Washington University psychology professor Todd Braver was Adam Savine's mentor and alerted school officials about problems with Savine's research. The university then launched an investigation, which resulted in Savine being placed on probation for three years. Braver's responses came in too late for inclusion in the Health section Thursday.
Q. How long did you serve as a mentor to Savine?
He began as a graduate student working in my lab in Fall 2007.
Q. How was it discovered that the research was questionable?
Adam had finished his dissertation research, submitted his final document, and was preparing for an oral defense in August of 2012. He was set to leave the lab to start a post-doctoral position at University of Michigan. As part of this, I asked Adam to make all of his analysis files and documentation available to me. Moreover, I was beginning to have some suspicions about Adam’s ethical integrity, so I asked him to walk me through each step of his analysis and files, to make sure it met my satisfaction. He stalled on this for a while, which made me even more suspicious, but eventually made the files available — though seemed to be avoiding the "walking me through" phase. As a consequence, I started digging into it myself. I started to notice some surprising discrepancies. At first, I chalked this up to my confusion in decoding some of his files (w/o his assistance), but I decided I needed to confront him directly. I told him that I would not let him go through with his dissertation defense until I was satisfied that his data were legitimate, and that he would need to attest that they had not been compromised in any way. Eventually he acknowledged to me that we needed to call off the defense. I was incredibly shocked by this admission, but reported it at once to the Academic Integrity office at WUSTL, so that they could go through a full objective investigation.
Q. What was your reaction upon hearing about the problems with the data?
I was shocked, angry, and incredibly disappointed. It felt like the worst kind of violation of my trust, time investment, and efforts towards Adam's mentoring. Being a graduate supervisor is in some ways akin to parenting, so it was like finding out that your child had just skipped town after robbing your and your neighbor’s house.
Moreover, I want to be completely crystal clear that in Adam's graduate training both through the department, and my own one-on-one supervision, Adam was repeatedly given instruction regarding ethical conduct in research. We discussed very explicitly what would be considered data fraud, academic integrity violations, etc., and what would be the consequences for doing that. So, there could never have been any ambiguity in Adam’s mind that what he was doing was unacceptable and would never be condoned.
Q. What happens to the federal grant money that supported the discredited research?
Adam's tuition and stipend were supported by graduate fellowships that were awarded to him directly (NSF graduate fellowship and NIH predoctoral awards). Because of his actions, Adam will not receive a degree from WUSTL.
The research funding was awarded to me as a PI, and only a small part of those funds relate to the research that Adam was involved with. Once the papers are retracted, and the data is released to me, I will do everything possible to thoroughly and accurately re-analyze the data to find out what findings are correct and what are not. I take very seriously my obligation as a steward of the research funds that allowed for the studies to be collected. Of course, it will take a fair bit of time for the data to be thoroughly re-analyzed.
Q. How are graduate students supervised, and will that process change now?
There are, of course, individual differences in how supervisors carry out mentoring and supervision of students. For my own part, I operate very closely with the students at the beginning of their training to make sure their skills are firm and that they have carried out analyses appropriately and correctly. However, as they progress to more senior stages of their training — in preparation for being independent investigators — I give them more responsibilities and independence. I believe that this is a very standard practice among supervisors, and is important. Students need to learn how to operate autonomously so that they are prepared to do this when they move on to the next stage of their career.
Nevertheless, this affair has caused me to think about what we can do to move towards more standardization in increasing the transparency and reproducibility of our data and results. I believe that this will be an important discussion not just for my own lab, but also for the field in general, and has been one that is receiving increasing attention.
Q. How common is it for doctoral candidates to appear as lead authors on research papers?
It is very common in the field of Psychology and Neuroscience. Indeed, having high-quality lead authored papers is one of the primary means that senior graduate students demonstrate they are ready to move on to the next stages of their careers (as post-doctoral fellows and/or Research Scientists/Professors).
Q. Have you ever been in a situation like this before?
No never, nothing in any way close. And I hope and pray that it is the last time I ever have to experience something like this first-hand. It is gut-wrenching.
Q. What are the implications for other papers that reference the retracted works, particularly this January 2013 study: http://ccpweb.wustl.edu/pdfs/Braver2013Temporaldynamicsofmotivation-cognitivecontrolinteractionsrevealedbyhigh-resolutionpupillometry.html#B23
The study you referred to used an experimental preparation and methodology that were substantially different from the ones used in Savine's research. The theoretical questions and conceptual issues were very similar, which is why it served as an important follow-up. The positive results of the 2013 paper are very reassuring to me, in that they provide independent verification that the ideas and hypotheses pursued in Savine's research are still valid ones. I do believe that we have enough of these independent verifications that the basic ideas will hold-up even if we discover that some of the specific findings reported in Savine's papers are not valid or accurate. To me, this is one of the most important aspects of scientific research — that it is objective, should be replicable in principle, and is self-correcting. For my own part, although I have been deeply discouraged and disheartened by Adam's actions, it has not had any negative impact on my own views about the positive contributions of our scientific research, or of the scientific endeavor in general.
Lastly, even though you did not ask me about this, I also wanted to comment on one other aspect of the process, as it has unfolded in Adam's case. I’d like it to be clear that I was caught blind-sided and am pretty disappointed by the way that Adam's data fraud has come to light. Although I believe it is important for there to be public acknowledgment of these cases and Adam’s wrongdoing in particular, the way it occurred was extremely unsettling to me. Specifically, I was not informed or forewarned by either WUSTL or ORI (the Office of Research Integrity) about the facts discovered in Adam's case, nor that these would be reported by ORI on a publicly accessible website. You learned of the outcome of Adam's case before I did, which I am pretty upset about -- given that I was the one to report him in the first place. Although I gave testimony about my own knowledge of his wrong-doing, I was not given any other information throughout the case. I understand that WUSTL did this to make sure that I was kept from interfering in the initial stages of the investigation -- and so that I could be cleared from any culpability or responsibility (which I have been)-- but I am pretty disappointed to have been left in the dark for so long (the investigation was concluded in Dec), given that I am the senior author on (and thus have responsibility for) 2 of the papers that I just learned will need to be retracted. My personal opinion is that both universities and the ORI need to work on ways to improve the procedures associated with investigation of research integrity violations, so that the collaborators and colleagues of a perpetrator of data fraud due not receive undue hardship or collateral damage. It was very upsetting to learn from international colleagues and reporters about the public web disclosure of the outcome and specifics of Adam's discovered data fraud before I received any information from my own university or the ORI.