Wikipedia:Reference desk/Archives/Science/2013 April 28

Welcome to the Wikipedia Science Reference Desk Archives
The page you are currently viewing is an archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

April 28

why do overweight people don't get treated with Adiponectin?

i wonder if it could help in some cases. thanks. Ben-Natan (talk) 02:20, 28 April 2013 (UTC)

It's a protein, and proteins can't be taken by mouth because the stomach breaks them down. And even if somebody wants to go through the ordeal of injecting it into the bloodstream, it might be hard to maintain a steady level that way. In short, you would probably need something like an IV drip to administer it. Looie496 (talk) 02:44, 28 April 2013 (UTC)

Adiponectin#Pharmaceutical therapy lists some of the problems that would need to be overcome. Red Act (talk) 02:45, 28 April 2013 (UTC)

Actually I have just removed that whole section from the article. The text was written in 2004 (although the refs were added later), and although it was clearly written using the literature, it is equally clear that the person who wrote it did not understand what they were reading. Looie496 (talk) 03:44, 28 April 2013 (UTC)

Here is a better answer: for the same reason we dont treat obesity with leptin-- it doesnt work. Obesity is not a condition of leptin or adiponectin deficiency and therefore giving more is not an effective treatment for 99.9999% of obesity. Being a protein is no barrier-- we give lots of proteins therapeutically by injection: insulin, growth hormone, hCG, EPO, etc. alteripse (talk) 17:57, 28 April 2013 (UTC)

Looking briefly at the literature, it looks like there were some uncertain indications - it has something to do with it, but there are lots of exceptions. Because you can get at it via berberine as the article explains, and likely by many other humble means, taking it as a protein is an overly expensive approach. Wnt (talk) 20:56, 28 April 2013 (UTC)

Question about progress of time in world line?

1- Less than speed of light 2- At speed of light 3- Greater than speed of light — Preceding unsigned comment added by 75.152.167.86 (talk) 04:45, 28 April 2013 (UTC)

4. None of the above. Time does not have a speed. It might be easier to give a useful answer if you would tell us what the question is. Looie496 (talk) 04:59, 28 April 2013 (UTC)

Speed, or let's say velocity, is a function of distance and time. Something like v = d / t (example: 120 miles in 2 hours = 60 miles per hour). Which can also be expressed as t = d / v. That is, time is a function of distance and velocity (example: 120 miles at 60 miles per hour = 2 hours). Time doesn't have a standalone velocity. ←Baseball Bugs ^{What's up, Doc?} carrots→ 06:57, 28 April 2013 (UTC)

It been said that at speed of light time stop , thanks water nosfim — Preceding unsigned comment added by 81.218.91.170 (talk) 11:02, 29 April 2013 (UTC)

Heat capacity

Background:

"Early biophysical experiments revealed that the rate of induced mutations in the genetic material of cells is proportional to their radiation exposure. Extrapolating to the smallest possible dose, it might be concluded that even a single photon of UV light has the ability to cause genetic damage to the skin cell.

...An ultraviolet photon carries an equivalent nergy of about 10 eV.

You hear a proposal for the damage mechanism: that the photon deliers the energy into a volume the size of a cell nucleus and heats it up and the increased thermal motion knocks the chromosome apart in some way.

(b) Estimate the temperature rise and hence describe why or why not you think this is a reasonable proposal."

What is the mean volumetric heat capacity of a skin cell nucleus? Could I substitute the volumetric heat capacity of water? What is the most important factor to consider to answer thesecond part of the question? Plasmic Physics (talk) 06:44, 28 April 2013 (UTC)

This is a not uncommon question in undergrad physics these days, and a good one, as the photons may actually be 1800 MHz electromagnetic radiation from a cell phone, instead of your ultraviolet, penetrating your brain - some idiots think cell phones cause brain tumours.

The average density of animal flesh is about that of water as most of it IS water, so it is reasonable to use the specific heat of water, either the volumetric specific heat or the mass specific heat - they are the same in SI units. Googling "density of flesh" results in pages quoting all sorts of unlikley values, but it easy to spot common errors e.g. not allowing for the lower density of bones, air in lungs etc. However a site I find usually reliable is Engineering Toolbox - it gives the specific heat of human flesh as 3470 J/kg.K vs 4178 J/kg.K for water at 40 deg C.

What is important in the second part is that to cause chromsome damage, there must be sufficient energy to break a chemical bond - you need to know this bond energy and estimate what fraction of photons of sufficient energy penetrate the nucleus and hit DNA, DNA being by volume a fraction of the nucleus volume, and taking into account transparency of cell plasm.

In estimating the effects of cellphone radiation, most people assume the human head 100% absorbs all radiation emitted in a direction so as to enter the head - estimating the temperature rise is then trivially simple, as the heat is carried away by the blood stream (thermal capacity assumed to be roughly that of water) flowing at a known rate (and relating the resulting miniscule temperature rise, the inevitable conclusion is that cellphones simple do not have enough power to be a problem)

However you do the claculation, the actual probability of permanent damage or a tumour is very very much less as each cell has effective mechanisms to repair DNA, and faults that do result in rogue cells are usually dealt with by cell mediated mitosis, most that get past secrete odd things that trigger by inter-cell mediated mitosis, and on top of that the body has further means of supressing tumours, which can't grow anyway beyond a few mm unless they figure out how to secrete a substance that brings in blood vessels.

Wickwack 121.221.214.246 (talk) 09:18, 28 April 2013 (UTC)

That being said, this question deals with a single photon, with a known energy, and the nucleus of a cell, as opposed to the whole cell. Plasmic Physics (talk) 09:59, 28 April 2013 (UTC)

I've covered that, but in a round about way without key data, as I think you've posted a homework question. Normally I would have responded "Do your own homework" to this not-uncommon question but I know you make lots of good contributions to Ref Desk and know a thing or two, so I've compromised. Wickwack 124.182.140.226 (talk) 10:42, 28 April 2013 (UTC)

I'm not asking for the question to be done for me, so it doen't really fall under that category. I'm not familiar with cellular biology, which is why I'm getting a bit lost. Does that mean that I can substitute the HC of water for the HC of the nucleus, I thought you were refering to the whole cell? It just sounds to me like you've made the problem too complicated by introducing too many variables. It seems that, by the nature of this question, assuming perfect conditions, I must figure out what the probability is that the increase in temperature causes damage. How do I relate those two values? I don't even know which bond is being broken, or how to incorporate the value into the relation. Plasmic Physics (talk) 11:13, 28 April 2013 (UTC)

OK - so first of all, you need to calculate the volume (and from that, the mass) of a nucleus. You then need to translate the photon energy into SI units. You then use the figure for the specific heat of flesh (or water) to calculate the amount that the nucleus will be heated by the photon. You then compare that with normal temperature variations in living creatures to ascertain whether chromosomal damage will be likely. HTH. --Phil Holmes (talk) 11:59, 28 April 2013 (UTC)

Perfect! Thanks. Plasmic Physics (talk) 12:08, 28 April 2013 (UTC)

It's strange though, denaturisation (which I assume is a key factor) is only introduced much later inthe textbook. Plasmic Physics (talk) 12:12, 28 April 2013 (UTC)

It's not as simple as that. Unstated in Phil's approach is an assumption that the photon is captured somewhere in the nucleus, which is mostly NOT chromosome material. The heat energy obtained by capturing the photon spreads out thru the nucleus and is largly lost to the rest of the cell and the tissue mass generally. This means that the probability of a photon breaking a DNA chemical bond by being absorbed in the nucleus generally is very low, but the probability of a photon breaking a DNA chemical bond by actually striking a DNA molecule is very considerably greater.

