The CrowdFund Science Crowd Mistakes "An Experiment" for "Doing Science"

Oct 14 2014 Published by under Conduct of Science, CrowdFund

I had a revelation that clarified some of my points of poor understanding of the science crowdfunding enthusiast position.

In skirmishing on Twitter with some guy associated with "Experiment.com" I ran across a project on brain inflammatory responses in a fetal alcohol model from the Lindquist lab. Something I can readily assess, being that it is a substance abuse, drug-toxicity investigation in rats.

The investigators are seeking $6,000 for the following goals:

  1. Determine whether ibuprofen protects against the changes in locomotion (hyperactivity), emotion (elevated anxiety), and cognition (poor memory) typically seen in rats exposed to alcohol shortly after birth, modeling third trimester alcohol consumption.
  2. Determine whether ibuprofen is most efficacious when given during pregnancy or after birth. Thus, ibuprofen will be administered during the alcohol exposure period only or after the alcohol exposure period.
  3. Alcohol binding to TLR-4 "turns on" microglia which change their shape when in an activated or "inflamed" state. Thus, the level of neuroinflammation in particular brain regions will be quantified by counting the number of activated microglia in alcohol-exposed and control rats following ibuprofen or saline treatment.
  4. If additional funds are raised beyond our original goal: Investigate pro-inflammatory protein expression in alcohol-exposed rats across development.

The budget breakdown is simple:

  • Lab supplies $1,000
  • Alcohol supplies $300
  • Microglia immunohistochemistry $2,000
  • Subjects (purchase) $200
  • Subjects (housing) $2,500

The model involves drug treatment just after birth so I'm going to presume the rat purchases are for dams, possibly already pregnant. That lets us assume a decent-ish sample size given litters of 8-14 pups. [Sidebar: Is this a bizarre pet fancier/lab scientist difference to call rat offspring kittens instead of pups?] This assumption is critical because the project describes four treatment groups, if you read around the edges. The key comparison is within alcohol exposed rats, one group treated with ibuprofen and one not. But in the histology goal, you can see mention of "control" rats which are also to be treated with either saline or ibuprofen. So four groups of rats to be compared. I don't think $200 gets you four decent sized groups unless their vendors are much, much cheaper than mine (not impossible, they might breed them in house at a discount over traditional lab rat vendors).

Outcome measures are three behaviors and one post-mortem tissue marker of microglial activation.

I'm looking at this and thinking that if everything goes perfectly then maybe this is publishable in one of the semi-respectable dump journals. Maybe. Most likely reviewers are going to demand a lot more, yes, even at journals of fairly pedestrian profile. This is not a fantastically brand new idea.

And, of course, things rarely go perfectly in actual science. They might not land on exactly the right doses of alcohol or ibuprofen at first. Behaviors might be equivocal. Post-mortem inflammatory activation effect sizes might not be as big as originally estimated.

All of this scientific reality leads up to "more work". Of a similar scope. Which costs money. So even assuming that these investigators clear the bar on this appeal for crowdfunding, can they duplicate it to respond to the initial review of their manuscript?

So what is the goal of the "crowd funding"? It does not look to me like this particular project stands much chance of ending up as a published paper without substantial additional research being conducted. Maybe it will, but the odds do not seem high to me.

I think a seat of the pants understanding of what it takes to generate a publishable study underlies much of the skepticism some working scientists express in response to the outlandish claims of crowd funding advocates.

I think I am finally grasping that when the crowd funding advocate goes on about "science" they really mean "an experiment". I am starting to suspect that they really don't think beyond that to the stage that I would call "doing science". Which means, at a very minimal level, "generating a publishable paper's worth of experiments".


Returning to an older theme, note what is not included in their budget.
1) Overhead - is the Uni picking this up? There will be a limit
2) Staff salary - for PI and grad student. Where is the effort of the PI coming from?

