ecoli

cool, any high school student can now consider themselves a mathematician.

The more common portmanteau I've heard is multiverse. Is this another erroneous case of mixing latin and greek? edit: Just looked this up: multiverse is correct. Versum is latin for "something rotated, rolled, changed" and multi is a latin prefix. Poly- is greek so polyverse would be a hybrid word. There are multiple Dr Maybes? 'cause there's only 1 ecoli.

! Moderator Note Are you sure you posted that in the right thread, Ben Bowen?

Bostrom says that there are empirical reasons why the universe might be a simulation. AFAIK, he doesn't make strong guarantees. By stating it in this way, you're making the general philosophy sound a whole lot kookier than it is. Again, Bostrom postulates the possibility of multi-level simulations, but I don't think he even says it's likely. Only if the probability of the existence of each multi-level simulation is independent, which isn't likely, even by the reasoning of the simulation hypothesis: For one thing since, as Bostrom argues, if the fraction of observers that live in a simulation is close to 1, then the fraction of posthuman civilizations that are interested in running or survive to run ancestor-simulations is close to 0. This seems to be contrary to the simulation within a simulation scenario, since the likelihood of each successive simulation advancing to a stage where that kind of computation is feasible is very low and you have to assume that creating simulations is allowable by the rules of that simulation. Further, you have to argue that stacked simulation doesn't degrade computational efficiency, which isn't true in our own observable universe. if the universe already ended, the probability of us having this conversation would be zero, so no. Well yeah, but more from lack of evidence than by your reasoning above. Even Nick Bostrom puts the probability of the simulation argument being true at around 20%... which is still pretty insanely high. Postulating that reality could be an illusion has long been treated by Philosophy, from Descartes and back to the Greeks. Do you have a more substantial complaint?

really? You need that level of precision for practical uses?

I'm pretty sure that Nolan Chart questions are designed to emphasize the libertarian/anarchist position in surveys. http://en.wikipedia.org/wiki/Nolan_Chart

Here's an R package for image analysis: http://rss.acs.unt.edu/Rdoc/library/EBImage/html/00Index.html

huh?

Pretty words, but I'd argue that many of the gov't-driven initiatives that improved tech and innovation during the 1960s was a direct result of people's fears during the Cold War and that the level of greed had nothing to do with it.

what is it that you DO want to argue because I don't really see a coherent point to this.

that's a pretty strong claim for which you've provided zero evidence

I don't think anyone would deny that greed drives people... especially politicians. An appeal to mass catharsis seems pretty far fetched though. People respond to incentives, not vague moral appeals.

Publishing in a peer-reviewed journal would be a good start. That way the scientists who produce mainstream work can evaluate it. Most users here aren't qualified to evaluate modern physics.

It can get confusing, but especially for popular topics it would be impossible and undesirable to keep members from contributing. If this is happening then its important to review the whole thread and keep track of who is saying what and if there's potential confusion, politely point it out. These threads belong to everyone and the point is to facilitate conversation. At one point, we had a board dedicated to structured debates between two people. If there's interest in that sort of thing, we can talk about bringing it back.

You can do a little better than this by adjusting your alpha.If you think of the statistical test as a Bernoulli trial, with a random, binary outcome. If the null hyp is false for all trials, this gives you a distribution of outcomes at a probability equal to the statistical power. Therefore the probability of not making any type I errors is [math] .95^{N} [/math] where N = the number of trials. So for multiple trials, to retain the same confidence, you adjust your alpha to 1- .95^(1/N). Which is very stringent for multiple comparisons and doesn't seem to be done much in practice. edit: I just noted a small error with what you said. PPV isn't the ratio of true associations to false associations (that's the specificity). PPV is the ratio of true associations to the number of assocations tested as positive (so true + false positives). This may have not been what you said.

What kind of jobs are you applying to? If you're interested in research jobs, try to volunteer/intern in a basic science lab. That'll get you the experience you need to 'prove' yourself for other jobs. If these are BS-only level jobs, and you're in a STEM job already I would think the transition shouldn't require going back to school (which would honestly be a waste of time, considering you already have a science degree). If you prove that you're smart and motivated, maybe a lab would even hire you into a research tech position. In my lab, there's a tech who came from a TV production background without a science degree and she's doing just fine.

try checking the sites on isup.me and then ping the URL from a terminal and see if you get a response

This can be handled with good stats, though it often isn't (like correcting for spurious correlations when making multiple comparisons, autocorrelation analysis, etc. Thanks for the references! Also, there have been multiple studies showing immune system reponses changing to engineered or changed microbiota in germ-free mice and in recolonized humans. Often with the accompanying relief of IBD symptoms. In mice,SFB has been shown to induce TH17 responses upregulating immune defenses and decreasing pathogen susceptibility (all with prospective, molecular techniques) Seems to me to survive the spirit of Koch's postulates, minus the requirement for culture based isolation, which isn't really necessary with genomic techniques.

