Questions...

[Gankfest]

I got some Tin Foil hat questions about science research funding and Ethics...

 

When it comes to scientific research from federal and private institutions around the world.

 

How much do politicians, spooky people like George Soros, and Evil villains influence the funding and research of these institutions?

 

What's stopping scientists from cooking the books just to keep their jobs...?

 

Is it possible to cover up the facts I learned about global warming recently here?

 

I'm just curious because my friends thinks this, and personally how are you going to argue that... Anyway it really got me thinking what the integrity of the scientific method used today, and how the funding is spread around? I couldn't really find anything in depth on the interworking of the science community and the proccess of coming to a consensus. A video on Youtube is good enough.

 

Also what's this fallacy called?

 

If 1 person is lying, than everyone is lying.

 

If the FDA lies about a drug; than the CDC is lying too.

 

Than You!

[swansont]

I got some Tin Foil hat questions about science research funding and Ethics...

 

When it comes to scientific research from federal and private institutions around the world.

 

How much do politicians, spooky people like George Soros, and Evil villains influence the funding and research of these institutions?

Government funded? Not much. Currently much of the funding glows through nonpartisan entities, though AFAIK programs of record are approved directly by congress and are more open to tinkering.

 

The current threat is from the opposite side of the aisle than those supported by Soros. It's the GOP that is attempting to put political restrictions on funding.

 

 

Private foundations and individuals can (and do) directly fund anyone they want to.

 

What's stopping scientists from cooking the books just to keep their jobs…?

Other scientists doing similar research, checking up on them, and funding based on quality rather than getting a predetermined answer. Doing poor research is a good way to keep yourself from getting more funding from the government; owing to tight budgets, the competition is pretty fierce.

 

If you're funded by someone who wants a particular result, there is obviously a conflict of interest.

[CharonY]

Also note that publications are essentially open records. Eventually faults will be revealed if it is not possible to replicate results or if subsequent experiments fail to substantiate predictions. A common misconception is also that scientists are somehow all work together to keep their funding. The reality is that all are competing for the same money and the infighting between scientists can be vicious. There is no formal way to build a consensus but if many scientists independently and using different approaches come to same results, it will be used a consensus until a better explanation (e.g. model with better predictive powers) comes a long.

Edited February 18, 2014 by CharonY

[Greg H.]

Also note that publications are essentially open records. Eventually faults will be revealed if it is not possible to replicate results or if subsequent experiments fail to substantiate predictions. A common misconception is also that scientists are somehow all work together to keep their funding. The reality is that all are competing for the same money and the infighting between scientists can be vicious. There is no formal way to build a consensus but if many scientist independently and using different approaches come to same results, it will be used a consensus until a better explanation (e.g. model with better predictive powers) comes a long.

It never fails to amaze me how many psuedo scietific types seem to hand wave away the competition among scientists, as if they all have a vested interest in propping up some false system. They seem to forget that discoveries that were to, say, replace GR with something better are career makers for scientists. They're not trying to prop anything up - on the contrary, scientists work very hard trying to expand, refine, or even replace established theories because that's what earns them funding.

 

Beung a synchopantic yes man in the scientific community earns neither accolades nor funding, despite what the psuedo-science crowd would have us believe.

[Gankfest]

I just figure if something wasn't a real issue like global warming; than scientists would just go research something else. Then people would pay people to do whatever else scientists or the world think os a problem today. Also with someone like George Soros, let's say he is pumping money into research that has conflicts of interest and are skewed. I figure every other scientists would call them out for it, and it would be so many it couldn't be ignored. Then it would get shut down, or just ignored.

 

I finally ran into this... After looking it up a different way.

 

http://www.skepticalscience.com/Hansen-1988-prediction.htm

http://www.skepticalscience.com/Hansen-1988-prediction-advanced.htm

 

But if anyone has anything else that will help, feel free to post.

