I think the Common Core is actually pretty good

Just read this article on NYTimes.com about people pushing back on the Common Core, which is about how a lot of parents and educators are pushing back against the "Common Core", which emphasizes problem solving skills and conceptual understanding over rote application of algorithms, i.e., plug and chug. I can't say that I'm super familiar with the details of the Common Core, but I can say that I now that I have taught the undergrads at Penn who are the products of the traditional algorithmic approach, it is clear to me that the old way of teaching math was just not cutting it. Endless repetitions of adding and subtracting progressively longer numbers is not a way to train people to think about math, as though the ability to keep numbers ordered is a proxy for conceptual understanding. I think many of the critiques of Common Core actually show many examples that seem to highlight just how much better the Common Core would be at teaching useful math.

Take as an example adding two digit numbers. I don't know anybody who does math as a part of their job who does the old "carry the 1" routine from elementary school. To me, a far better way to add numbers (and is the basis for the algorithm) is realize that 62+26 is 60+20 + 2+6. This is exactly the object of ridicule #7 in the previous link. I've been teaching my son how to add and subtract this way, and now he can pretty easily add and subtract numbers like these in his head (and no, he is not a math genius). From there, it was pretty easy to extend to other numbers and different situations as well, like, say, 640+42 and the such. I see absolutely no point in him even bothering to learn the old algorithms at this point. I think that those of us who developed mathematical skills in the old curriculum probably succeeded more despite the system than because of it.

The results of decades of algorithmic learning are students who have little conceptual understanding, and even worse, are frankly scared to think. I can't tell you how many students come to my office hours who essentially want me to spoon feed them how to solve a particular problem so that they can reproduce it correctly on a test. The result is people whose basic understanding is so weak and algorithmic that they are unable to deal with new situations. Consider this quote from the NYTimes article from a child complaining about the Common Core:

“Sometimes I had to draw 42 or 32 little dots, sometimes more,” she said, adding that being asked to provide multiple solutions to a problem could be confusing. “I wanted to know which way was right and which way was wrong.”

Even at this young age, already there is a "right way" and a "wrong way". Very dangerous!

I'm sure this Common Core thing has many faults. Beyond the obvious "Well, if it was good enough for me, grumble grumble harrumph!" reactions, I think there are probably some legitimate issues, and some of it probably stems from the fact that teaching math conceptually is a difficult thing to systematize and formalize. But from what I've seen, I think the Common Core is at least a big step in the right direction.

Why bother studying molecular biology if the singularity is coming?

Perhaps I’m just being hopelessly optimistic, but I believe Ray Kurzweil’s singularity is going to happen, and while it may not happen on his particular timetable, I would not be surprised to see it in my lifetime. For those of you who haven’t heard of it, the singularity is when the power of artificial intelligence surpasses our own, at which point it becomes impossible to predict the future pace of change in technology. Sounds crazy, right? Well, I thought it was crazy to have a computer play Jeopardy, but not only did it play, but it crushed all human challengers. I think it’s a matter of when, not if, but reasonable people could disagree… :)

Anyway, that got me thinking: if artificial intelligence is the next version/successor of our species, and it’s coming within, say, 50 years, then what’s the point of studying molecular biology? If we consider a full understanding of the molecular basis of development to be a 50-100 year challenge, then what’s the point? Or cancer? Or any disease? What’s the point of studying an obsolete organism?

In fact, it’s unclear what the point is in studying anything other than how to bring about the super-intelligent machines. Because once we have them, then we can just sit back and have them figure everything else out. That spells doom for most biomedical research. You could make an argument for neuroscience, which may help hasten the onset of the machines, but otherwise, well, the writing’s on the wall. Or we can just do it for fun, which is the only reason we do anything anyway, I suppose…

Google Docs finally has "track changes"!

I love Google Docs! It's where we store tons of information in the lab–it is essentially indestructible, easy to use and easy to share. And, perhaps most important of all, it makes collaborating on documents SO much easier. No more endlessly sending around XYZ_manuscript_final3_actuallyfinal8.docx. The Word Doc sharing method is prone to errors and is slow, especially if you have many collaborators.

However, the one feature Google Docs was missing from Word was the infamous "track changes". Warts and all, this is an essential feature, since it allows editing while also showing you directly where the edits happened. Google Docs had great commenting features and had some version history feature, but it just was not as good as good-old track changes in Word, end of story. But now it exists! It's called editing in "suggest" mode, and you can change modes with the little pop-up menu towards the top right corner (it's in "edit" mode by default). So awesome! Now witness the firepower of this fully armed and operational Word replacement!