Nice... I'd want to take regular blood serum samples, microbiome samples and as much patient meta-data as possible to control for confounders. Are there antigen panels you could run from blood samples? I could definitely design an experiment if I put thought into it. You made the claim that ex post association isn't enough (though there are legit ways to deal with this, statistically) so the burden is really on you to propose an alternative.

Which is often how interesting research starts. The BRCA mutations are 'merely associated" with breast cancer risk, yet its common medical practice to undergo prophylaxis for people with suspected-harmful BRCA variants. So while much of the mechanism is worked out now, but certainly not for all variants. If researchers had scoffed at this mere 'association' millions of people would not currently understand their risk for developing cancer. You seem to be conflating a philosophy or method of knowing (ie - "Science") with something teleological and thinking. People use science to find things out about the universe but science isn't really a proper noun. I agree that researchers shouldn't be satisfied with conjecture and that hypotheses should be tested and retested until the molecular details of any biological system is teased out. I've never claimed otherwise. However, it seems like your saying that correlation & association isn't evidence. Well with that you're wrong. Its true that correlation doesn't imply causation, but its also true that you can't have causation without correlation. So finding robust associations is often 'good enough' for early stages of medical practice and is a frequent starting point for experimental validation (provided you know your stats). What kind of evidence would be good enough for you? Because whether or not you agree with the hygiene hypothesis, many scientists certainly do and certainly the NIH does because research related to heterodox immune-microbe interactions IS getting funded.

I think a trade organization CAN be that even if the field requires rigorous training. In an ideal world, dieticians would be trained like doctors (or at least, doctors would recognize nutrition as a vital part of their practice). People have always been bad at evaluating the quality of evidence. What's the difference now? Just that so much more noise is available thanks to the web? Dispensing bad medical advise is one thing though (taking snake oil rather than cancer drugs is obviously bad). I'm not so clear that unlicensed nutrition advice is so much worse (or even often better) than "official" nutrition. Just take the old school food pyramid: for years and years nutritionists recommended supported your diet with carbs and lumping in all oils, fats and (somehow) sugary foods. There was some common sense to it, but it wasn't based off any cohesive research. I'm not even entirely convinced that the new pyramid (which still emphasizes grains over veggies) is all that great. Now, "alternative" dietiticians have been pointing this out for years and years. I'm just not sure that silencing unlicensed dieticians would do more good than harm.

Yes, I think there's got to a be an understanding that dispensing general advice is not the same as treating a patient. I mean, whats the logical conclusion here... journalist licenses and a ban on unofficial news blogs?

space cows!

or so says the alarmist headline over at Forbes. Curious to hear others' thoughts on this. Should dispensing out nutritional advice be under the protected domain of those licensed to do so or is this a violation of free speech? My take: nutritional advice given by a "registered" dietitian is probably no better than picking out a nutrition blog at random: ie - mostly bunk and not research-based.

I have to disagree in part. While the mechanistic details aren't all worked out yet, reporting on associations between immune-related disorders and microbes is perfectly reasonable as evidence. I think you're being overly demanding. The hygiene hypothesis is a general framework of thinking about immune related disorders as arising from a lack of key microbe-host interactions. Only with recently sequencing technology has it become feasible to even study the hypothesis in a meaningful way. I do agree with your skepticism about overdrawing conclusions, but your asking for general proof for a hypothesis that doesn't really make general claims! The only serious claims made by scientists is that some microbes may be factors in the prevention of some immune-related disorders. Therefore, scientists have been evaluation and testing this claims in human and animal studies one at a time. I'm not sure if there's a single experiment that could support or refute the many claims that the hygiene hypothesis is making, because the mechanistic details are probably very different depending on the disorder and microbe factor. I mean, this is kinda like demanding 1 experiment to prove that genetic polymorphisms mutations cause cancer. Well, this claim is certainly true to a limit, but SNPs are only risk factors and difficult to predict outcomes from and usually only provides a clue about mechanism. Reasonable claims about the hygiene hypothesis are similar to this: microbes have been evolving with mammalian hosts for millions of years, so its not unreasonable to assume that they've evolved stable and even beneficial relationships with our immune systems: perturb these relationships and you might get diseases. Is this really so different than what we already know about the causes of other diseases? agreed, but that's what you get if your exposure to probiotics info (and science in general) is from sites like webMD. The actual science is making much more reasonable, carefully researched, and well-supported claims.