Wikipedia:Reference desk/Archives/Science/2015 July 29

= July 29 =

Merger
The question is straightforward: should inventory control be merged with Inventory control system?Lbertolotti (talk) 01:29, 29 July 2015 (UTC)
 * Have you tried discussing that on those articles' talk pages? ←Baseball Bugs What's up, Doc? carrots→ 03:00, 29 July 2015 (UTC)

Yes, so far nobody said anything.Lbertolotti (talk) 03:13, 29 July 2015 (UTC)


 * Although there is no real harm in raising the issue here, the Ref desks are not really the ideal means to solicit additional opinions as relates to the best policy approach to an issue on an article. I would recommend you explore Proposed mergers, WP:RfC, and (for this particular case) Wikipedia talk:WikiProject Business.   The first is a noticeboard to promote merger discussions, the second is a process page which will guide you in how to attract outside input to article talk page via a method known as a "request for comment", and the third is the talk page for a Wikiproject (collection of editors with a common interest) for business-related articles.   One of these methods should surely attract some attention to your request for additional input.  And, of course, if you feel very confident that you have reviewed the relevant policies and that a merger would be warranted in this instance, you could always WP:BEBOLD and institute the change yourself (since no one has commented despite your best efforts to raise the issue on the talk Page) and then if someone objects or reverts the change, you can invite them via their user talk page to comment on the article talk page so you can get their views and a better feeling of how to proceed.   Regardless of which route you choose, I applaud your efforts to approach the change in a slow and cautious manner and to go above and beyond to seek additional input and consensus.  :) I will give my own opinion on the talk page shortly, but you should keep the above methods for outreach in mind for the future.    S n o w  let's rap 09:51, 29 July 2015 (UTC)

Reference_desk/Archives/Science/2015_July_21
I'm updating the pictures description, if there's no objection. Lbertolotti (talk) 01:34, 29 July 2015 (UTC)

The handedness of a dead body
Could a coroner tell if the person was left-handed or right-handed? Anna Frodesiak (talk) 06:13, 29 July 2015 (UTC)


 * It is possible if the dominant hand was used in repetitive activities, such as archery or metalworking, as the musculature will develop more and produce bone deformities. I'd say (not being a coroner but being a physical therapist) that modern life in general doesn't produce such gross deformities, the evidence would be more nuanced. --TammyMoet (talk) 08:42, 29 July 2015 (UTC)


 * I am not a coroner, but if I were asked I'd check the middle fingers for calluses. There is likely to be one on the dominant hand, caused by writing. DuncanHill (talk) 08:50, 29 July 2015 (UTC)
 * That might be harder to do these days, with people doing less writing: I can't see any obvious external differences between my middle fingers, though perhaps an export could. According to this, "[Forensic anthropologists] can tell whether the person was right or left-handed. There would be more muscle attachment on the bones on the dominant side". By the way, a coroner is a legal officer (at least in the UK), so they would not be the one doing the physical investigation. AndrewWTaylor (talk) 08:57, 29 July 2015 (UTC)


 * Use of a mouse on a computer affects the wrist, indicating handedness. Constantly moving a finger over a screen causes a callus, indicating handedness. That does not include muscle mass. Most people are stronger in their dominant arm. In boot camp, everyone had to work out. It was easy to identify the left handers because they had trouble lifting weights in the right hand that were easy in their left. For everyone else, lifting weights in the left hand was harder (except for the two who were body builders previous to boot camp, one put on weight control specifically to lose muscle mass). 209.149.113.45 (talk) 16:11, 29 July 2015 (UTC)
 * Not sure this would be deterministic. My father is left handed and before wireless mice made switching easier, he simply used his right hand.  I am right handed but left eye dominant so some things like shooting a rifle is left handed, while a pistol is right handed.  I use the wireless mouse on the right side, but the dot mouse on the laptop I use my left index finger.  Not sure why, it's just natural for me and it's not ambidextrous where I can use my right finger for the dot mouse or my left hand for the wireless mouse.  Could be just training. rather than handedness.  --DHeyward (talk) 21:21, 29 July 2015 (UTC)


