Bunny Hopping
December 4, 2007
A recent comment from whimple is pretty self explanatory:
Say I work on the mechanics of bunny-hopping. My papers get sent for review to colleague bunny-hoppers, my grants are reviewed by the bunny-hopping study section, and there is never really an opportunity (ESPECIALLY with the study section) for a non-bunny-hopper to stand up and say, “look, other than the bunny-hoppers, nobody really cares about bunny-hopping, and I think we already know all we need to know about bunny-hopping for now,” and close down the field.
I’ve been thinking about this a bit more. Too lazy to find the link but I recall Physioprof busting on human functional Magnetic Resonance Imaging studies, this might have been at someone else’s blog, I don’t recall. I’ve been known to take the occasional potshot at knockout transcription factor de jour studies.
The question of the day, Dear Reader, is what type of science could you really do without? Or at least cut way back on? What really ticks you off that “your” next R01 is being wasted on such obvious crap? C’mon, fess up.
I was also thinking about a neat little exercise for the home reader so if you like this one better go nutz. Take bunny-hopping and go with it. Assume you are submitting your next proposal to a study section with a large number of bunny-hopper scientists. Write up some Aims of interest to you that relate to bunny-hopping. Maybe toss out some experiments. This is brainstorming mode (think Program Project!) so just throw out some ideas they don’t have to be fully formed. Dang if we had a few more scientists blogging this would make a decent meme… Anyway, an example:
Aim I: To determine if long distance bunny hopping results in increased liability to abuse drugs.
In this Aim bunny hopping will be employed as a model of compulsive exercise akin to a proposed “running addiction” phenotype in humans. The goals are to determine if a compulsive hopping paradigm established in the laboratory disrupts general reward pathways leaving animals at increased liability to develop compulsive drug taking behavior.
DrugMonkey 
December 4, 2007 at 5:52 am
That bunny-hopping Aim reads so good that it’s scary.
LikeLike
December 4, 2007 at 6:55 am
What’s odd about the original critique is that the opposite is often true (especially for predoctoral and postdoctoral F awards with less specialized committees). There’s a good chance at least one person reviewing a proposal will have minimal interest in a topic and will either blandly evaluate it or try to sink the proposal.
LikeLike
December 4, 2007 at 8:57 am
That’s the flip-side of the bunny-hopping: decreased enthusiasm for non-bunny-hoppers. Combined with increased enthusiasm for the other bunny-hoppers, it can be a potent force in scientific review.
LikeLike
December 4, 2007 at 9:07 am
Okay, whimple, sounds like bitter experience talkin’ here :-). Any chance of getting a leeeetle more specific? is it the case that you need to seek out a different study section entirely?
Or is this a larger scientific issue where you think everyone is spiraling into navel inspection and missing the real point? is this related to comments over at MWE&G about human-focused research, for example? If this is the issue, well, prepare yourself for a long slog…
LikeLike
December 4, 2007 at 9:19 am
Aim 2: Does competition invariably increase the desire to seek performance enhancement?
For these experiments, short-sprint bunny hopping races will be deployed as a model for competitive sport competition in humans. Bunnies will participate in races daily, with the “winner” receiving a reward. Back in the home cages, some bunnies will be given access to either anabolic steroids (the rabbit “clear”), muscle-building protein supplements, or simple vitamins. Do rabbits increase their intake of these performance enhancers with increased success or increased failure?
– Sub-Aim: does a lack of performance by bunnies lead to an increase in anxiety and depression surrounding the daily races (as measured by other behavioral assays and stress hormone levels) and ultimately result in an increased intake of antidepressants, also made available in the home cages?
LikeLike
December 4, 2007 at 9:47 am
Nice one Noah. Mine will fit in the same component with yours, likely.
Aim I: To determine the relative efficacy of training and doping techniques to enhance sustained aerobic output.