I sincerely hope I never write about STAP again

Got a few negative comments on my last blog post about the STAP fiasco. Seems like some folks think I’m being overly apologetic or that I have no idea that she faked data. Haha, reminds me why I should just stick to blogging about Chipotle! Also, for the record, in case anyone cares (which I sincerely hope they don’t), I do not have, nor am qualified to give, any opinion as to whether the papers should be retracted or whether STAP cells are real. This is about the vilification of Obokata.

First of all, let me just say (in case it’s not obvious) that if someone has faked data, well, then they should be out of the game, permanently. I think everyone agrees that intentionally fabricating data is a capital offense.

Where, however, is the line between sloppy science and fabrication? Is it the intent to deceive? Are we sure Obokata had such intent? Let’s look at the evidence that I am aware of (and for those out there who think I was completely clueless about the context, I had considered most of this before writing my post).

Here is the RIKEN investigation’s report. At issue are two main problems: splicing of the lane of a gel and duplicating some teratoma figures from Obokata’s thesis. To the former, well, it was an inappropriate manipulation (scaled and spliced a positive control lane from one gel into another because it looked a bit better), but the original gel data was there and, while the manipulation was certainly not a good thing to do, it seems as though she wasn’t even aware that this was a bad thing to do. Moreover, her raw data doesn’t appear, from what I can find, to contradict what she portrayed. I’m not saying that this was a good thing to do, just that it seems like an honest mistake. The RIKEN report does not contradict this sentiment, by the way.

Then there is the more damning issue of the use of teratoma images duplicated from Obokata’s thesis. Yes, this is outright fabrication. Again, are we absolutely sure there was intent to deceive? Yes, I admit it does seem a bit weird that she would have forgotten what pictures came from where, especially from her thesis. That said, Obokata says that she has since provided what she claims to be correct images, although the data trail is very weak (which, by the way, is the case for imaging data in almost every biomedical research lab I’ve seen). Indeed, the RIKEN report itself says that she and Sasai provided the “correct" images just before the investigation began when notified of the issue, and so they didn’t even think they needed to provide further explanation. Obokata also maintains that she submitted a correction to Nature.

Certainly, this is beyond sloppy, probably worthy of retraction. But it also seems true, based on the report, that alternative images showing the effect exist that are not duplicates of her thesis (be these images legitimate or not). So, is she a fool or a knave? To the former, if we assume that she actually has the correct images and they show the right thing, as she maintains, then what other rational explanation is there for the images than an honest mistake? I mean, it beggars belief that she would purposefully show duplicated images when she had the right ones in hand. Now let’s assume she’s a knave and that she didn't actually have teratoma images and needed to manufacture evidence. Again, there is STILL absolutely no motivation for her to intentionally use images from her thesis when she had other non-duplicated fake-worthy images in hand. So she’s either a (yes, very) sloppy scientist, or an utterly incompetent faker.

I’ve never met this woman, and I’m not here to defend her honor, but personally, without seeing more of the evidence, I would feel reluctant to destroy a young woman’s life by branding her a liar and a fraud. I also wonder if we’d all be giving her the benefit of the doubt on all this if her results were immediately replicated by a dozen labs (and Wakayama did actually replicate the experiment, with coaching from Obokata). Would we then be asking for an erratum instead?

As I pointed out before, it’s absolutely nuts to deliberately fake results like this. What is the endgame? It’s obvious that people will attempt to replicate immediately, and so any “fame” would be fleeting at best, followed by probably the darkest period in your life. Whether she misinterpreted her data and those putative STAP cells are actually dying cells or contaminating cells or whatever, is a separate issue than whether she intentionally misled people in her work. Being (honestly) wrong in science isn’t the end of the world, nor should it be.

Or whatever, maybe I’m just hopelessly naive and she really is a faker and a cheat. I, like others, am working with limited information. If so, fine, I was wrong, and maybe she does deserve to be piled on. I’ve already written way more about this than I ever intended, and the whole thing has taken up way more of my brain space than I wanted. I suspect I'm not the only one to feel that way.

Why there is no “Journal of Negative Results”

I think everyone at one point or another in their scientific career says: “Why isn’t there a Journal of Negative Results?” Turns out there is, actually, and you could argue that PLOS ONE would also be an appropriate venue. However, it’s worth thinking about exactly why we don’t generally publish negative results. There are the obvious political reasons, one being that by disproving someone else’s work, you have now made yourself an enemy, potentially a very powerful one, and another being that you typically don’t get much “impact credit” for a negative result. But there is a scientific reason as well, namely that it is much easier to confirm a finding than to definitively disprove it.