 * Many things could interfere with such presumptions. I generally carry heavy things with whichever arm has less joint pain that day.  When I'm at my computer and not typing, my right hand generally rests on the arrow keys and my left hand on the mouse, so that my mousing hand has less far to move to reach the letter keys.  When I shoot a pistol one-handed, my ‘wrong’ hand is steadier. — There is a faint difference between my middle fingers, but nothing like the very obvious groove I had at age 16; I no longer grip a pen with such force! —Tamfang (talk) 06:58, 31 July 2015 (UTC)


 * These arguments are anecdotes vs data. Suppose I take a million computer users. I identify a spread in the wrist bones on 90% of them due to mouse use and verify that the hand use for their mouse is the hand they normally use for everything. I can then state that, after identifying a spread in the wrist bones, I am about 90% confident in identifying the handedness of the person. I work with this type of argument every day. For example, over 99% of people who have an HbA1c>6.5% are diabetic. Still, people want to talk about an anecdote about some guy they heard of who had an HbA1c of 8%, but wasn't diabetic. Therefore, they think they've invalidated the millions of other data points that support using HbA1c to identify diabetes. They are wrong, very wrong. In other words, the plural of anecdotes is not data. 209.149.113.45 (talk) 16:55, 31 July 2015 (UTC)
 * With all due respect, I don't see much in the way of hard data here; it's mostly anecdotes vs conjecture. —Tamfang (talk) 06:56, 1 August 2015 (UTC)
 * Agree entirely. I actually know someone who's right handed but because of RSI related problems started using the mouse left handed. I would agree however this is likely a real minority, but "likely" is very different from hard data. In particular, I'm far from convinced the percentage of people who are left handed and who uses the mouse left handed is anything close to 90%. I found this survey which has a number of activities. The activities where over 90% of left handed people do use their left hand (not sure whether they gave any clarity on what to say if you use both) are writing/drawing, brushing/combing own hair, holding a toothbrush and using a spoon by itself. Sadly using a mouse isn't anywhere on the list. Interesting enough, using a mouse while writing is a one sample answer for the advantage of being left handed. Based on the earlier stats, we can assume this person likely means they are at least comfortable enough with using the mouse right handed that they can use it while writing. There are plenty of other anecdotes of people using the mouse right handed despite being left handed. These aren't particularly useful in themselves, but they do give pause for thought particularly when you consider the reasons they do so and in the absence of actual data demonstrating such a high percentage of left handed people use the mouse left handed, reason to doubt it. (Obviously a real minority of computer users but it seems the first gaming mouse for left handers may have been in 2010  which may be part of the reason for ). It does depend on what the IP means by the statement. Since the percentage of people who are left handed is so small, it doesn't matter such much if you misidentify many left handers as being right handed in terms of overall numbers, your prediction accuracy will still be decent. In fact, if you just predict every as being right handed, you could probably achieve 90%+ accuracy. But if your purpose is to determine if someone is left handed or right handed, I'm not sure how useful a test is if you incorrectly predict a big percentage of left handers as being right handed. OTOH you may very well prefer a test which will correctly predict 98% of right handed people and 50% of left handed people to one which will correctly predict 75% of both left handed and right handed people. But then again, your ultimate preference is probably for something better than both, so it's probably better to look for something rather then settle on one of two equally poor options. P.S. You may assume that if someone uses both hands for the mouse, this will lead to a detectable skeletal pattern so it doesn't matter if someone uses both hands often instead of just one, you can predict this person is probably left handed. On the other hand (just noticed the possible pun), you then have to demonstrate that this skeletal pattern can be reliably detected, compared the the possibility the person just doesn't use the mouse enough to have any pattern, or whatever. Ultimately it comes down to me and I presume Tamfang's main point. You have a hypothesis, but little actual data presented here to support it. And given the data available, the alternative hypothesis is IMO still equally valid. Nil Einne (talk) 18:40, 1 August 2015 (UTC)