The idea of comparing the nucleus temperature rise with the range otherwise experienced naturally is a good one, but unfortunately it may not be valid - that depends on the energy required to break DNA chemical bonds and the fraction of nucleus volume occupied by DNA. So, yes, you do need to know the energy required to break bonds. Really, it may turn out that the probability of a bond break from a direct strike turns out to be low enough to ignore, but I think you need to show that. However, calculations I've seen that estimate the probability of chromosome damage from cellphone radiation, which are based on knowing the bond energy, assume such "direct DNA strike" probability is negligible anyway. Note that except for skin, the extremities of limbs, and the testes, tissue temperature is held to a very precise value (in humans, 37.4 +,- 0.5 deg C as I recall) except in illness.

Wickwack 121.221.81.83 (talk) 12:23, 28 April 2013 (UTC)

OK, so back to square two. I'll assume that the genetic material is at the same temperature as the nucleus, because of thermal equilibrium; and I'll assume that the nucleus is a closed-system because of the timescale of this scenario. Is that not a reasonable method? Plasmic Physics (talk) 12:50, 28 April 2013 (UTC)

I don't think so, because the mean distance from photon absorption point to a DNA molecule is of the same order as the distance from the absortion point to the edge of the nucleus. Rather than asume a closed/isolated system, assume the nucleus is is a great sea of substance of the same thermal conductivity as still water. The temperature at the absortion point will then be above the norm by a unit impulse multiplied by a factor, and then exponentially decay back to normal as the heat flows out. The temperature at a point on a DNA molecule will rise from steady state to a maximum (which will be substantually less that the absoprtion point max temperature, and occur later), then decay exponentially. Visuallise the system temperature profile as dropping a drop of oil into a sea of oil - immediately there is a step at the drop point, which then propagates outward as an ever decreasing amplitude ring wave, after a while passing the DNA island, by which time the oil at the drop point has dropped a fair way back to normal sea level. Hopefully someone else with both a knowlege of appropriate physics and laplace transformations will respond for you. Wickwack 121.221.81.83 (talk) 13:41, 28 April 2013 (UTC)

The basics of this question are that you take the energy received from the photon, divide by the volume of the nucleus (very roughly on the order of a cubic micron; I'll let you look for a better figure), calculate how many degrees that much energy raises the temperature of that much water, and laugh it off. Because of course the photon doesn't generally "energize" all the DNA in your nucleus and cause damage to it by some hazy mechanism afterward; it interacts with one specific thymine dimer or the like. Caveat: DNA can transfer energy or current short distances, acting a bit like a radio antenna, but there's a vast amount of DNA and the transmission range, so far as I can hazily recall, is only some hundreds or thousands of basepairs, and in any case, the photon that doesn't get absorbed by such a kooky means will just fly through to the next strand and interact with it fair and square. Wnt (talk) 20:51, 28 April 2013 (UTC) Oh, and note UV and cell phones are on opposite sides of visible light - what they have in common is that DNA can interact with them and common pigments don't so much.

What temperature is required for it to damage the chromosome then? I don't need a precise answer just an accurate one, since I just need to compare values it seems. I'm not meant to determine whether the proposal as a whole is true or false, just whether the precepts of the argument is true or false. Plasmic Physics (talk) 23:03, 28 April 2013 (UTC)

I didn't look at this very carefully, but [1] gives some numbers around 5 kcal/mol for the activation energy of thymine. Translating this into a "local temperature" on the DNA should have something to do with the Boltzmann constant, but I haven't really thought over this type of scenario to make a confident assertion about how much. I think it should be pretty complicated to figure out, depending on what kinds of plasmons move back and forth along the DNA and how far, unless there's a way to define the higher temperature in a region that moves as they move? The real answer requires some heavy-duty biophysics - maybe someone else can chime in here. Looking for "dna vibrational modes thymine dimer" quickly turns up references like [2] but to be honest I haven't yet really looked them over. Wnt (talk) 03:47, 29 April 2013 (UTC)

Relativity: Moving faster than light

An electron is approaching the Earth with a velocity very close to that of light [eg, its velocity is c - g (x 1 sec)], while the Earth's gravitation accelerates the electron's velocity by an acceleration - which does not decrease (and even increases) - as far as the distance between the Earth and the electron gets shorter. What prevents the velocity from crossing that of light? Please note, that the acceleration caused by the gravitational force, is not influenced by the electron's (increasing relativistic) mass. HOOTmag (talk) 06:45, 28 April 2013 (UTC)

Keep in mind that gravitational effect from the earth diminishes at a distance. I'm thinking it's an inverse-square ratio, but I don't recall for sure. But the electron (or whatever particle) would be going so fast that it wouldn't be within the earth's gravitational pull long enough for there to be any measureable effect. And I hasten to add that I'm just trying to reason this thing - I'm sure there's a more rigorous explanation - beyond just the axiom that nothing can exceed light speed. ←Baseball Bugs ^{What's up, Doc?} carrots→ 07:12, 28 April 2013 (UTC)

Yes, it's an inverse-square ratio, but the gravitational effect from the earth increases as far as the distance gets shorter. Anyways, this effect remains rather constant (approx. g) - at sea level. HOOTmag (talk) 08:22, 28 April 2013 (UTC)

Correct me if I'm wrong, but isn't an object's acceleration due to gravity a consequence of its increase in energy? If that is so, then there is more than one way for an object to exhibit such an increase in energy, other than increasing velocity. Instead of increasing in velocity, the electron could simply be getting hotter. Plasmic Physics (talk) 07:18, 28 April 2013 (UTC)

Yes, the electron gains energy from gravity, and this energy goes to increase the momentum and energy of the electron, but only some of the increase appears as an increase in velocity. The rest appears as an increase in "relativistic mass". As the velocity approaches the speed of light, less of the momentum increase appears as velocity, and more appears as "mass". Another way of looking at the situation is to say that as the velocity approaches that of light, it takes more and more energy to produce a small increase in velocity. Please note that this is only a rough layman's view. No doubt an expert will be along soon to give a more formally-correct explanation. Dbfirs 07:48, 28 April 2013 (UTC)

Darn, I forgot about RM. Embarrasing, considering I studied cosmology. Plasmic Physics (talk) 07:56, 28 April 2013 (UTC)

I guess the old addage is applicable: "If you don't use it, you lose it". Plasmic Physics (talk) 07:57, 28 April 2013 (UTC)

Yes, the electoron's mass increases as far as its velocity increases, but the gravitational force increases as far as the electron's mass increases, so the acceleration is not influenced by the electron's mass, and remains rather constant (approx. g, at sea level), and even increases as far as the distance between the electron and the Earth gets shorter, so I don't see how your comment explains what prevents the electron's velocity from crossing that of light. HOOTmag (talk) 08:22, 28 April 2013 (UTC)

No, an object's acceleration due to gravity is a consequence of the curvature of spacetime, which in turn is due to the energy distribution of the universe (as represented by the stress-energy-momentum tensor). Temperature has no meaning for a single electron. For a large collection of particles, temperature represents the kinetic energy--i.e. the velocites--of said particles. --Bowlhover (talk) 08:21, 28 April 2013 (UTC)