Let’s say someone had a finding about, say, expression of a gene going up in condition B as measured by RT-PCR. Then we in our lab try and confirm the finding using RNA FISH to measure gene expression. If we confirm the finding, great, done, that was easy! But if we find that the expression is unchanged, then what? In our lab, this is usually when we stop. But if we wanted to publish this negative result, we would not have satisfied the burden of proof. First, someone would say “well, your RNA FISH thing doesn’t work right, at least not in this case.” So now we have to do RT-PCR using their primers and conditions, etc. Then, “well, how do you know your conditions are exactly the same?” Now we contact the authors and try to replicate everything down to the smallest detail. Maybe we don’t get their RT-PCR results. Not useful, because while it might reveal that the original authors were incompetent, they could just keep saying “Well, you’re not doing the exact same experiments as us, so whatever.” Once the result is published, the burden of proof is on you, not on them. Suppose on the other hand that we do confirm their RT-PCR results. Now we’ve just bought ourselves months of work to try and figure out why there is a discrepancy between the methods. Not exactly glamorous work, especially when you already know in your heart of hearts that their result is wrong. Overall, the work required just doesn’t justify the reward, which is only compounded when you factor in the political risks.

That said, there is a type of negative result that would be less politically dangerous, although perhaps not quite paperworthy. Those are the little methods details that are typically buried within a lab’s collective brain. Like “oh yeah, don’t try to use the RNA FISH method in XYZ organism because the high autofluorescence will kill you.” I tried to do that for RNA FISH by setting up a FAQ, and I think it has proven to be useful. If you’re a methods-head, I would definitely give it a try.

Why is everyone STILL piling on that STAP woman?

[Update, 6/30: follow up post]

There is something weird about the STAP “scandal”. I still have no idea why there is such a hardcore investigation into these papers, now with talk of it shutting down the entire institute. Much of the inquiry seems to circulate around some apparently duplicated lanes of a gel and “plagiarism” of materials and methods (seriously?!?). Umm, well, the plagiarism of materials and methods stuff is just a stupid criticism. The duplicated lanes in a figure? Potentially more serious, but I’m inclined to say it’s just sloppy. Why? Because NOBODY in their right mind would publish this work without truly believing that the protocol worked. The only other logical option is that this woman is a pathological liar, because she must have known that everyone in the world will be angling to replicate this result immediately–there is nowhere to hide. And, admittedly from afar, I don’t think she’s seems like a liar. Rather, I think that either she made an honest mistake and got fooled by biology–wouldn’t be the first, won’t be the last–or the result is actually valid and just requires more time for others to replicate. I think it would be wrong to dismiss this latter possibility, by the way.

And what if the papers end up being wrong? So what. Papers are wrong all the time. I personally know of several papers that are wrong in high profile journals, ones that we and others have wasted time and treasure on following up. I’m sure you do, too. (Buy me a beer and we can trade stories.) In fact, there are entire subfields that I strongly suspect are bogus. Nobody is hunting these people down and debating whether they should close MIT or Stanford or Harvard, even if everyone in the room quietly agrees the work is bogus. Perhaps some of this work is not quite as high profile as STAP, but I don’t think that should matter. It’s all still in the record, helping the careers of those who write it and fooling those who read it. I’m not saying that it’s a good thing, I’m just saying there’s a serious double standard here.

I think it’s worth comparing the fallout here to that of two recent high profile physics “mistakes”. One was the faster than light neutrinos. The overall reception was “Well, it’s probably wrong, let’s just wait until they figure out what happened.” And it turned out to be a not fully plugged in USB cable or something. Fine. I don’t think there was any huge fallout calling for people to shut down OPERA, although the lead investigator did somewhat mysteriously step down from his post. But still, not as crazy as the firestorm surrounding Obokata. Then consider the very recent discovery of gravitational wave signatures of inflation shortly after the big bang. Turns out it might just be a bunch of dust. Oops! Still, nobody’s losing their job or shutting down an institute or being investigated like a criminal. And, with all due respect to STAP, finding particles that move faster than light or seeing signatures of inflation in the early universe are some pretty big results. So why this crazy witch hunt? Is it because she is a woman? I feel like maybe everyone is focusing all their “the biomedical research system is broken!” energy onto this one particular case, as though that will solve anything.

Whatever, all I’m saying is that you better hope you’ve got someone who believes in you when the science gestapo comes looking for blood. Because, in a slight bastardization of JFK, while success has a dozen coauthors, retraction just needs one scapegoat.

Is my PI out to get me?

I've been delving more into the world of science blogging lately, and it is a wonderful place, with many exciting voices, both young and old. Among the younger crowd, though, I sometimes see these negative comments towards PIs, usually along the lines of “All my PI cares about is getting her papers published and grants funded and not about training me with real skills that I need or helping me get ahead in my own career”. Essentially, some variant of “PIs only cares about themselves and not about me and the incentives in science are all wrong in this regard, etc.”