 * The problem with diabetic example isn't the anecdote or conjecture side, but it highlights the problem with the method he proposed. The correlation is only one way.  Having the particular gene may very reliably predict that the holder of that gene is diabetic.  But not having that gene doesn't correlate to not having diabetes.  So if the question is "Was the person diabetic?" having the gene is a high confidence "Yes",  and not having the gene isn't much better than the population percentage and the usefulness as post-mortem tool is only as great as the gene expression itself.  You would have to know the predictiveness as it applies to predicting each hand.  Correlations are probably not symmetric so statistical tests would have to be applied to each null hypothesis (i.e. if the right hand is larger than the left, it's 60% higher chance over than the normal population chance that the person was right handed.  If the left hand is bigger than the right, there is no correlation).  The asymmetric nature makes test harder (i.e. if I simply say every deceased person is right handed, I'd be right 85% of the time - the test would have to be better than that).  --DHeyward (talk) 19:59, 1 August 2015 (UTC)


 * This book has a chapter titled "Skeletal indicators of handedness". SemanticMantis (talk) 14:55, 29 July 2015 (UTC)


 * Sink your teeth into this! There is another way to find a deceased's handedness which I found out just 2 hours ago! I took my sister to the dentist today and had a good chat with her (the dentist).  I was telling her that I had heard right-hemisphere dominant people chew on the left side of the mouth and vice versa.  She told me it was true, and this affects the distribution of cavities.  She then went on to tell me that right-handed people brush the teeth on the right of their mouth harder (better) than on the left (and vice versa).  So, right-handed people have fewer cavities and less plaque on the teeth of the right side of their mouth!  Elementary my dear Watson! DrChrissy (talk) 17:22, 29 July 2015 (UTC)
 * Was this what she was reporting the dentist had told her, or was it her own opinion? Is she a dentist herself?  --   Jack of Oz   [pleasantries]  19:30, 29 July 2015 (UTC)
 * Apologies for my unclear posting. This was what the dentist was saying to me. (My sister's input to the conversation was "Arrrggggh", "Ugggguuuuuhhh" and "thank God that's over". DrChrissy (talk) 20:48, 29 July 2015 (UTC)
 * Are you sure that you have not reversed right and left for brushing. A comment by my dentist implied that it was the other way round.    D b f i r s   06:55, 30 July 2015 (UTC)
 * In my excitement at hearing the relationship, I might have mistaken it. However, I am right handed, and it feels like I put more pressure on the teeth on the right side of my mouth. (sample size N=1 !) DrChrissy (talk) 12:00, 30 July 2015 (UTC)


 * A person who writes with the left hand is more likely to use the right hand to support the right cheek, which as a result will likely tend to be more concave than the left cheek. (I searched for a reference, but found none.)  This question is relevant to WikiProject Medicine/Participants (including User:Doc James, User:Bluerasberry, User:Jfdwolff, User:Mattopaedia, User:Richardcavell, User:Looie496, User:Ozzie10aaaa, User:CFCF, and User:Peter.C).
 * —Wavelength (talk) 23:35, 29 July 2015 (UTC)

Thank you all!! Very, very interesting indeed!!! Pity, though, that I can't see the book with the "Skeletal indicators of handedness" chapter. Anna Frodesiak (talk) 05:51, 30 July 2015 (UTC)


 * Late to the party, but FWIW, in both the novel and film In the Heat of the Night, Police Detective Vergil Tibbs determines that a suspect is left-handed (and therefore likely not the perpetrator) simply by feeling the musculature of his forearms. Unless the author John Ball thought this up off his own bat (is that a USA-known term?), it's presumably possible to do the same to a corpse. {The poster formerly known as 87.81.230.195} 212.95.237.92 (talk) 13:25, 30 July 2015 (UTC)
 * Interesting. I will start to feel everyone's arms to see if there's a difference. Anna Frodesiak (talk) 09:17, 2 August 2015 (UTC)