Edited February 18, 2014 by Gankfest

[swansont]

I just figure if something wasn't a real issue like global warming; than scientists would just go research something else. Then people would pay people to do whatever else scientists or the world think os a problem today. Also with someone like George Soros, let's say he is pumping money into research that has conflicts of interest and are skewed. I figure every other scientists would call them out for it, and it would be so many it couldn't be ignored. Then it would get shut down, or just ignored.

 

 

Yes. Even if the globe weren't warming due to greenhouse gases, those studying the climate would probably still study the climate, because it would be a benefit to learn what we can about any natural fluctuations that exist. That's the nuance that gets lost in the discussion sometimes — these government scientists are paid to study the climate, period, and present their results. Not to come up with a particular set of results.

 

 

One problem is that the biased research is aggressively pushed by those who paid for or benefit from it in an attempt to drown out the others, just as we saw with tobacco, and (apparently) with the NFL's research into concussions. Eventually the weight of unbiased research showed the biased claims to be bunkum.

[Gankfest]

From the reading I've done so far; which I can't even describe how much I've learned about the climate and everything else just in the past few days. Ok, so how accurate are the models that we have today? From reading I know NASA has a super computer GISS that they use. Is there any other climate modeling computer that is comparing the data.

 

In the end, what matters is the accuracy of the predictions. If it was 2% accurate and 98% of error rate in the predictions; than that really isn't good information to take global warming seriously. I figured the the accuracy is better than that... I hope! Anyway, you get my point. I'll soon figure it out though.

Edited February 18, 2014 by Gankfest

[Dekan]

It never fails to amaze me how many psuedo scietific types seem to hand wave away the competition among scientists, as if they all have a vested interest in propping up some false system. They seem to forget that discoveries that were to, say, replace GR with something better are career makers for scientists.

Replacing GR by something better, might be a career-maker for a young new scientist. But suppose he has to apply for funding from a panel of older scientists, who've spent the whole of their scientific careers working on and developing GR theory. How are they likely to react to this whipper-snapper, wanting funds to research a new theory which could prove that their life's work has been based on a false idea.

 

Given human nature, would it be right to suspect that some of the older guys might be disinclined to grant funding?

[Greg H.]

Replacing GR by something better, might be a career-maker for a young new scientist. But suppose he has to apply for funding from a panel of older scientists, who've spent the whole of their scientific careers working on and developing GR theory. How are they likely to react to this whipper-snapper, wanting funds to research a new theory which could prove that their life's work has been based on a false idea.

 

Given human nature, would it be right to suspect that some of the older guys might be disinclined to grant funding?

I'm not a professional scientist, so I'm not sure how the grant process works, but I assume there are checks and balances in the system that try to prevent this sort of bias.

 

Then again, maybe not.

Edited February 19, 2014 by Greg H.

[swansont]

Replacing GR by something better, might be a career-maker for a young new scientist. But suppose he has to apply for funding from a panel of older scientists, who've spent the whole of their scientific careers working on and developing GR theory. How are they likely to react to this whipper-snapper, wanting funds to research a new theory which could prove that their life's work has been based on a false idea.

 

Given human nature, would it be right to suspect that some of the older guys might be disinclined to grant funding?

 

Potentially game-changing physics research gets funded all the time. If this came up (and it's not clear such a scientist would have sufficient influence on granting funds) I think you could argue the ego working in the opposite direction as well— the scientist would be confident that his work is correct and valid, and would not fear a challenge.

[CharonY]

Considering how established GR is, it would have to demonstrate in the proposal a certain likelihood of success. If that is worked out well, the reviewers would have a hard time refuting it. In addition the reviewer are generally not in contact with each other before they submit their reviews. Finally the panel often does not only consist of said referees and for controversial projects it is not uncommon to have panelists arguing either side.

If everyone is established in the panel, there is actually a bigger chance that some want the unknown one to succeed just to spite their rival, with whom they may have built up decade-long animosity (several layers of cynicism have been added which may not represent reality).

[Arete]

So, my adviser and I recently got a grant from the NIH to do some work in a relatively novel field, exploring a somewhat novel study system. In order to get the grant, we generated some pilot data that demonstrated proof of concept and that the study was feasible. This is a pretty common thing to do, at least in my field.