Experiments will be conducted using the bunny-hopping model to examine the relative impact of hypobaric sleep, high-altitude training, blood doping and EPO on aerobic athletic performance. Groups of bunnies will be trained on laboratory models of the 10m-to-hedgerow sprint, 5acre-open-field middle distance and escape-the-dogpack long distance events. Groups will then be subdivided and exposed to control or each of 5 experimental treatment conditions. A mixed within/between analysis design will be employed to determine if the experimental treatments differentially improve hopping performance in each of three events. Implanted radiotelemetry will be employed to measure resting, event and post-event recovery of heart rate.
LikeLike
December 4, 2007 at 11:24 am
You guys are losing me with all this bunny-hopping crap. Anyway, the more important issue in terms of allocation of funds between fields is not the composition of study sections and the inclinations of members of study sections, but the constitution, dissolution, and scission of study sections.
That is, what study sections exist? By creating a study section focused on a particular area of inquiry, CSR automatically guarantees that the top X%ile of grants in that area get funded. If you have a “Cellular and Molecular Bunny-Hopping” study section, more bunny-hopping grants are going to get funded than if you only have a “Cellular and Molecular Bunny-Hopping and Bunny-Galloping” study section, because now the bunny-hopping proposals are also competing against bunny-galloping proposals.
LikeLike
December 4, 2007 at 11:33 am
“By creating a study section focused on a particular area of inquiry, CSR automatically guarantees that the top X%ile of grants in that area get funded.”
Yeah, this is why we could use a little more in the way of specifics from whimple. Is it study section fit? Is it being so out there that there is no study section just for you? or, more specifically, which study sections would you cancel so as to eliminate their special pipeline 🙂
LikeLike
December 4, 2007 at 11:40 am
Down with bunny-hopping!!
LikeLike
December 4, 2007 at 12:07 pm
You are just jealous that the Hopping Bunny Neuroscience or Biology (HoBNoB) Society Annual Meeting is held in Maui.
LikeLike
December 4, 2007 at 6:15 pm
You guys know you’re going to be outed when your bunny hopping proposals arrive in study sections with your real names attached :-).
Is nobody willing to state their choice of “bunny-hopping” fields even anonymously?
I have one, but I won’t be first. I think if we really did it, what we’d find is that there’d be strong defenders for each of the bunnies even in this little group of commenters.
LikeLike
December 4, 2007 at 8:01 pm
Very nervous about stating this, even anonymously.
LikeLike
December 4, 2007 at 8:16 pm
electroacupuncture
LikeLike
December 5, 2007 at 7:23 am
Must combine with neurophysiologists favorite topic.
Aim 3: To functionally localize the regions of the brain that response to compulsive long-distance bunny-hopping
We will use a novel bunny virtual reality system (bsci et. al. 2007) to allow bunnies to simulate hopping within an MRI. By localizing the specific brain regions involved using fMRI, we will be able to better design, test, and target drugs to reduce the compulsive hopping behavior
LikeLike
December 5, 2007 at 8:07 am
“Aim 3: To functionally localize the regions of the brain that response to compulsive long-distance bunny-hopping[.]”
I feel sorry for any PI whose functional imaging grant or manuscript gets reviewed by me. Of course, that will never happen. It appears that even on the journal side, there are functional imaging quotas at the top neuroscience journals.
LikeLike
December 5, 2007 at 8:44 am
As an actual functional imaging person, I recognize and share the annoyance people have. The neuroscience stuff that gets the most publicity isn’t just fMRI, it’s often the worst or silliest of fMRI. Still, there is quite a bit of quality fMRI work and I do hope you don’t discriminate against grades based solely on the technology used. 🙂
You’ll notice that even in the bunny-hopping example, I make clear that the goal is to provide a target for more detailed studies using additional methods.
LikeLike
December 5, 2007 at 10:35 am
So bsci, I assume that you would concur with me regarding the irresponsibility and ridiculousness surrounding this?