Overhead is not just the University stealing from research funding. Look right at this project "All procedures in the current experiment will be done in consultation with Ohio State veterinarians and in accordance with the Institutional Animal Care and Use Committee (IACUC)." As well they should be but...who is paying for that? I don't see it listed in the project. This is what overhead on federal grants pays for, in part.

Alcohol supplies? Many institutes have chemical safety and/or environmental health and safety that is required here. Maybe it is simply built in to the pricing...or maybe those institutional services are paid for in part or whole by grant overhead.

Animal husbandry? Well, sure maybe that per diem is total cost recovery for all of the support services (including the vets and IACUC mentioned above) but usually this is not the case.

I see no rent for the lab space that will be required. Something is picking up the bill on that as well.

33 responses so far

  • I think a seat of the pants understanding of what it takes to generate a publishable study underlies much of the skepticism some working scientists express in response to the outlandish claims of crowd funding advocates.

    Certainly to some degree. But I've been around long enough to have seen new models of doing things go from things that people mocked as hopelessly naive to becoming accepted reality. Consider open source/GNU software. When I first heard of the GNU project in the late 1980s people openly mocked the idea as impractical dreams of ex-hippies like Richard Stallman.

    People used to argue that Sun's version of UNIX cost hundreds of millions of dollars to develop and that there would be no way for open source operating systems to compete. Today Sun doesn't even exist anymore and the compute clusters of nearly every university and research institute run on Linux (not to mention most routers, Android phones, etc.).

  • zb says:

    Your detailing two problems.

    First, is whether they will really be gathering sufficient data to produce a publishable result. I think this is a major problem with this kind of study, and, actually, it looks a bit like a "science project" study. That is, a study done to generate a presentation for a student, a Intel quality project, even. How many of even the winning Intel projects are published, as is? My guess is that the ones that are published are continued by a mentor/supervisor to the point that they become publishable in a journal. Most are published in a preliminary report format or published in the public press as a feel good human interest story. Just this morning, I stumbled on a science fair project done by two teens in Portland, on multi-tasking and performance in teens. They got a result in conflict with most studies (for a subset of individuals, high multi-taskers, doing simple attentional tasks while listening to music improved performance). A result that excites people, but, on a well-studied topic, and one that is prone to statistical issues. The article (in the Wall Street Journal) has the following quote "[They] said they have begun seeking a scientific journal interested in publishing their study." Good luck, huh? The study seems decently designed, but ultimately publishable? Not without a lot more work (and as you say for the rat work, the idea is not all that new).

    Your second issue, that there are hidden costs, is true. But, whether a university will allow them to do the work with hidden costs probably depends on whether there's a good argument that allowing the work is in a larger mission of the university (i.e. training/teaching, etc.). The work can be done by undergraduates (isn't a lot of it, in psychology, anyway, and aren't those students often not paid? even the grad students? who might get paid for teaching, and not for research?). The PI might be paid, as well, for teaching, and not need to support his salary. The time for this project comes out of his volunteer labor. And then, the question is whether other charges, say, tuition, can be reasonably attributed to paying for the marginal cost of IACUC oversight, lab space, etc.

  • zb says:

    You are, (though I hate comments to correct spelling/editing).

    The open source/blogging/etc. example just shows the trend towards unpaid labor driving out paid labor. Whether that will happen in science depends on whether people are willing to volunteer labor (quite possibly, they already do in many ways, by working extraordinary hours for fairly low pay) and resources. Resources are a bigger problem in science. In the case of unpaid programming, unpaid graphic design, unpaid photography, unpaid writing, a big part of the trend was the decreasing cost of the necessary resources. Photography, video, blogs, books became crowd-fundable when people started buying cameras, computers, . . . for themselves (in addition to volunteering their labor). Labs are harder to buy on the side, and, there are ethical considerations as well. I really, really don't want people doing experiments on rats in their basements (even if they are allowed to -- not entirely sure of what the regulations on rats are).

