Hypothesis as a Brand

Jan 22 2018 Published by under Uncategorized

Some time ago I was talking with another PI about a current paper. It builds on some old work I started as a postdoc. I'm pretty satisfied how it has turned out so far. There's just one thing. Way back as a postdoc I posited that a particular variable 'A' was mainly responsible for variation in outcome 'Z'. I speculated about it in my old papers, mentioned it at conferences way back then, etc. Turns out I was wrong. My current work shows more clearly that it's more like variables 'A' through 'F' are responsible for outcome Z. This is not an unusual situation for science. Turns out it was more complicated than I thought. So, I got it wrong but now I have a better idea of what is going on. We're making progress. Yay.

Other PI furrowed her brow. "oh no! don't say that you were wrong."

In her estimation your brand as a researcher is about what hypotheses you support. Not a general theoretical view, not a reputation for rigorous work, or anything like that. Pick something you think is true and stick with it. Here the worst out come is being wrong. Not uninterpretable data, or no data. Other PI is not alone in her attitude. I've heard similar from a few other senior researchers.

Think about structuring your career about never being wrong. About some thought you had as a 2nd year postdoc having to be correct. Sounds stressful but I think some well known folks in my field see it that way. This came to mind because of the back and forth about researchers being "attacked" about work supporting their hypothesis (e.g., stereotype threat, social priming). I wonder if some of these researchers see their brand as a specific hypothesis being correct. That would make any critique of the data supporting the hypothesis feel pretty stressful.


8 responses so far

  • drugmonkey says:

    This is terrible and horrifying that people think this. Direct link to fraud.

  • ecologist says:

    I would be more encouraging. It seems to me there are more options for describing the situation than "right" vs. "wrong". Maybe this is what OtherPI was referring to.

    Don't say you were wrong. Say "Turns out it was more complicated than I thought." And no, those are not the same thing. The process you describe, where you hypothesized that it was A causing Z, but after some work it appears that it's A,B,C,...F causing Z is not being wrong, it's the way things are supposed to work. Probably (at least often), your appreciation of the role of B,C,...,F comes about because you started by thinking it was A.

    So, the alternative to saying "I was wrong" is not saying "I was right and won't give up my hypothesis", is saying "It was more complicated than I thought, and here is how it looks now...".

  • Anon says:

    "Don't say you were wrong. Say 'Turns out it was more complicated than I thought.'"

    In other words, you were wrong. (Yeah, ecologist, that's called being wrong.)

    I respect people who can call a spade a spade. I find that an invaluable quality in science. But I may be in the minority.... How sad and how awful for science!

  • Ria says:

    Bashir, thank you very much for posting this, as it's a succinct way to address something that's been bothering me for a long time (since grad school).

    DrugMonkey is right...this is a direct link to fraud. I've run into too many scientists who think this way, and in all cases, it results in harm to the careers of the junior scientists that they are to mentor, to the scientific enterprise as a whole, and even to society. Science is all about recognizing what is factual in your data, and what is not, and attempting to infer a glimmer of truth from those facts. It is NOT being able to say "I'm right, never wrong."

  • qaz says:

    The most powerful thing one can say in science is "I was wrong." I hypothesized this, predicted that, turns out I was wrong, now I know this new thing. That's progress.

    Not enough scientists say this and own it. In my experience, the ones that really do, particularly when they say "Now we know this new thing", are the ones who are most successful and most respected.

    I think a large part of the problem is that people are often sloppy about theory, hypothesis, and experiment. Being careful about what you know, what you don't know, and what you are working on, allows you to be wrong in the right way.

    Personally, I love being wrong. It means I learned something. Sometimes it means I learned something no one ever knew before.

  • shrew says:

    Never ceases to amaze me when people like this say the quiet parts loud.

    I can't say that I love being wrong, but I can say that I always learn a lot from it. Sometimes I've had to put data away for weeks or months (or even, once, years) before I could understand why I was wrong and get a better idea of what might be going on instead - of course, said 'better idea' is probably also wrong.

    You know, we're always saying 'ideas are cheap', because they are, but maybe some of these people think they're only ever gonna have one halfway good idea? What a sad way to be a scientist.

  • Dave says:

    It's a shame that science has become about who is "right" and who is "wrong". There is an awful lot of value in publishing data that turns out to be wrong. If it moves the field forward and enhances our understanding of pathway X, then all is good. If it brings other people into the field to build on the story, that's good too.

    My personal opinion is that a lot of this has been driven by the "replication crisis" crazies, who often push for papers that are wrong to be retracted. People are scared to death of being caught up in that mess, hence the resistance to accept when something is just wrong. We need to tone down that rhetoric, and accept that science cannot progress unless we learn from our mistakes.

  • sciliz says:

    This is silly. If you have no particular allegiance to The Truth as determined by The Data, you may as well go full post-fact and simply pretend you said it was due to A-F all along.

Leave a Reply