I feel that we share the same attitude with regards to fMRI, although admittedly, I have never conducted research utilizing the technology. As with all techniques, one must be careful with how far the interpretation goes beyond the scope and resolution of the methods.
“Quota” is a strong word, but certainly cognitive neuroscience and fMRI studies must be represented at all publication levels. Let’s just leave it at that for my opinion.
LikeLike
December 5, 2007 at 11:05 am
In considering my own challenge, I come to a few conclusions.
1) While I may think I have a “Why do we fund that crap area” commitment in fact I do not. What I mean is that I find a pretty broad diversity of science interesting and can find “my” tack on many areas with a little reflection. I often can see where the most basic of science fits into not just public health but into public health that I find interesting or important. Therefore, it is always a matter of degree. That I wish particular approaches or techniques had a smaller share of the NIH pie or that fewer people were “wasting” their time on X when I think they should be working at X’ or X” or even Y.
2) Some of my top candidates for “toss it all from the NIH” are things like comparative cognition, the college sophomore subject variety of cognitive psychology, Olde Skoole behaviorism (yeah they still get NIH $$!) and single unit recording. WTF? This is stuff that I actually like and was trained on! The point being that I hate the execution (comp cog), refusal to consider real application (behaviorism) or have lost confidence that the “basic science” technique selected is the right one for good progress (single unit). So what I conclude is that if they’d “just do it my way” I would find it valuable. Once one reaches this conclusion one may be invited to STFU with the “independent investigator initiated” stick as reinforcer.
3) I have a somewhat old fashioned and universalized view of what constitutes “good science”. I have an aversion to the deification of styles and techniques that fly so firmly in the face of my understanding of good science. I am trying to understand, really I am. Why demonstrating something works once is a finding, as opposed to evaluating the statistical chances that it works each and every time, i.e., is “true”. Why “fold change” is somehow biologically meaningful where “reliability” of observed changes is ignored.
4) NCCAM is not even low hanging fruit. It has dropped off the tree, rolled down the hill and into your cave. Too easy.
5) Boondoggles. More a NIH structural thing than any area of science in particular. Although dedication of major new initiatives, Congressional mandates, Centers or what have you to a given area, strategy or technique can really point the finger at a specific scientific domain, this is not really the root problem.
6) Immediate early genes and transcriptions factors in behavioral neuroscience. Not an area, not a technique but a set of techniques applied to a given area at this stage of understanding. Take animal. Expose to drug of abuse, learning sessions, etc. Note that expression of c-fos or CREB or some crap has gone up in some brain area. whoo-hoo. start tracking down some internal signalling pathways. on the other wing start throwing different drugs or behavioral tests at ’em. look for “consistencies”. argue that you have “the drug abuse pathway” or “working memory pathway” under investigation. ignore fact that rodents barely have a frontal cortex, locomotor responses to stimulants (well gee, we needed a rapid screen you see) have fuckall to do with human drug abuse, ignore fact that your mouse is basically a liver wrapped in fur because “well, we can only do the genetic KOs in mice so that’s why we’re searching under this street light instead of up the block where we actually lost our keys, officer”.
LikeLike
December 5, 2007 at 11:29 am
“As with all techniques, one must be careful with how far the interpretation goes beyond the scope and resolution of the methods.”
Indeed. The trouble is that we must apply whimple’s “bunny hoppers” critique to even our own sacred cows and Emperor’s clothes. People have a tendency to get in a circle patting each other on the back about the superiority and application of their own styles and techniques. Oft reinforced by a circular logic of the GlamorMags and Impact Factors. Go read Chad and Janet on where lies the interface between physics and chemistry, think hard about levels of analysis and then go ahead and reflect on the way “does/does not demonstrate mechanism” is used in current bioscience to elevate the “quality” of one paper over another. can you really listen to one of these chumps in molecular biology bray about “mechanism” again without saying, “let me check with my physics buddy on that one”? It is all just levels of analysis. Which makes it preference, not OneGoldenTruth.
