More Recent Comments

Sunday, January 13, 2008

Scientific Mistakes

During a recent discussion with undergraduates, they mentioned that it would be a good idea to discuss the more recent scientific papers in class. They seemed to be very impressed with a course that presented papers published within the past few months.

I pointed out that there's a problem with that kind of course. If the goal of a course is to teach fundamental principles and concepts then it's very unlikely that recent papers are going to advance that goal. Why is that? Because much of the scientific literature is either trivial or incorrect. You don't know that it's trivial or incorrect until some time has passed and other scientists react.1

If the goal of a course is to teach how science is done on a day-to-day basis, then a key part of that course should be to drive home the concept of skepticism. Don't believe everything you read in the latest journals. An important part of that teaching goal is to pick examples of important mistakes in older literature.

John Dennehy has helped us out this week by posting a "citation classic" that turned out to be wrong [This Week's Citation Classic: Being Wrong]. In my opinion, it's far more important to look at examples like this than to expose undergraduates to several dozen hot new papers that are supposedly at the cutting edge.

The paper that John choose is by Paul Boyer who subsequently won the Nobel Prize in Chemistry [Nobel Laureates: Paul Boyer and John Walker] for his work on the mechanism of ATP synthase [How Cells Make ATP: ATP Synthase].

1. Sometimes it takes a long time for scientists to react to mistakes in the literature. Wrong ideas can be perpetuated for decades after they've been refuted, especially if the original papers were widely referenced. I was reminded of this the other day when listening to a graduate student seminar—coincidently, on the structure of ATP synthase. The student posted an old-fashioned, out-of-date view of the citric acid cycle as an introduction to the function of ATP synthase. I have challenged my undergraduate biochemistry class to find a single example of a web site that gets the entire citric acid cycle correct. There's a prize. They can't use the IUBMB site, they can't use my sites, and they can't make one of their own. So far nobody has collected the prize.


  1. Seriously, some students want to miss out on all the traditional teaching and go straight to paper reading? How can you get to that level without learning the fundamentals? What some students forget is that the fundamentals don't change - they've been established and been subject to years of scrutiny. My undergraduate physics degree would have been a jumbled mess of knowledge if it was taught the alternative way. Learning about classical experiments and theories help you understand how the new ones developed. You end up going through some of the same thought processes that the older scientists did when you see what they had to deal with and try to explain. I would say that if anything, student's need more of the fundamentals and the historical context in which these scientific facts were discovered. I wish there was a course like "history of biochemistry" offered, something like the "history of math" course offered by the math department. I think what students are intimidated about here is that the burden is on them to learn all the important science that has been done before them. That is a daunting task, but each generation of scientists must acquire at least a working knowledge of what is accepted scientific fact in their field, otherwise, ideas and experiments get repeated and recycled needlessly. I've seen this happen a few times at the poster sessions of the Biophysical Society annual meeting.

  2. I agree that a history of biology or biochemistry would be a good idea. In geology there's the historical geology course that not only serves as foundation for understanding the discipline but is also a good survey course for non-science majors (journalism comes to mind).

    A course in the history of the merging of biology and geology would be really cool too - From Hutton/Leeuwenhoek to the Hot Springs of Yellowstone to Mars and Beyond would be a good course title.

  3. I wonder if anyone could give some advice to any undergrads out there (like me) who know what field they want to get into in grad school and beyond and who want to learn more about their chosen area of science (beyond what is on offer at their school), in my case genetics?

    I've picked up a copy of Genes IX (which I believe is more advanced than Intro to Genetic Analysis that we used in my genetics class) to work through over the summer, and I am also intending to search pubmed for review papers related to my field from the last 10 years or so, which I assume will themselves contain references to classic papers which I can then follow up on.

    Does this sound like a good plan?

  4. I supposed you're implying that the scientific process isn't really complete (if it ever really completes) until a published paper has been subjected to months, perhaps years of critical scrutiny. It's the old dialectic between creativity and criticism.

  5. This post is so novel! (When you read that sentence consider the source.)

    I often find that when I read papers I lack the sufficient background so I'm guessing you gave them easy papers or you just summarized the papers for them in your lectures.

  6. This is one of a series of IMO excellent posts putting forward the idea that science doesn't result from a contest of belief systems, but from what survives healthy skepticism and a demand for evidence.

  7. "If the goal of a course is to teach how science is done on a day-to-day basis, then a key part of that course should be to drive home the concept of skepticism. Don't believe everything you read in the latest journals."

    Three years ago I met a someone whom I had known as a student around 20 years earlier. She told me that the most important thing she had learned as an undergraduate had come from my kinetics lectures, when I emphasized that students couldn't beleve everything they read in textbooks. On the other hand I was very glad that she had remembered this and that she thought it important, but I was also surprised that no one else (apparently) had conveyed the same idea in their lectures.

  8. I'd be very interested to learn what the common errors are in depicting the Krebs cycle, as well as a link to a page that has it right. Thanks.

  9. The most common errors are ...

    1. Equations not balanced, especially protons.

    2. FADH2 incorrectly given as the product of the succinate dehyrdogenase rection.

    3. Forgetting inorganic phosphate in the succinyl-CoA synthetase reaction.

    Of these, the depiction of FADH2 as a reaction product is the most egregious error.

    Even the IUBMB site gets it wrong, although, in fairness, the two mistakes are probably typos.