Maxwell's demon and the Carnot efficiency
I'm curious whether there is a simple, straightforward way to relate the maximum efficiency of a Maxwell's demon to the maximum efficiency of a Heat engine, i.e. 1-Tc/Th. There is an impressive paper that does so but it is somewhat difficult for the non-expert to process, and I'm not sure if the quantum mechanical features they focus on there are important for making this connection or just a distraction. Also there's a discrepancy between that paper and our entropy article on one hand and the Landauer's principle article on the other; the former use kB ln 2 for the entropy that must be produced elsewhere and/or energy cost, while the latter uses kT ln 2. I'm thinking the latter is measuring entropy in terms of joules and the former two doing something else but I'm not quite sure why. Is there a straightforward derivation by which you can start with this expression (whichever one) for the cost of erasing a bit and end up showing that the demon has the same maximum efficiency as a heat pump for reservoirs of the same temperatures? Wnt (talk) 23:46, 29 July 2015 (UTC)
 * Sorry, Wnt, there's no formal definition for the efficiency of a demon, so there's no meaningful way to compare it to the efficiency of a heat engine. Nimur (talk) 00:03, 30 July 2015 (UTC)
 * But surely it could be measured empirically? {The poster formerly known as 87.81.230.195} 212.95.237.92 (talk) 13:33, 30 July 2015 (UTC)
 * Sure, if we can only find a cost-effective and safe way to obtain a statistically-significant number of demons! I'm still looking for a cost-effective and safe way to obtain a statistically-significant number of neutrons, which I consider to have more interesting thermodynamic properties... regrettably, the best I can find is a one-of-a-kind spallation source on the opposite side of this continent - but it's available to anyone with any experience or credential level, so long as the research proposal is meritous!  Nimur (talk) 16:25, 30 July 2015 (UTC)
 * Not sure it's applicable but see here Bose–Einstein condensate. Using lasers to cool through coherence seems pretty neat.  QED and QM seem fundamental.  For some reason, Maxwells demon seems like the simple aerator on my pool.  The hot molecules leave as evaporation, cold molecules stay and the pool is cooled.  It's not a closed system though but the "demon" is a pump.  At some point the demon would have too many collisions with the door (i.e. similar to space-charge region in a plasma where the atom has so much less velocity than electrons at a given temperature that a charge region forms until the collisions balance and the current is zero).   I would think that barrier could be QM related distance with tunneling and other effects.  Black hole formation and evaporation may also have relevance.  Just a thought.  --DHeyward (talk) 04:02, 30 July 2015 (UTC)
 * I'm surprised if there isn't a formula for the efficiency, but to give an example: today I had to deal with a car issue and was waiting for a long time in a room with a soda pop vending machine.  (For some inexplicable reason, the supposed pleasure of having a cooled soda pop outweighs the prolonged and annoying noise of the contraption; as I don't actually refrigerate soda, it is all very mysterious to me)  Anyway, the point is, I'd like to rip out the condenser and coils and replace it with a sort of refrigerator magnet lining the box that has a thin layer of water in it and many, many demons that put the hot water to one end of the layer and the cold to the other, so as to act as a heat pump and do the refrigeration.  My understanding is that the power consumption of these demons is linked to the number of times that it measures the temperature (speed) of a water molecule approaching the gate, opens or shuts it, and then has to forget what it decided to do - that last part, oddly enough, is where the entropy is charged.  So just like a refrigerator, there has to be a plug in the wall to keep it running; but fortunately one expects not to hear those bazillion gates clattering open and shut every second.  And so the efficiency should be absolutely, directly comparable; but it can be written either as Carnot efficiency assuming that the demon can't power some kind of steam engine that powers the demon, or as the cost of erasing a whole bunch of bits of data.  I think someone somewhere might have explained the relationship of those two means of calculation in terms even I can understand. Wnt (talk) 18:46, 30 July 2015 (UTC)