November 28, 2007

U. of Washington Scientist Fabricated AIDS Research, Report Says

An AIDS researcher at the University of Washington altered images and falsified data, according to a 2003 report that recommended he be banned from ever working there again and that was disclosed only after The Seattle Times fought a legal battle to obtain the document.

The report, which was based on a 16-month investigation by the university, said that the scientist, Scott J. Brodie, a research assistant professor, had fabricated data in 15 instances, including manipulating an image of a single cell into “two distinct images presented as different types of cells” to support his paper’s findings. Some of his data were published in scientific journals.

Oddly enough, despite the fraud, other scientists have confirmed Mr. Brodie’s findings. But in 2002, a researcher reviewing the work discovered inconsistencies and reported them to the federal Office of Research Integrity, which, in turn, notified the university.

The report was not released until now because Mr. Brodie had sued both the university and the newspaper to block its disclosure. A judge ruled for The Seattle Times last week. Mr. Brodie, who resigned from the university in 2003, now lives and works on the East Coast, but the newspaper could reach neither him nor his lawyer for comment.

The university submitted the report’s findings to federal authorities, who are in the process of their own investigation. —Mary Andom

Comments

1. Wow, this is ridiculous people in the science community are supposed to have more integrity than this and falsifying one’s research is terrible, it’s like an athlete who uses steroids to get better no matter how you look at it it’s still cheating. i think drastic measures should be taken and i don’t believe this is the first time a scientist has done this.

   — Elijah Cavender    Nov 29, 10:17 AM    #

2. I don’t think steroids. Probably – marihuana. And this would explain why others were able to confirm his “research”

   — Michael Pyshnov    Nov 29, 10:57 AM    #

3. there is a 5 in 100 chance that a study finding reported as significant is really insignificant (accepting a false hypothesis). There is there is a 5 in 100 chance that another researcher could confirm this first study and 25 in 10,000 that the two studies could happen one after the other sequentially. A third and a fourth confirmation would be much rarer but could occur. The problem is that we have no information on data on how many studies failed to confirm these findings and where these studies occurred in the string of replicating studies. This disconfirming studies would “discount’ and “net out” the confirming studies and give a better estimate is the confirming studies were statistical rarities or if they also should be closely examined for fraud. Cheating in science is really to “break the covenant bigtime” even given the immense pressures of various kinds and although meta-analysis helps a great deal to minimize the effects on these fabricated studies in the forming of generalizations, the qualitative importance of some individual studies cannot be under-estimated or discounted in terms of how other scientists and reviewers of scientific studies behave. Cae in point, this study was perceived as being important enough to replicate and apparently several times!! Lastly, there is a phenomenon with certain kinds of data (often visual) called the phantom effect (see Chapman and Chapman) where groups of professionals will see effects and relationships in data that are not there objectively. This defect of the ‘cogntive apparatus” adds another layer of difficulty and may account for the “replications’ perhaps in part if not all but a closer examination of the replication studies should be done to understand the impact and mechanisims of fraudulent studies. After all, one “replicator” went ‘whoa’ something is wrong here and found the fraud. We want to better understand this person, but also the people who did not go “whoa’ and replicated the study. Both will help us to improve training and the conduct of science.

   — jc    Nov 29, 11:36 AM    #

4. Seems to me there’s another possibility. Most grant-funded research falls in the realm of what Kuhn called “normal science”—i.e. extending what we already know. Proposals are based on the prior work, and typically are incremental in the new ground they seek to cover. A reasonable proposal is reasonably likely to work. Perhaps the fraud was due, not to a bad idea, but to insufficient skill/patience/luck to do it right. A good idea, perhaps, that actually worked in the right hands! BTW, This by NO means exhonerates the cheater; as previously noted, the scientific enterprise is built on trust, and would collapse like a house of cards if fraud weren’t diligently identified and villified!

