4 ms·
This metaphorically reeks of being the sort of result that in six months will be quietly retracted. But at least there's a mechanism proposed that isn't utterly
by jerf 13y ago
This metaphorically reeks of being the sort of result that in six months will be quietly retracted. But at least there's a mechanism proposed that isn't utterly ludicrous...
(See expansion below, which describes why this triggers my BS meter. YMMV, but it's not just "reflexive".)
- groby_b 13y agoPlease do expand on your thought there. What makes it likely this will be retracted? The experimental set-up seems fairly sound. They controlled for the obvious factors that could influence the outcome. It is somewhat consistent with what we know in other areas - yes, pain is felt less under increased stress, and yes, the presence of other males is a stressor. The interesting part is that it's an inter-species stressor. But then again, testosterone as a male hormone is shared across a lot of species - it's not entirely unreasonable to assume that scent markers might be similar too. So, what makes your BS sensor go off? (I ask in all seriousness. I'm always interested in calibrating my sensor, and it didn't go off for this).
- jerf 13y agoLike I said, and I meant it seriously, it does at least propose a non-silly mechanism that makes sense. But it A: proposes a sweeping result which, if taken seriously, ought to invalidate a lot of other experiments and B: pushes a moderately political button with the gender issue. Neither of those is a disqualifier, but A in particular at least somewhat gets into the realm of "an extraordinary claim that will require extraordinary confirmation". The proposed mechanism would also seem to imply that female hunters don't exist. I would have preferred a proposal that you're inclined to hide your weaknesses around a male in general (lest he either take your woman or harass you in other ways), and the mechanism is likely keyed off of testosterone in general rather than mouse smells in particular, which seems reasonable. My head isn't going to explode if this is established by further experiment. But it's suspicious to me. To siyer's point, impeccably-done research can still come up with coincidental results, unfortunately.
- sp332 13y agoNo, the mechanism is increased stress, and it has already been experimentally verified (although I'd like to see it repeated). Mogil has a guess about why that might happen, and it's called out as supposition, not an "extraordinary claim". Other potential explanations include a few male rat-farmers who were rough with the rats, which made the rats edgy around males later in life.
- CamperBob2 13y agoNo, the mechanism is increased stress, and it has already been experimentally verified (although I'd like to see it repeated) Until it has been repeated, it has not been verified. That is how experimental science works.
- sp332 13y agoScience is repeatable, not repeated. It's not true that one experimental result is meaningless but two are gospel. The first has a probability of being a fluke, and the second reduces that probability, that's all.
- CamperBob2 13y agoIndeed, this is true as well. Science isn't the right place to look for absolute certainty. But still, there are some obvious things you can do to increase confidence in a result, and one of those is to get other people to try to reproduce those results.
- siyer 13y ago"and the mechanism is likely keyed off of testosterone in general rather than mouse smells in particular, which seems reasonable." Yes, this is what Figure 1c and 1d of the paper demonstrate. (To be precise, not testosterone, but adrostenone, androstadienone, and 3M2H). "My head isn't going to explode if this is established by further experiment. But it's suspicious to me." Supplementary Fig. 1 and Supplementary Fig. 3 describe replications performed in two other laboratories, one at the University of Alabama, Birmingham, and the other at another laboratory at McGill. Look, I'm all for scientific skepticism, and it's true that there's a lot of low-powered stuff out there. However, papers are best evaluated on what they actually say (which, incidentally, is why open-access/archiving is important), rather than on a filtered impression of what a press release says. If you'd like a copy of the paper, let me know, I'd be happy to send you a pdf.
- siyer 13y agoThis is just reflexive HN negativity. As a pain researcher who's completely unaffiliated with the group that did this work, I find this to be an excellent paper - carefully done, with good controls, standard sample sizes for the field, and sensible assays. People interested in more of the details can read the Nature Methods article here: http://www.nature.com/nmeth/journal/vaop/ncurrent/full/nmeth.2935.html http://www.nature.com/nmeth/journal/vaop/ncurrent/full/nmeth... It also fits into a long history of work out of the Mogil lab about the limitations in our current assays for pain. It's worth noting that this is by no means an easy problem. Pain is inherently subjective, and measuring and quantifying it in humans is difficult, let alone in rodents. The only way forward is more work like this - that carefully quantifies the sources of potential bias and error in our assays, and proposes solutions for how to get around them. Edited to add: Vox has a good interview with Mogil that goes more into this. You can read it here: http://www.vox.com/2014/4/28/5653444/mice-get-stressed-out-from-men#interview http://www.vox.com/2014/4/28/5653444/mice-get-stressed-out-f...
- CamperBob2 13y agoI'll take the liberty of rephrasing jerf's comment: this is the sort of paper I'll be more likely to believe after it's been reproduced by multiple labs. Until then, I'd just as soon not even be told about it. It sounds too much like any number of "xxx affected by cell phone radiation!!1!!" results that historically have given rise to much panic, intense discussion, and clicking of links, but have never, ever, ever proven to be consistently replicable. At some point, skepticism and even a bit of "HN negativity," as you put it, becomes the most rational response to initial reports of such phenomena.
- siyer 13y agoMy point is more that there are differing levels of the prior probability of a study being true. Yes, 'xxx affected by cell phones' stories are bunk, however, those stories tend not to be published in Nature Methods, tend not to come out of highly respected laboratories, and tend not to have voluminous documentation and results, with 9 Supplementary Figures. I've been reading this paper closely over the past 30 minutes, it's important enough that I'm sure it'll be heavily discussed at conferences and meetings this year. I can't think of a single thing that these researchers could have done that they haven't done. They've approached the question carefully, looked into a ton of second-order explanations and effects, and provided a plausible discussion of how and why they see their effects.
- anigbrowl 13y agoThis metaphorically reeks I see what you did there.
- nuser 13y agoYour comment reeks of being the sort of claim that you will pretend you never made if nothing changes, but you will hold up as evidence of your incredible knowledge of it isn't. There's literally nothing of value in your comment (unsurprising, as nearly all of your comments are just you restarting your tired and dullheaded Libertarian beliefs). You're upvoted solely because HN is a shithole that was long ago ruined by nontrepreneurs like you.
- sesqu 13y agoThis particular study seems fairly solid, but the linked one about a cardboard cutout appears contradictory. The comment about predators seems careless, in that light, but the person making the comment has clearly been theorizing about the effect for a while. So I don't expect the study to be retracted, but I expect half a dozen more papers and then a theory that will be eventually debunked in favor of a more politically convenient one.