  • eeke says:

    The monkey is too excited for no reason. The PI does not propose to publish results from the experiment (not anywhere that I can see from a quick glance). Instead, this looks more like a small investment for an experiment that could serve as preliminary data for a larger grant application. I was invited to submit a proposal to this site, but declined. The university I work at seemed open to accepting funds from this site without overhead. I get the impression that successfully funded experiments such as this could be a good investment that might potentially lead to a better chance at funding. I've also seen requests through this site for supporting an undergrad for a summer position, or for smaller aspects of already-funded projects. Anything helps.

    The only fee that is required (instead of overhead to the univ.) is what gets paid back to the company - 5%.

  • zb says:

    "The PI does not propose to publish results from the experiment (not anywhere that I can see from a quick glance)."

    But then, they're overselling on the experiment.com site, which reads as though they are actually going to get an answer (to me, as a first step, that means a publishable result). I agree that that is what they're doing -- overselling, rather than presuming they can publish with $6K. Intriguingly, as I read the site, I had the thought run through my mind, that if I could get all that for 6K, I could find the money. The crowdfunding is a marketing tool.

    In some ways, the issues you are raising are the same as the crowdfunding donations at Donors choose, for schools. There too, people are being asked to fund marginal costs, while other sources fund the major costs (of labor and facilities). There's less overselling, though, since the teachers don't promise definitive answers/results for the trumpets/carpets/art supplies people donate.

  • theLaplaceDemon says:

    Yeah, I think it depends on what the PRIMARY goal is here. If the project is, say, an undergraduate summer project at a primarily undergraduate institution - where the primary goal is training the undergrad and actually obtaining publishable results is secondary - I can see this making sense. But if that's the case, the proposal should be a little bit more upfront about it.

  • becca says:

    How much is your blog kicking back to your institution for overhead to pay for internet bandwidth?

    How much is the university kicking back to experiment.com for good PR/advertising?

    When I'm Empress of everything, I'm fixing Amazon's wage theft of not paying for people's time when they are screened for theft way before I'm getting around to fretting about these kinds of things.

    BTW- this is Not the Lindquist lab I thought of initially. Makes me feel better, really (the other Lindquist is a bucket of awesome). I mean, TLR4 and alcohol?? My PhD involved TLRs binding to some funky things, but I've never heard of that (NB: I would have no trouble with ethanol -> upregulation of TLR4, which then responds to endogenous ligands of unknown identity... it's direct binding that sounds funny to me- if anyone has any info on this I'd appreciate it).

  • Davis Sharp says:

    Medical schools and universities regularly raise money from various sources (rich dead people, alumni, and civic organizations). This isn't all that different except that the funds are earmarked for an experiment. Maybe the PIs will ask for more $ to reproduce the experiment, do follow-up experiments, and publish. But how does one report this on the Active Support section of the biosketch and what would reviewers think of that?

  • drugmonkey says:

    The monkey is too excited for no reason. The PI does not propose to publish results from the experiment

    The target of these comments is not "the PI". The target of these observations is the corpus of outlandish claims made by the CrowdFundScienceEleventy waccaloon types.

  • dsks says:

    $6000 for that?

    Pfft.

    I've got an experiment that could cure cancer* for a third of the cost. I'm headed to kickstarter.

    * in vitro

  • Dave says:

    TLR4 is hipster. Would not fund.

  • becca says:

    TLR4 was hipster before there was an internet. Nowadays, TLR8 is hipster.

  • Comradde PhysioProffe says:

    This shittio is basically 10 year old in the basement with a chemistry set pretending to be a scientist. We spend $3000 per fucken *day* just to run a modest size lab with good productivity.

  • K99er says:

    I realize that perhaps you chose this proposal at random and that you are mainly trying to criticize crowdsourcing as a viable model for funding real science. But, I think this particular proposal and this particular PI is undeserving of the scrutiny that being featured on your blog brings down upon him. Looking at this guy's pubmed listing and google scholar page, it's pretty clear that he's a new assistant professor. He therefore surely has a startup package that can supplement this project, and he probably has some leeway for a few years with regards to having grants with indirect costs. So really, he's asking for an amount of money that would help HIM complete these three experiments. This may be very different from what someone else in a different situation would need.

