a court of law where Orr should be able to obtain a fair hearing afteri all the evidence is presented. If U of T does not have this evidence….best for them to settle this case and pull out before they further shoot their own feet because it is my opinion that honest error will not meet the bar to revoke a degree in a court room. Without strong evidence of deliberate manipulation of data they are harming Orr’s reputation along with their own… no graduate student would risk attending such a school, knowing that years later the university might have the power to revoke a degree as a result of possible errors made by a committee or mentor years earlier. Hopefully all this is is honest error and we can all move on but the most important aspect of this case is for any graduate student at any university in this country to be given the opportunity to obtain due process – so the best that can happen here is for new case law to be embedded for others to rely on. Universities are NOT courts of law and the revocation of a degree is extremely serious and should not be the decision of a few graduate students on a committee.

  • D Brown March 18, 2016 at 8:18 pm

    Any information about the review which was on March 5, 2016??

  • Post a comment

    Threaded commenting powered by interconnect/it code.