 

To generate this initial data, we used some money from my adviser's start up package. When you start a professorship at a university, some research funding is generally part of the negotiated package. Because this is given by the university, it can be somewhat less restricted than other sources of funding and therefore can be used to explore more risky ideas. In our case, the small investment of start up funding on the pilot data paid off by netting a bigger pot of dollars to carry out a comprehensive study.

 

If start up isn't available, many scientists will appropriate funds from other grants to explore new ideas - I've seen some very creative account juggling to try and fund riskier work on the side of stable, fundable projects.

[CharonY]

Some unis with core facilities also allow you to generate preliminary data using their facilities (typically things like sequencers, mass spectrometers, NMRs etc) on a tab under the assumption that you would reimburse them once the grant hits.

[chadn737]

Science is incredibly competitive. You are competing for limited jobs and limited funding. High profile and high impact research rarely goes unscrutinized. In genetics/genomics, I have seen high-profile papers ripped apart within hours/days of publication. Currently there is a paper on a new method of generating pluripotent stem cells that is being challenged due to the difficulty of replication. Low-profile research can fly under the radar, but its low impact.

 

There exists incentive for scientists to do something novel, but if an idea is well-established, the evidence will have to be much higher to overturn it. Global warming is well established for a reason, an accumulation of evidence.

 

Getting funding for more radical research, at least from the federal government...that can be tough. If program directors and review panels are doing their job, then they will try to fund the best proposals. They tend to fund established and productive labs because it will typically give them more bang for their buck. This has the negative effect of excluding younger PIs who may have new and exciting ideas, but lack the track record of productivity and success that gives them that edge. Its the same reason why as an investor you may choose to buy stock in a Fortune 500 company rather than that new startup. You are playing it safe and going with what will most likely produce results.

 

Another major problem is the public availability of raw data. This is gradually becoming standard practice. Most journals, at least the higher level ones, require genomics projects to deposit raw data into a public database for everyone to access and reanalyze. For climate data, much of it would likely be available as much of it is probably generated by federal agencies. I don't know what policies are standard for private data sets in this field because I do not work in climatology.

Edited February 21, 2014 by chadn737

[CharonY]

Well, to be fair, raw data is rarely useful even for researchers, much less than the public. For example sequencing reads are can be essentially fluorescent images (such as of the clusters in illumina systems). You could upload them but would it be worthwhile? Instead one would upload processed data (i.e. at the minimum after base calling). Same goes for large data sets such as mass spec data (I usually rack up a few gigs per hour) or, serial diffraction data (terabytes of data, much of it being empty images). Although sometimes it could be worthwhile to re-load data and run it through improved algorithms to e.g. deconvolute data, but even with today's storage capacities especially smaller labs would be struggling to maintain the raw data for a long time. Sometimes it is easier to replicate the experiment with newer equipment instead.

 

For modeling papers it is usually customary to identify the data sources (as modelers do not often collect the data themselves) and present the parameters of their model. So it is generally possible to replicate them, although some complicated models would require serious computer time to do so.

[chadn737]

Well, to be fair, raw data is rarely useful even for researchers, much less than the public. For example sequencing reads are can be essentially fluorescent images (such as of the clusters in illumina systems). You could upload them but would it be worthwhile? Instead one would upload processed data (i.e. at the minimum after base calling). Same goes for large data sets such as mass spec data (I usually rack up a few gigs per hour) or, serial diffraction data (terabytes of data, much of it being empty images). Although sometimes it could be worthwhile to re-load data and run it through improved algorithms to e.g. deconvolute data, but even with today's storage capacities especially smaller labs would be struggling to maintain the raw data for a long time. Sometimes it is easier to replicate the experiment with newer equipment instead.

 

For modeling papers it is usually customary to identify the data sources (as modelers do not often collect the data themselves) and present the parameters of their model. So it is generally possible to replicate them, although some complicated models would require serious computer time to do so.