I was using the very loose meaning of temperature (random thermal motion, fluctuations in its energy distribution among it's degrees of freedom). Plasmic Physics (talk) 09:12, 28 April 2013 (UTC)

FYI, 'g' does not stand for velocity, but acceleration due to gravity at sea level. Plasmic Physics (talk) 07:18, 28 April 2013 (UTC)

I didn't say that g is a velocity; g is the gravitational acceleration at sea level. However, if the electron is at sea level and its velocity is c - g (x 1 sec), then it's expected to cross that of light within one second. HOOTmag (talk) 08:22, 28 April 2013 (UTC)

If you're not saying that 'g' is velocity, then why are you attempting to subtract it from a velocity. You can't take three apples away from five oranges. Plasmic Physics (talk) 08:50, 28 April 2013 (UTC)

I was using c and g as pure numbers. Anyways, thanks to your comment, I've just replaced the expression "c - g" by "c - g (x 1 sec)". HOOTmag (talk) 09:26, 28 April 2013 (UTC)

Let's forget gravity, because that brings in general relativity, and replace it with a constant force pushing the particle. That way, we can answer the question in a purely special relativistic framework.

In layman's terms, when you exert force on an object, you give it momentum. In special relativity, the momentum of an object approaches infinity as its speed approaches the speed of light. So you can never get it to the speed of light, because you can never give an object infinite momentum.

To see what happens to an accelerating object: by definition, force is equal to the time derivative of momentum: ${\vec {F}}=\mathrm {d} {\vec {p}}/\mathrm {d} t$

In special relativity, the momentum of an object is equal to ${\vec {p}}={\frac {m_{0}{\vec {v}}}{\sqrt {1-v^{2}/c^{2}}}}$

Without loss of generality, let's assume that the momentum, force, and velocity are all in the x direction. Then we can take the time derivative of p, and after some algebra, find that:

$F_{x}=m{{(1-v^{2}/c^{2})}^{3/2}}a\,$

where a is the acceleration.

So the closer v is to c, the higher Fx needs to be to achieve a given acceleration. As v approaches c, Fx needs to be infinitely large to achieve a given acceleration, or alternatively, the acceleration would be close to 0 for any force. --Bowlhover (talk) 08:21, 28 April 2013 (UTC)

I haven't been talking about a constant force, because the acceleration caused by any constant force decreases as far as the electron's (relativistic) mass increases. I've been talking about the gravitational force, by which - the acceleration caused - is not influenced by the electron's (increasing relativistic) mass. HOOTmag (talk) 08:36, 28 April 2013 (UTC)

Obviously, the problem is that your assumption is false. The acceleration caused by gravity does depend on the relativistic mass. Or more precisely, the apparent acceleration due to gravity expressed as the change in velocity of the particle in the rest frame of an inertial observer appears to decrease as the speed of the particle approaches the speed of light. For a particle at rest it is certainly true that:

$m\,a={\frac {dp}{dt}}=m\,g$

Where m is the mass, p is the momentum, a is the acceleration, and g is the gravitational field. However, only the second equality holds in the relativistic limit. Specifically, if you start in the frame of the accelerating particle then:

${\frac {dp}{dt}}=m\,g$

However in the inertial frame of the lab the right hand side becomes is:

$m\,g\rightarrow {\frac {m_{0}\,g}{\sqrt {1-v^{2}/c^{2}}}}$

The left hand side goes to:

${\frac {dp}{dt}}\rightarrow {\frac {d(mv)}{dt}}={\frac {m_{0}\,a}{(1-v^{2}/c^{2})^{3/2}}}$

Setting these two pieces equal tells us that given a constant gravitational force, the apparent acceleration measured in an inertial frame is not constant, but rather:

$a=(1-v^{2}/c^{2})\,g$

Hence the effect of a constant gravitational field on relativistic particles is what one would expect at low speeds ($a=g$), but decreases towards zero as the speed of the accelerated particle approaches the speed of light. Dragons flight (talk) 11:30, 28 April 2013 (UTC)

Oh, that was instructive. Thanks! Btw, according to your explanation, it seems there's something wrong in our article negative mass - which points out: "two objects of equal and opposite mass would produce a constant acceleration of the system towards the positive mass object". If I'm right, then would you care to fix this? HOOTmag (talk) 14:14, 28 April 2013 (UTC)

It doesn't need fixing because it (implicitly) assumes non-relativistic speeds. Dauto (talk) 15:12, 28 April 2013 (UTC)

The acceleration is independent of mass and velocity, but that acceleration is not dv/dt but dφ/dt, where φ is the rapidity. This follows immediately from the equivalence principle (it's actually you who are accelerating upward while the falling object moves inertially, v ≈ φ for low velocities, and the "velocity addition formula" for rapidities is actual addition). From v = tanh φ, you get dv/dt = (1 - v²) dφ/dt, which is equivalent to Dragons flight's final equation. -- BenRG 01:48, 29 April 2013 (UTC)

Thank you for your comment. Anyways, just as velocity is not rapidity, so acceleration is not the derivative of rapidity. Please notice, that I'd been asking - why the electron's "velocity" does not cross that of light - assuming the "acceleration" (i.e. the derivative of velocity) is constant. If I had used the term "rapidity" instead of "velocity", I would have asked - why the electron's rapidity does not become complex - assuming the acceleration (i.e. the derivative of velocity) is constant. Dragons flight answered that my assumption was wrong: the acceleration (i.e. the derivative of velocity) is not constant, because it depends on velocity. That's right. To sum up: The whole issue of rapidity has nothing to do with my question (about velocity) nor with Dragons flight's answer (about the dependence on velocity). HOOTmag (talk) 06:35, 30 April 2013 (UTC)

What species of bird is this?

http://commons.wikimedia.org/wiki/File:Seabird,_Elwood_Beach.JPG Flaviusvulso (talk) 09:47, 28 April 2013 (UTC)

Looks like a juvenile Kelp gull, Larus dominicanus, scroll down to third image. Richard Avery (talk) 13:44, 28 April 2013 (UTC)

Actually, to me it looks more like a juvenile Pacific gull. Cf. the image here, and note the shape of the bill. Deor (talk) 13:47, 28 April 2013 (UTC)

I concur. Richard Avery (talk) 14:56, 28 April 2013 (UTC)

Data capacity of vinyl record

How much data capacity (in MB equivalent) is there on a vinyl LP record? 86.179.119.114 (talk) —Preceding undated comment added 12:52, 28 April 2013 (UTC)

Only a rough figure can be given. Assuming you mean the digital data equivalent of the recorded audio, it can be estaimeted from the playing time (15 to 20 minutes per side), the maximum playback frequency with a typical cartridge (15 kHz) requiring a sample rate of 30 kilosamples/sec, and the dynamic range possible (about 70 dB), which implies 15 bits/sample, the total is about 800 MBits. This ignores the low level white noise inherent in both the recording and playback process, which is not applicable to digital methods. Most records were recorded with a dynamic range compressed, 50 dB being typical, equiv to roughly 300 MBits.

I'm not aware of LP's ever being used to store data, but in the early days of personal computers, some software was distributed on 45 RPM singles, using kansas city modulation to be compatible with systems using compact cassettes for storage. The capacity using both sides was then only 140 kbits.