LikeLike
December 5, 2007 at 11:51 am
Noah, I absolutely agree with you regarding that op-ed. In fact, after stewing on it for a few days, I wrote a letter to the public editor asking what fact checking the opinion section does on any pieces, when opinion writers need to state a conflict of interest, and whether any of the editors had conflicts of interests. I haven’t seen anything back from Mr. Hoyt yet, but perhaps he needs more letters on this topic.
As for fMRI resolution, the method has it’s problems, but someone else’s standard response to physiologists is why are measurements from 20 semi-random neurons in 2 monkeys a respected publication, but a relative measure across the whole brain from 15+ volunteers is irrelevant. One of the more interesting developments I’ve seen is that now that people have been comparing fMRI directly to LFPs, spiking, etc and we’re seeing much of the neural information isn’t in spikes, there’s an interested interest in some of these other measures in physiology.
LikeLike
December 5, 2007 at 12:04 pm
“why are measurements from 20 semi-random neurons in 2 monkeys a respected publication”
dude. ’cause it is monkeys. You know, genetically closer to humans, have an actual cortex, blah blah. oh and they are “expensive” and “hard” to work with doncha know? and, c’mon. single cell firing! reductionist = good, brah…
LikeLike
December 5, 2007 at 2:19 pm
DM sez: “expression of c-fos or CREB or some crap has gone up in some brain area”
do I ever have the paper for you!
LikeLike
December 5, 2007 at 3:14 pm
[…] resorting to the video-cast to understand it. It seems like this is, in part, trying to address the bunny-hopping issue raised by whimple. Posted by drugmonkey Filed in Grantsmanship, NIH Careerism, NIH […]
LikeLike
December 6, 2007 at 12:27 pm
“DM sez: ‘expression of c-fos or CREB or some crap has gone up in some brain area’
do I ever have the paper for you!”
Do I ever have the Nobel Prize for you!
LikeLike
December 6, 2007 at 3:33 pm
Comparing mechanisms in physics and comparing mechanisms in biology are really apples and oranges. Even biological mechanisms have different meanings depending on the specific field of study. I am mulling over the topic of “different standards” for different fields, even in my small world of neuroscience. But as you can imagine, writing on that in my position can be a little awkward…just the challenge I need. (maybe that will be the post that finally gets me in trouble with the bosses at NPG).
I still like single-unit work. There are many questions to be asked and many techniques that can adequately address the question being asked, tetrode recordings and the like included. That is why I commented on interpretation.
For the record, I am insanely skeptical about immediate-early gene expression telling us anything with regards to underlying neuronal activity.
LikeLike
December 7, 2007 at 7:49 am
I don’t think there’s anything wrong with different standards for different fields even within neuroscience. The core standard is that the work must be interesting and contribute to our understanding of neuroscience – hopefully beyond just the field.
One of my neuroscience teachers said everyone gravitates to their maximum level of uncertainty. People who really hate uncertainty and want to control every factor of an experiment are studying proteins in neurons. Up the scale is channels, neuron-firing, systems, … human cognition. It’s impossible for a human cognitiion study to control as many factors as a protein structure study, but the best human cognition studies are the ones that might inform and guild research on the cellular and sub-cellular level. Whether those are the ones that get published in NatNeuro is an entirely separate question. 🙂
LikeLike
December 7, 2007 at 8:00 am
Actually, I kind of like thinking about subcellular receptor trafficking researchers as being control freaks, and TMS-using, scalp EEG-measuring cog scientists being relaxed, simple accepting their resolution-challenged fate…
LikeLike
December 7, 2007 at 11:22 am
[…] and maintain themselves and their peers through the grant review process. We’re using “bunny hopping” thanks to whimple and the NIH CSR calls this “clustering“. Note upfront that […]
LikeLike
December 7, 2007 at 3:17 pm
[…] 7, 2007 The DM recently used the pejorative term “boondoggle” (Wikipedia, fwiw) to refer to the larger NIH […]
LikeLike