Hmm. As a PI now (still feels weird to say it) I think it's worth making a couple points. First, by far the majority of PIs that I've met care deeply about their trainees and want nothing but the best for them. They are acutely aware of their successes and failures and want to maximize the former and minimize the latter. As in any relationship, I think some degree of conflict/frustration with your PI is totally normal and is in fact productive. Indeed, I think one of the most wonderful things about being a PI is working with younger people who have a healthy disrespect for experience, including my own. That said, thinking you know better than your PI is one thing, thinking that they don’t care about you is something else.

Another criticism I often hear these days is something along the lines of “My PI didn't teach me anything, so why is it called graduate SCHOOL?” I think this gets to the heart of what graduate school is about. Is it developing tangible skills? Yes and no. Yes, you will learn some specific skills along the way. But going to graduate school for a PhD is decidedly not the same as going to a trade school. Or at least it is one of a wholly different sort. To me, the most important thing you learn in grad school is critical thinking, which is hard to quantify but is very, very real. I was just on a thesis committee of a student who is about to wrap up her PhD. In the last year, she did a ton of work, and made fantastic progress–which, by the way, everyone on the committee was happy to see. And it was also clear that she had changed as a person. Perhaps a bit weary, maybe a bit less idealistic. But at the same time, it was clear she had developed tremendously as a scientist–and, by extension, as a professional. She handled criticism with aplomb, projected authority on her research field, and presented her work clearly and effectively. In short, she is a now a trained, mature scientist. It is this aspect of training that is the hardest to objectively describe and so gets ignored the most, but is far more important than any particular skill in the long term, no matter what career path you choose to follow.

So if this is the point of grad school, then how does the PI help make this happen? Partly to provide an exciting and vibrant environment for doing science, and make no mistake, this is a lot of work–something I didn’t appreciate as much until I became a PI. And yes, partly to provide some technical scientific skills. (I actually think this includes the elusive notion of “creativity”, which is a skill that can be taught, in my opinion.) But another big part of it that is much harder for trainees to appreciate is that we bring experience. Experience in navigating through the ups, downs, twists and turns (the “cloud”) as you struggle through a research project, experience in how to deal with failure and rejection, experience in how to deal with success. We have learned from many mistakes, and the point of the human learning is to try not to repeat mistakes others have made before. At the edge of knowledge, where the path is by definition not clear, this experience is invaluable.

I think we also have something valuable to share with students just based on the fact that we’ve typically been alive a bit longer. I think this manifests itself most often in the common belief that “My PI is not letting me graduate because they just want to get another paper out of me.” I’m going to just speak for myself here, and I’m sure there are many counterexamples, but I don’t think I’ve met any PI who’s done this. In fact, far more common in my experience is the opposite thought for the PI, along the lines of “How am I going to graduate this student who just has not been productive?”

Where does this notion of the PI holding back the student come from? I think it’s the Luke Skywalker/Yoda dynamic. You know, when Luke Skywalker wants to face Darth Vader and Yoda says “Complete is not your training” (or whatever Yodaspeak, leave me alone Star Wars nerds), but Luke leaves anyway, gets his hand chopped off, and then returns, at which point Yoda says “Complete your training is”. Students are young, with that energy and the aforementioned healthy disrespect for inexperience that tells them “I’m ready to get out there and do my thing!” PIs have a bunch of experience that says “Wait, why be in such a rush? Life is long, there’s more to learn here, and it will serve you well.” You know what, I think they’re both right. At some point, the student is ready to get out there and develop their own experiences–they, like Luke Skywalker, will be faced with a new situation, rely on what they have learned, perhaps make some mistakes, and then their training will be complete. I have often found myself now encountering situations where I think to myself “Oh, now I see what my advisor was talking about…” I think it’s just the natural cycle of the mentor and the trainee.

Again, though, I think it’s important not to mistake this for the PI acting against the best interests of the trainee. I would like to think that when faced with a choice, I would choose to act in the best interests of the trainee, and I believe that to be the case for myself (although perhaps some might disagree :). To be frank, though, I don’t think these choices come up all that often anyway–most of the time, I think the interests of trainees and mentors align quite well. Yes, PIs have their own career to think about, and I’m not going to say that it's not a part of one’s thinking as a PI. But consider this: if the interests were so misaligned, why is it that junior PIs, who face the most career pressure and uncertainty, typically make the best mentors for graduate students?

Anyway, I guess my point is that for the most part, PIs care about their trainees, and usually a lot. Sometimes PIs make choices that may not initially seem to trainees to be in their best interest. And sometimes PIs make mistakes, or sometimes we all fall victim to just plain old bad luck. But, for the trainees out there, please don’t feel like your PI is out to get you, because they probably aren’t.