If you chose a modulation method to best use the characteristics of LP's and cartridges to store digital data, say mPSK modulation, you could get about 40 MBits using both sides (taking 15 kHz as the maximum recording frequency and (as there would be no bass frequencies) minimum permitted groove spacing.

Wickwack 121.221.81.83 (talk) 13:23, 28 April 2013 (UTC)

Thanks, but why such a big difference between 800 MBits and 40 MBits? 86.179.119.114 (talk) 18:53, 28 April 2013 (UTC)

The 800 MBit is what need to (reasonably) faithfully reproduce the analogue signal of the music on the LP. The 40Mbit is the amount of digital data you can expect to reliably retrive from the LP. One is a bit like taking a high-res photographic image of a nicely set page from the Gutenberg Bible, the other is like the ASCII representation of the text printed on that page. Which one is relevant for you depends on the application. --Stephan Schulz (talk) 20:11, 28 April 2013 (UTC)

Theoretically those amounts should be almost the same. The channel capacity of the record is both an upper bound on the amount of data you can encode in it and a lower bound on the amount of data needed to encode it, and Shannon's theorem implies that you can get arbitrarily close on both sides. However I suppose they could be very different if data stored on the record needs to be recoverable even from a damaged record using cheap equipment, while the digitized record needs to cope with perfect condition and high-quality equipment. -- BenRG 02:22, 29 April 2013 (UTC)

Theorectically, those amounts most certainly should be different - you have ignored a few important factors. Shannon's theorem only applies with random noise. An analog channel has not just random noise, it has non-linear distortion (also known as amplitude distortion). The human ear is very tolerant to amplitude distortion, indeed second harmonic distortion is percieved by some people as an improvement to recorded music! Indeed, some rock and roll music of the 1950's and 1960's owes its' impact to judicously added distortion! In electronic circuits it is easy (when you know how) to reduce amplitude distortion to extremely low levels by the technique of negative feedback (comparing output with input to derive a correction signal) - 0.01% being routine in modern quality stereo equipment. However, no feedback or error correction method is possible for the transducing from disc to electric signal in the pickup cartridge. Distortion is introduced by cartridge non-linearities, elastic deformation of the vinyl under stylus forces, and tone arm tracking error (cartridge axis not maintained at exactly right angles to groove tangent) and tracing distortion (which arises because a stylus of finite width in a sine groove will not trace a sine path). A total distortion amounting to several percent and mainly second harmonic is typical, perhaps 1% in equipment of the highest quality, and it varies from unit to unit and brand to brand, so you cannot pre-compensate.

Analog sound reproduction such as LP record replay systems are also subject to phase distortion and non-level frequency response. This also varies from brand to brand and so cannot be pre-compensated for. The human ear is quite insensitive to non-level frequency response, and one gets used to gross errors anyway. Even very good cartridges may show peaks and nulls amounting to 4 or 5 dB, i.e., +,- 70%! The ear is just about totally insensitive to phase errors. However, any coding system that attempts to exploit the amplitude range of LP records as well as the frequency response is inevitabley sensitive to frequency and phase errors.

People were insensitive to, and even liked, the distorted sound of LP records, but the absense of all these distortion modes is why CD's sound much more transparent. Not necessarily "nicer" to all people, but definitely more transparent.

Digital error correction systems using extra parity bits, checksums, etc, can handle random (white) noise very well, but the distortion mechanisms of LP's and replay systems are systematic but very complex and unique to each unit, and techniques for handling it have not been developed.

Wickwack 120.145.90.62 (talk) 04:37, 29 April 2013 (UTC)

Motion Detector Sensor Type Most Used in Door Openers at Stores

What kind of sensor technology is most commonly used in the ubiquitous door opening sensors at most stores? 67.163.109.173 (talk) 13:10, 28 April 2013 (UTC)

Microwave (12 GHz) doppler radar. Wickwack 121.221.81.83 (talk) 13:11, 28 April 2013 (UTC)

Incidentally these are the same sources used in radar guns (usually the older X-band, but sometimes K-band) which is why they set off radar detectors. Shadowjams (talk) 14:55, 28 April 2013 (UTC)

And I've lived my whole life thinking they were just plain old Passive infrared sensors. Vespine (talk) 00:41, 29 April 2013 (UTC)

Cheaper ones, like you'd find in a home security system are. But I can attest that most commercial doors set off my radar detector. Shadowjams (talk) 03:25, 29 April 2013 (UTC)

There was an episode of Mythbusters where they got a motion detector to not detect them as they walked through a room by holding up a large white sheet and slowly passing by. I forget the details of their reasoning as to why that worked, but the white sheet consistently worked, they made it through the room without detection a couple of times. 20.137.2.50 (talk) 13:49, 29 April 2013 (UTC)

Yes, I've tried the same experiment with PIR detectors (and it works -- even a coat or jacket can be enough to shield the radiated heat from the body). It doesn't work with the radar type that set up a standing wave in the room. Dbfirs 14:27, 30 April 2013 (UTC)

Transport of materials within neurons

In most cells, diffusion is the primary means of transporting materials (proteins, rna, nutrients, etc.) within the cell body; also, cellular motors are used to move cargo from one place to another within the cell.

However, many neurons are orders of magnitude larger than the typical cell. Do neurons use additional mechanisms to move materials back and forth to/from the more distant parts of their anatomy? For example, it seems unlikely that a cargo packet could be trundled for a distance of 1 meter on a microfiliament.

Royjacobus (talk) 13:46, 28 April 2013 (UTC)

It does indeed use cellular motors. Dauto (talk) 15:05, 28 April 2013 (UTC)

We have an article on this, axoplasmic transport. It's pretty technical, so feel free to follow up if you have difficulty understanding it. (The basic story is that for fast transport cargo is trundled along a sort of railroad made of microtubules.) Looie496 (talk) 16:36, 28 April 2013 (UTC)

It's also important to note that there is protein synthesis at the synapses of these nerves too, so not everything has to be transported down the entire length of the cell. Fgf10 (talk) 20:16, 28 April 2013 (UTC)

Biology of exercise - two, very very nice questions

Why do some people (not many i assume) report the Release of Lactic acid after 5-10 minutes of Aerobic exercise (Mainly running) ?. is there any biological reason in the claim that such condition may be related to Breathing problems? Regards. Ben-Natan (talk) 16:28, 28 April 2013 (UTC)

Yes. See Lactic acid#Exercise and lactate for a full explanation of what happens. Briefly: aerobic exercise always produces lactate, which is removed by oxidation. Without enough oxygen the lactic acid builds up to cause acidosis, which is painful. This can happen to anybody, even a highly trained athlete, if the level of oxygen intake doesn't match the intensity of exercise. Looie496 (talk) 16:44, 28 April 2013 (UTC)

biology

Are females feel more hot in summer than in male,as they refer to wear small size clothes? — Preceding unsigned comment added by Titunsam (talk • contribs) 16:41, 28 April 2013 (UTC)

I would imagine that how "hot" somebody feels is more a factor of body size (are we talking somebody who's 100lbs or 450lbs?) and the type of activity they are partaking in. The idea that women prefer to wear fewer clothes than men is extremely culturally specific. Check out this photo as an example of what I mean. Fashions also change a lot, if you want to look a little closer to my home, you can find a man walking around in short-shorts in Magnum PI, and very well-covered women in the 1890's Falconus^{p t c} 17:50, 28 April 2013 (UTC)