   — Vickie    Nov 29, 12:18 PM    #

5. I believe that someone needs to step up and take responsibility for this mans actions. What he did was morally wrong. Changing research to prove that you were right all along? To me it doesn’t seem like he is much of a scientist. Aren’t scientists supposed to search for the TRUTH? Not just make up something to further their own career.. To me it just seems all wrong. And it really makes me wonder about other medical or scientific things I have heard in the past. It just goes to show you, you can’t always believe what you hear. Thats truly sad.

   — Shauna Mitchell    Dec 3, 05:01 PM    #

6. I don’t understand how a man can fabricate these scientific findings 15 times and not get caught sooner. How did he manage to get his work into scientific journals? Why would he even want to publish information that he knew was false? This isn’t just a case of cheating or morals, this is about possibly hurting others. He lied and that’s bad enough but he put his “findings” out to the public! It makes you think if one man has falsified scientific findings 15 times how many others have done the same thing, and still haven’t been caught.

   — Miranda Crace    Dec 3, 11:02 PM    #

7. I think that it’s terrible that so many people donate time and effort to help solving the AIDS crisis, and it’s horrible that there has been false evidence presented about this problem that’s killing thousands of people every year. It is concerning to hear how long it’s taken for people to realize that this has been going on. Research by scientists on such an important issue are things one should be able to trust. Anyone is capable of making a mistake, but the fact that Mr. Brodie did not admit to making a mistake and has been unavailable for comments ever since, shows that this probably no accident. It also seems extremely cowardice. These findings could effect people’s willingness to donate time and money to the AIDS crisis.

   — Jenny Ferch    Dec 16, 03:18 PM    #

8. This is a total railroad case by the Seattle Times. Seems they didn’t get much for spending legal money for this story.

   Think about it the University of Washington spent 16 months this guys info. for 8yrs of work and not one paper was retracted and not one grant was given back? Read Research Misconduct at the ORI.hhs.gov site. Institutions are suppose to spend 60 days for inquiry and 120 to investigate NOT 16 MONTHS.

   How many people worked in the lab? For years the UW never gave proper storage space or good computers for reasearch. I saw research slides stuffed under lab tech work benches and in the cleaning closets. 3 seperate labs work in a small space with no IT person to help. They all worked to fix stuff when computers crashed or not enough storage was available on computers.

   Check out the “investigators” they were anybody they could get to tell the UW’s story. Find out there names. The Ligget investigator (guy check his background)…he always had a personal beef. He enjoyed participating in the railroad job and the Times gave him even more glory by printing his words.

   If anybody is outright guilty of misconduct it is the University of Washington.

   ORI Misconduct Cases Highest Since 1996;
   Office Speeding Case Closures, Official Says

   The Office of Research Integrity closed more cases and made more misconduct findings in 2006 than it has in a decade, the office reported in March.
   ORI made 15 research misconduct findings and closed 35 cases in 2006, the highest number of misconduct findings and closed cases since 1996.
   John E. Dahlberg, director of the division of investigative oversight at ORI, confirmed the statistics and told BNA March 13 that since taking over the division last May, he and his staff have tried to channel more of their focus on bringing cases to closure, while at the same time working more aggressively with ORI attorneys to get the misconduct cases handled.

   ORI also closed more cases than it opened in 2006, with five to 10 cases that Dahlberg said his staff is aggressively trying to close. He said his division is on track to keep up a similar pace in 2007.

   Why has the ORI not yet looked into Brodie’s case?

   The UW should never release any information until both the local and especially the Federal level has concluded their investigation. My bet is the UW will be found guilty of misconduct themselves since the broke every rule they are suppose to follow and made-up a misconduct case since they were looking very stupid by taking 16 months. I find the UW suspicious themselves that they keep all the publications in print and they kept all the grant money.

   Now tell me who is really the guilty party?

   — George    Dec 18, 04:00 PM    #

Previous: U. of Minnesota Endows Chair Honoring Professor Who Brought Costly Fame
Next: Might a Bush Friend Be Pressing Spellings on Nelnet?