  • drugmonkey says:

    But, I think this particular proposal and this particular PI is undeserving of the scrutiny that being featured on your blog brings down upon him.

    actually I'm hoping some of you donate! good stuff.

    So really, he's asking for an amount of money that would help

    emphasis added. kind of the point.

    This may be very different from what someone else in a different situation would need.

    Not really. Experiments are experiments and papers are papers. Last I checked, nobody looks at your funding when assessing your paper in review.

    just to run a modest size lab

    I suspect there is very little about you that is "modest".

  • anonymous postdoc (shrewshrew) says:

    Two things:

    1) Pre-publication, pre-experiment peer review! Ibuprofen is not a great choice to use in a gestational model to address anti-inflammatory mechanisms. This is because in pregnancy women are already strongly advised against using ibuprofen and naproxen, because of a suggestive literature indicating birth defects including cardiac defects and facial clefts (http://www.sciencedirect.com/science/article/pii/S089062380100137X) and a general risk of reduced uterine perfusion due to vascular constriction which is an unavoidable side effect of most NSAIDs (http://informahealthcare.com/doi/abs/10.1517/14656566.5.3.571).
    The researchers would do well to read the literature describing intrauterine growth restriction as a result of reduced uterine and placental blood flow, as this exerts its own whammy effects on brain development. Perhaps they don't care, but it increases the likelihood that their NSAID-treated control dams will produce some fucked up offspring. However, I have seen interesting proposals suggesting aspirin as a potential pregnancy anti-inflammatory, because its blood thinning effects may counterbalance the vascular constriction. This may be a better choice for preliminary experiments, if that is indeed what the proposed studies are. If the studies are to try to develop a translational approach to treating FAS, which is what they say in their blurb, then they definitely would want to consider a different NSAID.
    Also, what TLR4 receptors are these people purporting to target? Neuronal? Placental? Just, you know, "inflammation", man? TLR4 has been shown to mediate CNS inflammation caused by alcohol, but I as I read it, there is no evidence that the teratogenic effects of alcohol are inflammation based. If this proposal is seeking funding to directly to test that hypothesis, fine. I'm okay with that. But they should say so, instead of insinuating that they are going to test a valid intervention, when their intervention is controversial at best.

    2) Crowdfunders have no fucken ability to provide any fucken peer review whatso-fucken-ever. This is a problem. Fucken.

  • drugmonkey says:

    I wasn't all that clear (the proposal is) but for rats they are arguing that the postnatal interval really maps onto human third trimester so the ibuprofen will be in neonate rat pups. So placental blood flow isn't in the equation for this model.

    translational recommendations are another issue, of course.

  • Ola says:

    Waaaaaay off the mark.

    Back in the day, when the first Affy arrays started becoming available, and people started freaking out about doing "a single experiment that costs $15k", we did a breakdown of our last half dozen papers and figured out they'd cost about $60k each, factoring in salaries, animals etc. $15k didn't seem so bad then, especially because it was the Wild Wild Ouest and you could just do a bunch of arrays and fart the data into a fancy diagram and call it a paper. Nowadays I would not expect a CNS paper to be had for less than $100k when all is said and done, and there are several labs who run a single glam mag paper a year on an entire R01 mod budget, putting those papers at a cool quater mil a piece. This gets scary when you pick up a print copy of a journal - your average IF4-5 rag probably has >$1m per edition in science, while the average edition of CNS (with 83 supplemental figures per paper) runs into the tens of millions. $6k just about pays for everyone in the lab to get fancy coffee (not pumpkin spice!)

    Of course, the advantage (for the sciencers) is the whole deal of crowd funding's raison d'être - if it doesn't work out, oh well, no problem, onto the next project, angry investors should have read the small print. Suckers! Given the delights of NIH's new progress report system, such carefree reporting requirements are looking very attractive from this side of the fence.