Note, do you know why dual temperature settings were invented for cars? It's because on a hot day, men like their side of the car cool and women like it warm. What does that tell you? It tells me that men have lots of body hair. Viriditas (talk) 01:01, 29 April 2013 (UTC)

It's not that women feel hotter. Western women have learnt to tolerate feeling colder in order to comform to fashion. And probably to show more skin because deep down they like us men looking at them. I've known my lady to expect to feel too cold while wearing a sexy miniskirt to a winter party, but she wouldn't dream of wearing something longer. That's what girls do. Wickwack 121.221.28.17 (talk) 01:29, 29 April 2013 (UTC)

Source please?I am not convinced the genders don't perceive temperature differently, or at least find different temperatures more comfortable. Every woman I have lived with (a handful) prefers the heating on higher then I do, my wife used to sleep with her electric blanket ON until I managed to convince her how it's a really bad idea. I've had this conversation with many friends and they pretty much always say that their wives take hotter showers (as has pretty much every girl I've every been with) and if they had their way they'd have the heating on full. I think they're lizards ;) Now I'm NOT saying that I'm not a biased sample and my perception is all confirmation bias, but it's enough for me not to take a comment like the above on faith, with no reference. Vespine (talk) 06:52, 29 April 2013 (UTC)

Actaully, that might be arguing the same thing, my wife likes it hotter in the house so she can walk around in a singlet, wher as I'm perfectly comfortable with a jumper on inside. Vespine (talk) 06:54, 29 April 2013 (UTC)

We are not disagreeing. My lady, the very same one who will put up with being cold wearing a miniskirt to a party, also likes one more blanket on the bed than I do. She also likes the house warm so she can wear less. Left on their own women might act rationally, but if there's someone to look, they follow fashion. As to how they like aircon or heating, additional factors come into play. In my experince women are more cost conscious, but they only consider their own costs. We have long agreed that I pay for electricity and she pays for the phone bill. That tends to make her consider the cost of phoning. With previous girlfriends, we had agreed on the other way around - in my experience that causes the phone bill to increase and electricity to be saved. My lady and I have gone long range hiking (~300 Km) on forrest trails, where we had to compromise between comfort at night and weight carried on our backs. I tend to go for light weight, she goes for comfort at night. Wickwack 121.215.16.186 (talk) 08:01, 29 April 2013 (UTC)

Once upon a time I was told that women feel the heat less than men because they have an extra layer of fat which helps them keep warmer in cold conditions. This may well be an old wives tale. --TammyMoet (talk) 10:23, 29 April 2013 (UTC)

The extra fat is in certain specific places though - not all over. I once exclaimed to my lady who was feeling cold that she had one extra layer of clothing covering part of her - to whit, her bra. She didn't think that made much difference though. One aspect that may contribute is that men need and burn more calories, weigh more but are not as much different in body surface area. So maybe there is a reason why females need an extra blanket as reported above. For instance, I weight 80 kg, my lady only weighs 57 kg, and we are both the same height. I eat about twice what she eats. Wickwack 121.221.219.237 (talk) 12:48, 29 April 2013 (UTC)

Im still very dubious about this claim: Western women have learnt to tolerate feeling colder in order to conform to fashion. Don't forget we put clothes ON in order to conform to fashion in the first place. Many primitive peoples wear very little and while I'm sure they have notions of what is "fashionable", they certainly don't think their women have LEARNED to tolerate feeling colder in order to conform to anything. Vespine (talk) 23:26, 29 April 2013 (UTC)

That's what we call the exception that proves the rule. Primative people generally didn't wear much because good clothing was not availble to them. And, whlie you are correct in saying any and all humans want to look fashionably good, they had a lot more to think about just surviving than we do. Wickwack 124.178.140.70 (talk) 05:20, 30 April 2013 (UTC)

There is no such thing as a "primitive" person. In general, each culture worldwide has an extremely rich and complex social life, with profound knowledges of their environments, and generally spend a lot of time focused on things that are not "about just surviving" (whether for religious purposes or leisure). As a matter of fact, many, if not most, of the people that you would be tempted to dismiss as just "primitive" survive by working a far shorter work week than somebody in the US or Europe would have to just to cover basic living expenses. Sure, their ways don't make much sense to us... But our almost complete lack of understanding doesn't mean that they are "primitive". If people who live up above the Artic Circle, with the resources available there, can figure out how to make really good clothes, anybody could if sufficiently interested in doing so (a lot of places where people don't have a lot of clothes have a lot more resources available for such a purpose). Falconus^{p t c} 13:24, 30 April 2013 (UTC)

Please also see Exception that proves the rule. I think wickwack has it exactly backwards. Vespine (talk) 23:11, 30 April 2013 (UTC)

Falconus needs to get out more, and away from his laptop. One example: Australian aborigines - their best idea for winter clothing before whiteman clothes became available was often nothing much more than a good coating of animal grease. Maybe a kangaroo skin draped over the shoulders if one is available. And yes, they typically spent MUCH more than whiteman's 8 hours a day getting food, and old photos typically showed the normal signs of insufficient and unbalanced diet - real skinny arms and legs and bloated abdomen. They didn't grow anything and only method of cooking they knew was baking (bry the animal with hot rocks in the ground), so the usable food available in most areas was pretty limitted. Coastal aboringines that mastered fishing did rather better, but they were a minority. Eskimos could make good clothes because they would die if they didn't - another exception that proves the rule. If traditional ways were easier or gave a better lifestyle, you'd get most Australians (white or black) follwing them - the reality is almost none (white or black) follow traditional ways, though sometimes a small number (white and black) do an occaisonal bit of traditional for a bit of "roughing it" fun. Wickwack 121.215.55.166 (talk) 23:41, 30 April 2013 (UTC)

Well of course; by the time those photos were taken, the Europeans had taken all the good land and forced them into the really poor land. Throw in a bit of smallpox etc and it's no wonder that that would make life difficult. Check out The Original Affluent Society by Marshall Sahlins. "Traditional ways" were not adopted by the Europeans because it wasn't their tradition. Let's look at it the other way, why were so many Native Americans so reluctant to embrace the colonial lifestyle? Your argument cannot only work one way... Falconus^{p t c} 00:50, 1 May 2013 (UTC)

Aborigine clans are thought to have been somewhat evenly distributed in Australia before Europeans came, due to territorial and clan strife, though clans were larger in better areas. The problem of Europeans taking the good land and forcing indiginous people off really only applied to southern coastal areas, so your thoughts on old photos are not right. American Indians have their own problems, and I don't pretend to know much about them, but Australian aborigines abandoned their traditional ways (except for some occaisonally doing some traditional things for fun as I said, Europeans sometimes do it too) with alacrity. They are not stupid - European large scale commercialism gives a far better lifestyle with a heck of a lot less work. Europeans are not stupid either. It has nothing to do with traditions - it is simply blindingly obvious to Europeans and aborigines that a subsistence way of life is too hard for too little gain. White or black, you'd have to be a complete nutcase to consider roaming the bush virtually naked in all weathers to occaisonally get lucky and kill a kangaroo instead of working in an office or factory 9 to 5 and get high quality food from the local supermart and buy or rent a house. Unfortunately alcoholism, government handout mentality, and "stolen generation" impacts have rather ruined things for aborigines in recent decades. However, my argumant does work, in that the fact that aborigines today all wear clothes in European fashions shows that the reason they did not traditionally wear clothes was only because clothes were not available then. Wickwack 121.215.55.166 (talk) 01:31, 1 May 2013 (UTC)