 

I think that kind of depends on the definition of "raw". While sequencing data in its "raw" format is the flourescent images, typically people think of the raw data as being the sequence read itself. At least in the context of next generation sequencing, these raw reads are typically deposited in public databases and there is huge value in being able to reanalyze them. In the past I have made great use of raw microarray data, which I reanalyzed to supplement RNA-seq data. Currently I reuse and reanalyze a great deal of bisulphite sequencing data in addition to new data. Its not merely a matter of trying to reproduce the original results, in all the cases I have used raw data it was actually to discover something new, not to verify someone else's results.

 

Interestingly, this is a very relevant discussion as PLOS has just announced a new policy that has kicked off a shit storm on twitter and the blogs requiring authors to make available all raw data.

Edited February 27, 2014 by chadn737

[CharonY]

Actually I am not clear from the above post whether you are talking about the tiffs or the sequence. Although since you mentioned microarrays, I assume you mean the images or one step below, intensity traces. While I do agree that there is some value in it, even for simple things like using new normalization methods (or maybe improved image processing if we talk about the real raw here), it gets incredibly unwieldy if you run a lab that relies on high throughput instruments and methods.

 

I do see both sides of the argument, although I still think that our ability to generate data has way outpaced our ability to keep them organized and accessible. The organization for DNA/RNA sequences is relatively straightforward, but it can be incredibly complex when moving to quantitative information on other biomolecules where massive deconvolution is done (e.g. metabolite/protein data,to some extent also RNAseq in this context). There have been all kinds of attempts to standardize elements (as e.g. MIAME), but often times biology is too exploratory to conform to them in a neat way.

 

There have been proposals from various agencies to create the infrastructure that would at least allow hosting of that data (although I am not really sure how the discussions about funding have progressed) but especially for quantitative data I am always concerned that the biological part (including e.g. cultivation and manipulation) is often neglected and makes cross-study comparisons exceedingly difficult, which could limit the value of raw data repositories (again, depending on they type of information you wish to extract, and to some extent one could make that point for the majority of biological research). Quite frankly, I have no idea what a more or less unified platform would look like, without messing everything up.

[chadn737]

I don't think that the tiff files are necessary in the context of sequencing. There honestly is not much more gained anymore from reanalyzing the tiff files. At best you gain a few more reads, maybe increase the quality a bit more, but with current sequencing costs, its just not worth it. With microarrays, the standard is to deposit the intensity traces in .cel files.

 

I grant you that it is a difficult process, but I think it is one that is ultimately necessary. In the context of high-throughput data (sequencing for example) There are relatively standardized protocols for this. The short read archive for all next-gen sequencing data is one example. The Gene Expression Omnibus (GEO), allows researchers to put everything from microarray data, to RNA-seq, to text spread sheets filled with rows of data, to qPCR.

 

It is a difficult problem, but I also see little excuse for not making research data available as possible, particularly when it is funded publicly. Researchers have an ethical obligation to be a detailed as possible in their methods and honestly I do not find data dumps to be all that more difficult. In some contexts, there is clear value in wanting to reevaluate the research. Some really high profile and high impact research has been published and found to be false positives...even when there is not deliberate fraud. The first thing that comes to mind was several years ago when one group claimed that RNA-editing was widespread in the human genome. Quite a radical claim and one later shown by reanalysis of the original sequencing reads, to be a methodological fluke. This becomes all the more critical in the context of drug/toxicity testing. Seralini et al. a couple of years ago published a horrifically done study claiming that Roundup ready crops caused cancer. While the study was easily refutable on statistical and experimental design grounds, I remember reading it and the images used to support it. How much easier would it be for individual researchers to evaluate the claims if they had access to data sources such as images from other individual mice in the study, etc?

 

I am always a bit wary of microscopy images in studies because they are always the "best" examples the researcher can find. Especially when the images are of split-YFP or something of the like. I understand, you spend hours at a confocal, you can't publish all hundred or so images and you want to make your case. But at the same time, the internet is already flooded with billions of images. Surely we can come up with a repository for microscopy images?

 

Reanalysis of raw data has to be done carefully, but as long as it is done with care, the data extracted can be extremely useful, not to mention saving time and money. Meta-analyzes of GWAS studies have produced so many Nature Genetics papers at this point that I don't think we can deny in any fashion the value of reanalyzing raw data.