  • Comradde PhysioProffe says:

    "Nowadays I would not expect a CNS paper to be had for less than $100k when all is said and done"

    Accounting for both direct and indirect costs, and figuring three years of work: one lead author $300,000 salary support, one PI at 20 percent effort $250,000, 20 percent effort of figure five other contributors is another $300,000. So we're already at $850,000 and still haven't paid for animals, supplies, reagents, or the amortized expense of equipment. I would say that your average multidisciplinary CNS paper with genetics, molecular and cellular physiology, and organismal physiology represents at least a million dollar investment.

  • drugmonkey says:

    ....to be published behind a paywall.

  • Josh VW says:

    CPP: I've been talking with my students lately about how much science costs per paper (a scary thought for me as I burn through my startup). We came to a rough figure, for a JACS-level work in an already well established lab where the big-ticket hardware is already paid for, of at least $100K. This is for experimental physical chemistry, where instrument-building could easily increase that to $1m to get the first paper out of an instrument, then $300K for the next couple papers until it's running smoothly. This is direct+indirect.

    Does this estimate sound right to others?

  • Molly says:

    "The model involves drug treatment just after birth so I'm going to presume the rat purchases are for dams, possibly already pregnant."

    Each dam can bear 3 litters of up to 12 pups each. The experiment requires 2 (ethanol exposed & control) x 2 (ibuprofen & vehicle)— 8-10 subjects/per group is standard for behavioral neuroscience experiments, meaning that the experiment would need 32-40 total rats. This could easily be managed by 4-5 dams who can then be used later for additional experiments.

    "I'm looking at this and thinking that if everything goes perfectly then maybe this is publishable in one of the semi-respectable dump journals. Maybe. Most likely reviewers are going to demand a lot more, yes, even at journals of fairly pedestrian profile."

    The primary goal of this experiment is to gather data for a larger RO1 grant. It may be published as a brief report but it will likely be rolled into a larger paper later on.

    "This is not a fantastically brand new idea."

    No, the idea of inflammation in FASD is not novel. However, it is not well-characterized, particularly in how it contributes to cognition. Several studies have been published using different models (earlier gestational periods, different modes of alcohol administration) and looking at other deficits (motor, structural) but there is not sufficient work on cognition- learning and memory, in particular, is lacking any solid research. Furthermore, 'fantastically brand new ideas' rarely get funded. You need solid pilot data and other research to back it up.

    "They might not land on exactly the right doses of alcohol or ibuprofen at first. Behaviors might be equivocal. Post-mortem inflammatory activation effect sizes might not be as big as originally estimated."

    The dose of alcohol is a standard dose that has been validated as being neurotoxic by the Lindquist lab and others (see Pamela Hunt, Mark Stanton, etc.) The dose of ibuprofen is one that has been used in other research on neuroinflammation in neonates, particularly in hypoxia. Pilot data with 5 animals/treatment group is nearing statistical significance (see experiment.com page)

    "Returning to an older theme, note what is not included in their budget.
    1) Overhead - is the Uni picking this up? There will be a limit
    2) Staff salary - for PI and grad student. Where is the effort of the PI coming from?"

    Overhead (lab space, animal care, etc.) is being covered by other grants, the PI start-up fund, and department. The PI has a teaching salary and the grad student has a 50% GAA position that covers her tuition and stipend. Undergraduates working in the lab do so for course credit.

    "The crowdfunding is a marketing tool."

    In a way, yes, it is helpful to raise awareness for the types of research being done as well as for FASD itself which is an understudied condition, despite affecting up to 5% of American kids and up to 30% of children in certain South African communities. However, it is also an excellent way for new labs to supplement start-up funds to gather enough pilot data for larger grants.

    "I mean, TLR4 and alcohol?? My PhD involved TLRs binding to some funky things, but I've never heard of that (NB: I would have no trouble with ethanol -> upregulation of TLR4, which then responds to endogenous ligands of unknown identity... it's direct binding that sounds funny to me- if anyone has any info on this I'd appreciate it)."