This is going way off topic. Your original claim was that western women in particular have learned to take clothes off (or wear smaller clothes) and tolerate it, for reasons of fashion and because society has taught them that they should like being looked at by men. I call bogus to this claim, I think you have it precisely backwards. The fact is as a race we have learned to put clothes ON, not take them off, we aren't BORN with clothes. If anything, western women have UNLEARNED to completely cover them selves up. Further, you completely missed my point about "primitive" people, what I was saying is that they go about with little to practically no clothes on and it is NOT considered "sexy" or "taboo" or "fashion" and they certainly DON'T do it so their men will look at them. To some of them there is absolutely nothing whatever noteworthy about bare breasts. This aversion to the human body is also something learned by more modern societies. Like the Catholics who went around gluing fig leaves on the little cherub penises because they were so offensive. Vespine (talk) 02:19, 1 May 2013 (UTC)

I agree with nearly all that you said, but you have two different concepts confused, or almost confused.

Concept A: Primative people did not wear full clothing because for them such clothing was not available; conversely, advanced people took up wearing clothing because clothing was invented and became available - it short they took up wearing clothes because 1) they could, and b) it met a need (keeping warm in cooler climates). Concept A has nothing to do with wanting to look nice or sexy.

Concept B: Once wearing clothing was taken up, clothing could be adapted or modified, and thus create fashions. And it happens that, women seem to be more keen on fashion than men (though both are). And it it seems that women tolerate discomfort (eg being cold or whatever) in order to look attractive, to a greater degree than men. It's culturally driven as others have said, it's really no different to certain central west African tribes whose women compromised their confort, flexibility, and health by artificially lengthening their necks with neck rings.

You are quite correct in saying that aversion to the human body is an artificial learned/cultural thing. There is a full range - primative peoples think nothing of bare breasts (and even bare pubes); Americans & Australians think nothing of a bit of cleavage or bare lower legs; to certain Arabs just the sight of a bare female angle is a turn-on (only because with their style of dress an ankle is not often seen). None of this has anything to do with, nor disproves, that women accept some discomfort in order to look how they want to look (Concept B); not does it prove or disprove that it has become possible to alter clothing to look how you want; nor have anything to do with primative people not wearing clothes because they could not (Concept A). In many cases, such primative people took up other cultural or fashionabel decoration that was possible, such as shell necklaces, the proverbial bone through the nose, or cords wrapped round the wrist etc - They had a primative life, but a brain no different to us. Wickwack 120.145.218.220 (talk) 02:47, 1 May 2013 (UTC)

I don't disagree with your analasys of my post but I'm not mixing anything up, you are just stuck completely off topic. The topic is "females feel more hot in summer than in male,as they refer(sic) to wear small size clothes". As far as I understand it, you said NO, the only reason women wear smaller clothes is for fashion and other reasons dictated by culture. All my posts have been a refutation of this proposed answer and I don't think anything you have said so far has furthered your hypothesis. Saying that women go through discomfort for fashion does NOT address the question at hand, because as you yourself noted, some women put MORE clothes on for fashion. All I was saying is that wearing skimpy clothes is not necessarily explained by fashion and cultural norms because in cultures where clothes don't play a significant role in fashion and flesh does not play a big role in attracting mates, women can and do still wear little clothes. To me, it sounded like you were dismissing the question by providing an alternative explanation to the evidence presented by the OP, however that evidence, even if false, does not disprove the opening hypothesis. All I'm saying is that in my experience, most women i've lived with do in fact behave as if their perception of tempterature is different, they take hotter showers, prefer more sheets on the bed, etc... I freely admit I might be the outlier, especially since I have never showered or slept in the same bed as other men, but in driks cooler conversation I've also found that most men feel the same way about their female significant others. I could still be wrong, but it's going to take more then a mere assertion to convince me, which is finally, why I asked for a SOURCE, this is the reference desk after all. Vespine (talk) 04:14, 1 May 2013 (UTC)

No, I'm on topic. If you go back and read what I said, you'll see that I said my experience too is that in situations where conforming to fashion does not apply, such as number of blankets on bed, yes, women seem to want to be warmer, or their perception cf men is that their surroudings are cooler - just as your dinks cooler colleagues report. You are not an outlier. As I said, that may be due to some subtle hormonal thing, or merely due to men on average being heavier and having a greater metabolism rate. My point is that when fashion does apply, their desire to look good overides their comfort - and looking good/attractive/sexy means showing more skin. Perhaps you should ask a few females the right question. As to sources, yes, an authoritive source or two would be good, but I'm not aware of any such on this topic. It would be difficult to find. Wickwack 120.145.218.220 (talk) 05:21, 1 May 2013 (UTC)

OK! So, i think we can boil it down to: you think that fashion can explain why women wear skimpier clothes and I was saying I don't believe that answers the question, which you never really disagreed with. Glad we finally got to the bottom of it. :) Vespine (talk) 05:51, 1 May 2013 (UTC)

This thread would be incomplete without someone mentioning the "Theiss Titillation Theory" (see William_Ware_Theiss#Theiss_and_Star_Trek). It states: "The degree to which a costume is considered sexy is directly proportional to how accident-prone it appears to be". Theiss did actual research on what young men find sexy - and the amount of skin left uncovered was not so much a factor as how likely it appeared that the costume might fall off at any moment! Either women (and their fashion designers) have failed to understand this principle - or they prefer skimpier clothes for reasons other than attracting young men. SteveBaker (talk) 14:12, 2 May 2013 (UTC)

Filariasis and cancer

What will happen if a person suffering from filarisis in one part of body ,sudddenly suffering cancer in that region? — Preceding unsigned comment added by Titunsam (talk • contribs) 17:21, 28 April 2013 (UTC)

You get a person who has both filariasis and cancer. Do you have any reason for asking this seemingly inane question? (I've changed the section title, as "biology" is not informative enough to be useful.) Looie496 (talk) 00:34, 29 April 2013 (UTC)

Difficult to answer, yes - inane, no. Please do not savage a poster until you have made a genuine effort to figure out whether there is something behind his question. Both conditions involve the lymphatic system, so it would be possible that the lymphedema caused by filariasis could change, say, the pattern of metastasis to lymph nodes. But I assume the OP is more interested in something like [3] where nematode-derived substances might put the immune system on a hair trigger to fight cancer. I think that figuring out adjuvants that could be injected in/near a tumor to make it more immunogenic should be a very productive approach for cancer research and I hope to see more of it. Wnt (talk) 13:27, 29 April 2013 (UTC)

Wnt, for someone who is so wont to answer medial questions you seem oddly loath to check the user contributions of obvious sockpuppet trolls. μηδείς (talk) 02:46, 30 April 2013 (UTC)

Alright, I see now that whoever it is asks a lot of questions and doesn't write them very well (possibly language or age issues) but luckily I am completely ignorant of how you would diagnose sockpuppetry from that. I don't see a troll here either - I see questions that are at least as answerable as the others we get. And I never even thought about filariasis and cancer together before, so I count this as quite a good one. My feeling is that if your first impulse when seeing a question like this is to investigate the person posting it, answering Refdesk questions may not be the best hobby for you. Wnt (talk) 16:49, 30 April 2013 (UTC)

accuracy of a poll

A poll of religious self-identification was published by the US Census Bureau at [4]. It uses a methodology used repeatedly for the same pupose in the past, covering over 54,000 responses.