[CharonY]

So in that case we are talking about at least one step-analyzed data, which can be quite a different beast depending on the analytical technique (serial X-ray crystallography comes to mind). I am more open to that, but again, I do find nucleic acid data the easiest to handle. I do agree in many ways that re-analysis of data is valuable and i would put biomarker research pretty much in the front for this (at least when it comes to validation purposes).

But again, for some of our projects we had to send out data using terabyte harddisks to collaborators, because it was simply not feasible to transfer that amount of data in a timely fashion otherwise. I actually do agree with microscopic image issues, but that opens up another can of worms. You could substitute best image with best set of images and the issue would not go away completely. The only way would collect all data in an unbiased way, including runs and examples that you may consider as failed. While they may contain information therein, the flipside is that there would be even more data to sort through and avoid spurious associations.

As it is now, crap is already abundant in published studies, if we add everything to the mix I am not sure how look through all of them. I would maintain that this more appropriate for strong, high impact claims (in which case a more thorough review is more appropriate than yet another me too paper).

 

I am not arguing that data should not be made available, but my main gripes are a) who is paying for the infrastructure and maintenance of the repositories? Grants certainly do not provide me with funds to do anything more than the experiments (and often barely so) and b) what would be a good system to make the sets useful for a range of applications so that we do not intellectually masturbate over data that is actually the result of bad experimental design?

 

Biological studies are often so diverse in the way we try to tease out functions that trying to pulling differing types of data together is not working very well (and I could launch into the whole systems biology rant, but I will refrain myself here). Again, I am all for sharing data, but at least in my field the infrastructure is not there yet.

[Arete]

I am not arguing that data should not be made available, but my main gripes are a) who is paying for the infrastructure and maintenance of the repositories? Grants certainly do not provide me with funds to do anything more than the experiments (and often barely so) and b) what would be a good system to make the sets useful for a range of applications so that we do not intellectually masturbate over data that is actually the result of bad experimental design?

 

Yes. I'm sure we all know data that's come from NCBI that's very suspect. In fact, I know of one lab who deliberately exclude information from their Genbank uploads, forcing you to contact them if you want to use their data in any meaningful way - completely unethical and in violation of almost all journal policies, but the data police haven't knocked down their door yet. Also, I've reviewed some papers who mine data and then combine it in erroneous ways that may seem ok unless you have knowledge of a particular study system - e.g. the taxonomy has changed substantially meaning sequence A published 10 years ago and sequence B published last week are not necessarily from the same species even though Genbank tells you they are.

 

An issue I've always had is that Genbank make it such a huge PITA to upload data - it almost makes having a paper accepted bittersweet, in that it means I then have to deal with Genbank. I much prefer using Dryad to make data publicly available. While it's not nearly as searchable as Genbank, I can upload data in a variety of far more useful formats (like Nexus alignments, vcf files, genepop files, etc) as well as non-genetic data, like information mined from bioclim, morphological data, phenotypic data, etc. I personally feel that having the data available in exactly the format I actually used it in my study makes it vastly easier to a) vet the analyses I published and b) incorporate data into a meta-study.

 

One thing I've had more issues with is papers where I've written my own code and the been barraged by people trying to use it. I'm far from a programmer, and my code is probably pretty awful from a computational standpoint. I totally get that it needs to be transparent so people can see what I did, but I wish I could put a disclaimer in the paper like "This janky piece of code did what I wanted it to do, on my data, on my system. If it doesn't work on your data, on your system, I don't know why. I don't have time to fix it for you. It's time to bust out "Perl for dummies" or collaborate with a computer scientist."

[Roamer]

I got some Tin Foil hat questions about science research funding and Ethics...

 

When it comes to scientific research from federal and private institutions around the world.

 

1) How much do politicians, spooky people like George Soros, and Evil villains influence the funding and research of these institutions?

 

2) What's stopping scientists from cooking the books just to keep their jobs...?