    Sure- while it is unknown if alcohol is physically binding to TLR4, it has been shown to interact with it in such a way that activates inflammatory cascades. See: Alfonso-Loeches et al., 2010 (Journal of Neuroscience), Drew & Kane, 2012 (Neural-Immune Interactions in Brain Function and Alcohol Related Disorders, Chapter 11), Fernandez-Lizarbe et al., 2013 (Journal of Neurochemistry), Pascual et al., 2011 (Brain, Behavior & Immunity), Pascual et al., 2011 (European Journal of Neuroscience), Pascual et al, 2014 (Alcohol & Alcoholism).

    "TLR4 is hipster. Would not fund."

    TLR4 is totally mainstream.

    "This shittio is basically 10 year old in the basement with a chemistry set pretending to be a scientist. We spend $3000 per fucken *day* just to run a modest size lab with good productivity."

    Perhaps you are inefficient. The crowdfunding site is just to raise money for a specific experiment- not to fund the day-to-day lab activities.

    "Ibuprofen is not a great choice to use in a gestational model to address anti-inflammatory mechanisms"

    The experiment is postnatal- alcohol is given during a 3rd trimester-equivalent period and ibuprofen is given in a period considered to be equivalent to early postnatal life in humans. Ibuprofen and indomethacin are used regularly for the treatment of patent ductus arteriosus in newborns and has been found to be safe. The goal of this experiment is not to make it safe for women to drink during pregnancy. Rather, it is to investigate a different use for an already existing drug for babies born who have been exposed to alcohol during gestation.

    "Also, what TLR4 receptors are these people purporting to target? Neuronal? Placental? Just, you know, "inflammation", man? TLR4 has been shown to mediate CNS inflammation caused by alcohol, but I as I read it, there is no evidence that the teratogenic effects of alcohol are inflammation based."

    Ibuprofen is administered systemically but crosses the blood-brain barrier so it will target Cox-2 in the periphery and CNS. See Drew & Kane, 2012 (Neural-Immune Interactions in Brain Function and Alcohol Related Disorders, Chapter 11), Tiwari & Chopra, 2011 (Journal of Neurochemistry), and Randall et al., 1989 (Annals of New York Academy of Sciences. Also see pilot data on the experiment.com site.

    "…translational recommendations are another issue, of course."

    Yes, translation is always an issue with rodents. However, non-human primates are usually not approved for use unless one can show that ‘lesser species’ are not sufficient. Research has shown good concordance between animal models (i.e. rodents) and humans with FASD.

    The experimenters would be happy to answer any questions or respond to any concerns that potential backers may have. However, posting them on an unrelated site that is frequented by a post-doc in a neighboring lab is not the best way to contact them. Messaging is available through the experiment.com site or you can always email them directly. Please consider supporting this small research project to open the door for more studies on treatments for FASD!

  • drugmonkey says:

    Molly- thanks for stopping by. My apologies for highlighting your project in service of a longstanding argument I have with the CrowdFunding nutjobs...but thanks for underlining my essential points. Best of luck with your project.

  • becca says:

    That was an excellent answer to my question. *throws money at experiment.com for Molly*

  • Josh: "Does this estimate sound right to others?"

    I've been told through unofficial channels that the Navy/DARPA expects a Physical Reviews X (or a PRL) per $100K over 2 to 3 years. Seems pretty reasonable overall, since a lot of that is salary for a few postdocs within that time span; and buying time from a PI and computer time etc. (I'm currently in a computational materials science group that mainly employs quantum physics PhDs. Since I'm just a physical biochemist, I'm writing the manual. Thank goodness I learned LaTeX.)