It states also that the estimates for groups under 75,000 total were aggregated to reduce sampling errors.

Christian Scientists were estimated to have 339,000 self-identified adherents in 2008. This figure is disputed at WP:RS/N with comments such as "it is very very likely to be off by a factor of three" and "A factor of 10 is what I would expect from the sample size."

I took a slew of probablity and statistics courses many years ago :) and I think the objections fail to hold water. Positing that 81 people out of 54,000 self-identified as part of a group, the simplistic use of "standard deviations" suggest more than a 50 to 1 probability that the figure is closer than 3 to 1 (about 6 standard deviations) and a bot worse than 1000 to 1 that it is off by a factor of 10 (8+ standard deviations). Are my stats off on this sample size? The claim was made that the "error rate" should be about 0.5% and this would mean the figures for any population under 1+% of the total are essentially meaningless. (" For very small groups like the Christian Scientists, this completely dominates the result.") Thanks. Collect (talk) 23:56, 28 April 2013 (UTC)

Their arguments are silly. Any error would have a proportionate effect not an absolute one. The standard deviation would be about nine and as you say a factor of three out would be six standard deviation which has of the order of 1 in some hundred millions chance of happening. If there is any error of that order it is not due to low figures and statistics - it would have to be due to people not reporting their religion correctly or the sample being very non random. And with such low figures I guess you would have some people ticking the box by mistake for instance. Or perhaps they have some religious reason for not reporting their religion, I don't know all the strange beliefs of them all. Dmcq (talk) 00:40, 29 April 2013 (UTC)

I'm afraid I can't parse the sentence that starts "Positing...", but I agree with you that the comments you cited seem to be off the mark. If you observe 81 hits out of a large N, the expected error in the number of hits is on the order of 9, so you would rarely expect to see an error larger than 20%. It's worth noting though that this only applies to sampling error. If there are measurement errors that cause people to be identified as Christian Scientists when they really aren't, or vice versa, then all bets are off and anything can happen. Looie496 (talk) 00:58, 29 April 2013 (UTC)

The statistic is sound; it corresponds with User:Collect's calculation; but more importantly, it has a very reliable source - the Census Bureau - and a link to a government document describing the methods and data in detail. Anybody who disputes the statistic needs to bring counter-evidence that is more reliably-sourced. The purveyors of dissident opinions may produce any calculations they like; whether the calculations are right or wrong is irrelevant. (In fact, even though our calculations are right, that detail is also irrelevant! For the purposes of Wikipedia articles - we don't need or want to cite our personal calculations in articles). What matters is, the Census Bureau says the number is X, and that's what our article should say - including a link to the source. Our article should mention the error as reported in the source - not as we independently estimate the error. And if anyone believes the data to be in error, the burden of finding evidence is on them - they cannot put forth independently-derived calculation here on Wikipedia. If the dissident editors are excellent population statisticians or census researchers who know more than the Census Bureau, then the activist editors will not need to edit Wikipedia to make their case - they can simply publish their results elsewhere in a reliable source, and that data will eventually get cited and trickle into the article! Nimur (talk) 03:34, 29 April 2013 (UTC)

Sorry, but you all have been hoodwinked by Collect, who has been less than honest about the source. The survey was NOT performed by the Census Bureau, as the Census Bureau makes clear. Here is the original survey, from which the Census Bureau extracted the data for their table: [[5]]. And as that source makes abundantly clear, the standard error is under 0.5 percent for the FULL sample.

As for the sigma squared rule that Collect uses, it works great with jellybeans in freshmen statistics class. It doesn't work at all with real people in a real-world survey on religion, for the reasons outlined here: [[6]]. Kudos to Looie496 for at least suspecting that there was more than mere sampling error at work here. Dominus Vobisdu (talk) 06:27, 29 April 2013 (UTC)

I actually read the study, and some of the peripheral information, before commenting - surprise! And if you carefully review my post above, you'll see that I never said the survey was performed by the Census Bureau. My line was, "...the Census Bureau says the number is X"... which they do. How they obtained that data, and who actually performed the data collection and processing, are secondary details that the interested reader can investigate. The fact that the phone-calls and number-crunching were performed by a third party does not change the fact that the Census report endorses those numbers.

And let me reiterate: our article should not use my interpretation, or User:Collect's numbers, or User:Dominus Vobispu's mathematical analysis. None of us are "reliable sources," no matter how much we believe we are right. Our article should report the numbers, the methodology, and the error parameters as published in a cited, reliable source. Nimur (talk) 07:44, 29 April 2013 (UTC)

All that might be a good (or not ;-) argument on WP:RS/N, but this is the science reference desk, where the underlying problems od the survey are more relevant than Wikipedia policies. I found the Gulliver Effect article a real eye-opener. --Stephan Schulz (talk) 07:59, 29 April 2013 (UTC)

@Nimur: Whether the Census Bureau "endorses" them or not does not make them any more accurate that the publishers of the study say they are themselves (and indeed, the Census Bureau does NOT claim that). You asked for a more reliable source, I gave it to you. The survey itself. And I gave you a link to why simple freshman statistics does not work with a data set like this, and why the survey designers reported such a high standard error. In short, the survey is not accurate enough to support the claims that Collect wishes to make in the article in question. He's going way beyond where the survey designers themselves went. Dominus Vobisdu (talk) 08:09, 29 April 2013 (UTC)

I'm surprised that the discussion has gone on for this long, in 3 different venues, with no mention of some basic concepts from high school statistics. Neglecting all types of sampling bias, there are 2 main sources of error when estimating the population of tiny minorities:

1. The inherent sampling error, due to the inherent randomness of the sampling process. Since everyone is either a Christian Scientist (CS) or not, the sampling error follows a binomial distribution with standard error sqrt(npq) = 9. If the best estimate for the number of CS adherents is 340,000, the 95% confidence interval would be roughly 270,000 to 410,000.

2. False positives, also called Type I errors, due to the fact that a small minority of non-CS would mistakenly identify as CS. If only 1 in 1000 non-CS Americans identified as CS due to sloppiness or malice, and only 10,000 real CS existed, there would appear to around 300,000 CS--far higher than the actual population. Dominus' link has a good explanation of this effect, but for some reason, decides to call it "the Gulliver effect". In actuality, this is a well-known effect that all freshman statistics classes cover (or should cover), and we have an article on it: the false positive paradox.

Here's a real-world example from our article on Type I and type II errors: "The US rate of false positive mammograms is up to 15%, the highest in world. [...] As a result of the high false positive rate in the US, as many as 90–95% of women who get a positive mammogram do not have the condition"

So a test that's 85% accurate overestimates the number of breast cancer patients by a factor of 10-20, due to the false positive paradox. Here, being CS is analogous to actually having breast cancer, the mammogram is analogous to the survey, and a false positive is equivalent to someone claiming to be CS when they're actually not.