 

3) Is it possible to cover up the facts I learned about global warming recently here?

 

4) I'm just curious because my friends thinks this, and personally how are you going to argue that... Anyway it really got me thinking what the integrity of the scientific method used today, and how the funding is spread around? I couldn't really find anything in depth on the interworking of the science community and the proccess of coming to a consensus. A video on Youtube is good enough.

 

5) Also what's this fallacy called?

 

6) If 1 person is lying, than everyone is lying.

 

7) If the FDA lies about a drug; than the CDC is lying too.

 

Than You!

 

I numbered the questions for some ease

 

1) The real question would be how much you and/or your friends contribute themselves,

this does not necessarily mean funding, but for example an essay on how these villains actually do things that we disapprove of.

 

2) Unemployment benefits.

 

3) How easy is it to manipulate you ?

If it is hard to manipulate you it might be easier to cover up the facts you did not learn.

 

4)There are some bigger organisations out there, they're maybe better targets for the tinfoil hat-approach,

but for the most part, humans just do.

 

5)/

 

6) People do not lie, who told you that ?

People say whatever serves them best, and if we do not like what they say we call them liars and drag examples from this reality that we live in to counter the "liar's" argument.

 

7)Hmm, i 'm going to assume those are american government agencies dealing with drugs, and you're not talking about them forging false results on some research.

Those agencies, they have their own set of standards, if those standards are met by a substance, they attach a label to the substance.

When you hear "drugs" or "medicine" and automatically get a picture of something good/evil(i suppose), then that's your doing.

If you go to their website(s) and actually figure out what their standards areand what they mean according to them, that'll be your doing as well.

[magnocrat]

It is difficult to make a moral case for spending enormous sums on Hadron Colliders or Martian exploration when there are many urgent problems right here on our planet. Science is a sort of golden calf in our technological age what ever goals it pursues are always accepted as worthwhile.

[swansont]

It is difficult to make a moral case for spending enormous sums on Hadron Colliders or Martian exploration when there are many urgent problems right here on our planet. Science is a sort of golden calf in our technological age what ever goals it pursues are always accepted as worthwhile.

 

Define "enormous sums". The National science foundation research budget in the US is a little under $6 billion. Americans spend more than that on Halloween. One big difference, though, is that science research may solve some of the problems we face.

 

The goal is to learn things. But it's research — you don't know if it works until you try.

[magnocrat]

Using that argument I could get away with any waste I like.

The man next door throws away a ton of food each year so why shouldn't I throw half a ton. The other great cry of the scientist is you don't know where it will lead , a sort of open- ended appitite for money.

Well tell me where did the internal combustion engine lead-- total pollution.

And now the scientist who contributed to the world crisis present themselves as gardian angels. We might get more deadly weapons than ever before known. Meamwhile back at a ranch called earth 30,000 starve to death each day.

[imatfaal]

Using that argument I could get away with any waste I like.

The man next door throws away a ton of food each year so why shouldn't I throw half a ton. The other great cry of the scientist is you don't know where it will lead , a sort of open- ended appitite for money.

Well tell me where did the internal combustion engine lead-- total pollution.

And now the scientist who contributed to the world crisis present themselves as gardian angels. We might get more deadly weapons than ever before known. Meamwhile back at a ranch called earth 30,000 starve to death each day.

 

What's your excuse for wasting time on a science forum when you could be out in the fields growing crops to help feed those 30,000 people or at least raising money to help them? You said "Using that argument I could get away with any waste I like." - well reusing your response one could condemn any activity that isn't directly related to saving those lives. Your argument are far too extreme - and leave no room for human compromise.

 

Your argument was that we could not afford the LHC etc when people were starving. The counter was made that science does a lot of good and yet costs less than the celebration of an antiquated feast of the dead. Which bit of the counter is wrong? Wouldn't it be better to criticise the hapless consumerism that leads us to spend more on pumpkins than on the search for new anti-biotics?

 

If you don't believe science is a worth-while endeavour what are you doing here? How do you reconcile within yourself the contradiction of decrying scientific contribution to global culture AND doing it on the internet!