  • neuropop says:

    @Josh, @PP: Even a purely theory/modeling paper in semi-fancy pants journal (read offspring of CNS) is: 1 lead author (say at 25% effort): 50K + 1 postdoc (at 100%): $100K = $150K. Even with no amortized equipment costs. Not to mention the indirects. In the days of yore, one could do a long calculation, summarize the results in a letter (PRL) and publish the details in a PRX format. The CNS offspring now demand, for the sake of completeness, all details in the supplements. All this effort for one publishable unit.

    So, no, science doesn't come cheap.

  • Molly says:

    DrugMonkey- Thanks. The point of your post is correct- biomedical research is very expensive and crowdfunding is not going to fix the funding gap. However, it can have its place (i.e. pilot data, social science research, etc.) and I hope it can continue to help small and/or new labs get the money that they need to get things working.

    No worries on the use of my project to make your point- it's a good one and I got a pledge out of it (thanks becca!).

  • DrugMonkey says:

    I hope it can continue to help small and/or new labs get the money that they need to get things working.

    My concern is the thought on the part of the NIH, grant reviewers, CongressCritters, small foundations, etc, that the availability of crowd funding means that they can dodge out on their respective responsibilities to try to get these small and/or new labs the money they need. Or that it is anything other than a hit and miss, unrepeatable curiosity.

    I suppose I would also like to see some more consideration of the expectations of the donors and the ability to deliver on the part of the people seeking funding. Your project has a participant track record of similar publication and a clear plan (as you nicely mention above) for how the additional work will be covered. From my perspective, these things make a donation more likely.

    What I don't know is whether the average donor cares, given they are likely to be 1) friends and family and 2) committed CrowdFund nutters and therefore the actual results may not be the point of their donation.

  • jmz4 says:

    "However, it [crowd funding] is also an excellent way for new labs to supplement start-up funds to gather enough pilot data for larger grants."
    -This seems like a pretty reasonable statement to me. DM, do you have any links to the people claiming that crowd-funding is the wave of the future for science (cause it probably isn't).
    Now if there was something like this for oligarchs with too much disposable income and guilty consciousnesses, then it might actually be able to fund full research programs.

  • JT says:

    I have perused the Experiment.com site and have a few comments.

    First, those saying $6K isn't sufficient to fully fund the project are probably correct. But look at the website... those with lower funding goals are the ones that get funded! The PI presumably reduced costs to the bear minimum in order to increase their odds of being funding (though it doesn't look like they will).

    Second, critiques that the study is not detailed enough are ludicrous. Again, look at the site. Space is limited and the experiments are intended to be understood by an intelligent layperson, not another scientist. The projects obviously do not, and cannot, contain the same level of detail that, say, a NIH grant proposal would contain.

    Third, I have found not projects that specifically state what will come of the project. Most probably will never end up published, as acknowledged by Molly above. This is an issue with Experiment.com, however, and one the biggest concerns as I see it. Some backers may expect to the see the project written up and submitted, others may not even know what is "normal" for an experimental project. Clearer and better communication (on the part of the website) is required to ensure that PIs and backers know what is expected....

    Fourth, DM used this project as an example. All those jumping on the PI are dumbasses in need of validation or trying to impress. It's one project among thousands, most of which are backed by the major universities.

  • drugmonkey says:

    How can you possibly have missed Perlstein's blather before he settled down into a standard issue biotech startup scrambling for VC backing?

  • jmz4 says:

    The "rogue scientist"? I thought his model was more hitting up parents of sick kids for cash and renting out lab space to other independent scientists (who would somehow have money).

  • Patrik D'haeseleer says:

    "Or that it is anything other than a hit and miss, unrepeatable curiosity. "

    That seems like a much more apt description for NIH funding these days. Excepting federal funding has become a very stochastic process. You can greatly increase your chance by writing 10 proposals a year, but who wants to spend all of their time proposal writing?

    Crowdfunding is not going to bring in the big budgets that you need for a serious research program, but to raise a small amount of money to produce some preliminary data, I think it is actually a much more reliable route. How much money you bring in is almost directly related to how much effort you put into the campaign.

Leave a Reply