To estimate the true number of CS adherents, we need two numbers: the probability that a CS adherent would claim to be CS, and the probability that a non-CS adherent would claim to be non-CS. These correspond to sensitivity and specificity respectively. We can probably assume that the former is close to 1. The latter is also close to 1, but even if it were a tiny bit less than 1--say 0.999--that corresponds to hundreds of thousands of false positives.

With respect to the dispute at hand, I have no idea how the Census Bureau, or the ARIS, accounts for false positives. Maybe they have some way of estimating specificity; maybe they don't account for it because it's too difficult to do. If they don't account for it, I would certainly be suspicious of their figures for extreme minority religions. --Bowlhover (talk) 08:11, 29 April 2013 (UTC)

The figures are for reported Christian Scientists, i.e. ones who say they are to the questioner, not those who really are.

Looking at the survey it seems it was mainly over the phone so I would hope they compensated a bit for that because of groups like the Amish Mennonites. They have more than fifty religions so one would have to have a particular factor about Christian Scientist that made it more prone to errors than other groups or for it to be really really tiny or for there to be a particularly large false reporting error. People reporting their own religion wrong would be much higher than 0.5%, so I'd guess the 0.5% is the person filling in the entries getting the reported religion wrong. One would need to allocate half those errors to the Christian Science box to get a factor of 3 out whereas randomly one would expect one in fifty and really less as they would be more careful over the minority ones. The thing that does strike me though is I wonder if there are a load of people out there who think they are Christian Scientists because they are Christian and scientists ;-)

As to Wikipedia use, Nimur is right. Government statistics offices are reliable sources and people complaining here are not reliable sources. These statistics may not have been collected by them but they are a reliable source in their own right and moreover have been 'sanctified' by the US statistics office. Dmcq (talk) 09:44, 29 April 2013 (UTC)

In short, ignoring the idiotic(") claim that I "hoodwinked" everyone by calling the Census Bureau the "publisher" here, the statistics are as I stated, the publisher is a reliable source, and those who asserted that an error factor of three was "very very likely" and that an error factor of ten was likely, were ill-schooled in statistics and probability theory, and my MIT degree is still safe (along with all my more than a dozen courses in math) <g>. Thanks. Collect (talk) 11:19, 29 April 2013 (UTC)

You have a very selective perception. Your statistical argument about random sampling error was sound. But that is not the only source of error - as the linked paper above explains, you must typically take into account 3-5% of reporting error for surveys, which has a much amplified effect on small subgroups. --Stephan Schulz (talk) 11:24, 29 April 2013 (UTC)

Eh? You thought For very small groups like the Christian Scientists, this completely dominates the results (shown to be exceedingly fallacious above by everyone) and DV asserted a "factor of ten was likely". Now you are reduced to the fact that every single poll ever made has -- that a huge number of people simply lie. The history of this poll, however, seems to belie that claim - when they interviewed so many people why would people specifically lie about Christian Science? It is not exactly the most prominent religion to lie about, and I would, in fact, suspect as many would deny as would assert the religion. "Wiccan" had a huge increast (factor of forty) over 18 years -- but I suppose no one would lie about that one? Pagan and Spiritualist also had huge increases -- has it occurred to you that religious self-identification is not the same as being a subscriber to a specific magazine? Absent any coherent rational otherwise, we use the statistics the government prints. Now read the books <g>. Cheers. Collect (talk) 12:20, 29 April 2013 (UTC)

Again, no. Random sampling errors are the one class where a statistical analysis using standard deviation is really useful in cases like this. A pollster making the cross in a wrong location is not a random sampling error. An interviewee misunderstanding the question who answers that he is both a Christian and a scientist and thus confusing the pollster is not a random sampling error. From the Gulliver paper (and there sourced to Takalkar et al, 1993): "We can safely say from this and preceding research, however, that even the least ambiguous self-report variables tend to exhibit between 3% and 5% error." (emphasis mine) Takalkar et. al also write "Overall, responses involving simple, permanent, and factual data (e.g., age) show the smallest proportions of error (1-10%), while response tasks involving either multi-category responses or clearly involving psychological factors produce the greatest numbers of errors (15-50%)". --Stephan Schulz (talk) 12:50, 29 April 2013 (UTC)

The results were for reported numbers. If the reported religion was placed in the wrong box 0.5% of the time then with the at least 50 boxes then we would expect about 5 more in that box if they were placed completely at random. However as I said before they would be more careful about the smaller ones and anyway there were probably more boxes. The numbers there are probably a good reflection of what the people said their religion was, if you want to find errors they would be in other factors like how did they correct for telephone polling if at all and what's the difference between what people say their religion is and some truth about what their religion is. Dmcq (talk) 12:41, 29 April 2013 (UTC)

what's the difference between what people say their religion is and some truth about what their religion is: [7]. You found the elephant in the room. Congratulations! Dominus Vobisdu (talk) 12:48, 29 April 2013 (UTC)

And by the way the Gulliver effect is a real thing even if I believe it isn't a major factor here. It explains why survey forms try and classify queries and put the more common possible answers first rather than sticking everything into alphabetic order. People often just tick a neighbouring box by mistake. Dmcq (talk) 13:04, 29 April 2013 (UTC)

The poll appears not to rely on "boxes" but uses multiple questions and answers to get the most accurate possible result. Including questions aimed precisely at the issue DV seems to be fascinated by. Apparently it is a pretty sophisticated poll at this point. Collect (talk) 13:38, 29 April 2013 (UTC)

• For the benefit of people who come across this discussion in the future, I hasten to try to correct some misconceptions promulgated here:

1. To repeat, the religious identification poll is emphatically not government data. The US census isn't allowed to collect data on religion.
2. Although the government includes the poll in its Statistical Abstract of the United States, that does not necessarily indicate some sort of unique endorsement of the poll. The next page in the 2012 Statistical Abstract presents a second table of data on religion: a table from the National Council of Churches Yearbook of Churches that lists denominational membership, largely based on numbers reported by the denominations. Christian Science isn't in the Yearbook table. If you compare the tables for other denominations, you can see big differences between the numbers. For example, for the LDS church, ARIS lists over 3 million adherents in 2008 and the Yearbook lists membership of over 6 million. For Episcopalians, ARIS lists 2.4 million adherents in 2008 and the Yearbook lists 2 million members. For Catholics, ARIS lists 57.2 million adherents and the Yearbook lists 68.5 million members. For Assemblies of God, ARIS lists 810,000 adherents and the Yearbook lists 2 million members. For Jehovah's Witnesses, ARIS lists 1.9 million adherents and the Yearbook lists less than 1.2 million members. This wild variability tells me that no one number can be treated as true -- and statistical error related to sample size cannot possibly explain all of the observed differences. It's likely that a lot of people self-identify with one religion in a confidential poll but publicly affiliate with another.
3. The link to the ARIS report provided above is a link to a summary of the study. Don't bother to look for data for Christian Science in that linked report, as Christian Science doesn't show up in the summary (Christian Science was lumped with several other small Protestant denominations in the data summary). --Orlady (talk) 14:22, 29 April 2013 (UTC)

Look, this page is for scientific questions, and I think I gave a decent explanation of false positives and why they might impact the survey. That said, I do think we should include the ARIS numbers in our Christian Science article. I think it's the most authoritative estimate we have, and if other sources disagree, that means nobody knows the actual number and we should report the range (like this: 100,000 - 340,000). --Bowlhover (talk) 18:09, 29 April 2013 (UTC)