Wikipedia:Reference desk/Archives/Science/2010 March 20

= March 20 =

Location of herpes hideout
I was reading this article about Swedes working on a herpes vaccine: and it gives me the surprising impression that the virus travels up the nerve fiber from the site of infection and hides out in the spine. Is that correct? I know it resides in sensory nerves, and I always assumed that meant the viruses would stay at the site of infection, but the article says "We found the genome of the virus in the spinal cord" and "somewhere between the mucous membrane and the spinal ganglion the infection lost its virility". 213.122.1.147 (talk) 00:57, 20 March 2010 (UTC)


 * Our article on herpes simplex virus notes that sites of latent infection do indeed include the sacral ganglia (particularly for HSV-2). TenOfAllTrades(talk) 01:52, 20 March 2010 (UTC)

Symbol for angular impulse
What's the standard or semi-standard symbol for angular impulse in physics? A few places on the internet I saw it as H, but they also seemed to denote torque or what they called moment with an M, whereas I was taught to use tau. Is H the correct symbol to use? 68.160.248.223 (talk) 03:09, 20 March 2010 (UTC)

Angular friction
I made up this expression which is supposed to calculate the amount of torque friction causes on an initially spinning object (like spinning something on a table that eventually comes to a stop). I'm not very experienced with calculus, but I felt like experimenting to see what I could come up with. Starting with $$F_f = \mu_k F_n$$, where $$\mu_k$$ is the coefficient of kinetic friction and $$F_n$$ is the normal force, I got $$d\tau_f = r\mu_kgdm$$, where $$d\tau_f$$ is the frictional torque, $$g$$ is gravity, and $$r$$ is the radius of mass element $$dm$$ from the center of mass. Integrating both sides gives $$\tau_f = \mu_k g\int rdm$$. Is this expression correct?

If this expression is in fact correct, could someone try checking my answer to the following? I tried evaluating this expression for a disk of mass $$M$$ and radius $$R$$ and got $$\tau_f = \frac{2}{3} \mu_k M g R$$. Does this seem right?

If all this stuff is correct, I will be very happy with myself. :P

Thank you for all your help!!! 68.160.248.223 (talk) 03:40, 20 March 2010 (UTC)


 * It looks right to me. Dauto (talk) 04:44, 20 March 2010 (UTC)


 * I don't see how the radius matters. First, are we talking about a round object spinning on edge, like a coin, or on it's face, like a puck ?  However, in any case, friction isn't theoretically dependent on the area of contact, but only on the coefficient of friction and the normal force (which is mass times g in both cases here). StuRat (talk) 17:52, 20 March 2010 (UTC)


 * I was referring to it being on its face, like a puck. The radius here does matter because we are talking about a frictional torque now and not a frictional force.  That's why I set up my equation to integrate the radius with respect to each infinitesimal element of mass.  I guess more generally, it should be $$\tau_f = \mu_k \int rdF_n$$ for situations where the normal force for the whole object is not necessarily $$mg$$ (i.e. - a puck coming to a stop from spinning on a ramp or in an accelerating elevator or something like that). 68.160.248.223 (talk) 18:17, 20 March 2010 (UTC)

<<>> Is it possible to excite BOTH light particles (in a small area) and the air surrounding them, with the SAME charge?
Greetings.

I have a question that has been bothering me for quite some time, and I hope somebody can answer it.

Imagine a cubic box (about 15cm X 15cm X 15cm) that is closed on 5 sides and open on 1 side.

Now, imagine that the air particles INSIDE THE BOX --to wit, the Nitrogen, Oxygen, and Argon atoms-- are all excited. And also, imagine that all light particles (photons, etc.) ENTERING THE BOX VIA THE OPEN END are also excited WITH THE SAME CHARGE as the air particles in the box (both positive or both negative).

My question is:

A.) Would this result in reflection of the light away from the box, total darkness in the box, or some other visual effect?

and B.) Can such a phenomenon be created using extant machinery and documented, scientific processes?

--Thank you for reading this. I know it's a big question, but it's also one that is driving me crazy! Pine (talk) 07:28, 20 March 2010 (UTC)


 * Photons do not carry electric charge. When an atom is in an excited state, it usually does not have an electric charge either - unless enough energy has been added to ionize it.  I think you have some confusion between energy and electric charge.  If, however, the photons have the correct frequency, they will interact with the atoms in the box, possibly resonantly.  Note that when a photon excites an atom, the atom is briefly excited, but later decays to a lower energy level and re-emits a photon in an unpredictable direction ("spontaneous emission").  This means that there will be light in the box.  Some technical terms for related processes include Atomic absorption spectroscopy, Compton scattering, spontaneous emission, stimulated emission.  If you carefully control the "box" and its resonance, you will have created a laser.  By injecting photons (energy) into the chamber, and controlling resonant frequencies, you can produce a coherent beam of light.  This is nearly impossible to do with regular atmospheric air mixtures, but works great with mixes of helium and neon, or other setups.  Nimur (talk) 09:27, 20 March 2010 (UTC)

Confused about Fourier Transform
How is a time-domain function represented in a frequency-domain function? Say we have f(t)=sin(t), a time-domain function. If we do Fourier Transform to convert it to f(v), then what do the range and domain of f(v) mean, and what do they look like?

In the article http://en.wikipedia.org/wiki/Frequency_spectrum, it's said, "A source of light can have many colors mixed together and in different amounts (intensities). A rainbow, or prism, sends the different frequencies in different directions, making them individually visible at different angles. A graph of the intensity plotted against the frequency (showing the amount of each color) is the frequency spectrum of the light. When all the visible frequencies are present in equal amounts, the effect is the "color" white, and the spectrum is a flat line. Therefore, flat-line spectrums in general are often referred to as white, whether they represent light or something else."

How does the frequency-domain graph (i.e. frequency spectrum) of a visible light look like and what does it actually mean?

Also, it's said that the values that Ff(u) (the FT function) take on are usually amplitudes and phase, both plotted against frequency. Then what is the y-value for the time-domain function?

—Preceding unsigned comment added by 70.68.120.162 (talk) 07:42, 20 March 2010 (UTC)


 * The map from time domain f(t) to fourier domain F(ω) creates a set of sinusoidal functions. F(ω) is a new function which represents the defining parameters for that set of sinusoidal functions: F(ω) gives you an amplitude and a phase for the sine wave at frequency ω.  Adding up all those sine-waves, with the correct amplitude and phase, recreates the original time domain signal.  So, the Fourier transform contains the same information of the original signal, but in a different format.  A fourier transform can transform any type of "y-value".  Scientists take frequency spectra of radio signals, light signals, height of ocean waves, comet orbital radius as a function of year (Rcomet(t)), stock price as a function of day ($(t)), and so on.
 * An NMR machine measures an RF or microwave signal as a function of time - this is the "y-value" of its signal in the time domain. That RF signal is probably around a 30 MHz-ish intermediate frequency downmixed by the machine from the actual NMR physical frequency.  This is an engineering detail you probably don't care about.
 * Light is a sort of pathological example, because we almost never think about light signals in terms of their time domain representation. (This is because they are extremely high frequency - 1013 to 1014 Hz. In other words, the oscillations are too rapid for most of our technology to work with in the time domain). But, if you wanted, you can describe the "y-value" as the time variation of the electric field for a light signal - there really is an electric field and a magnetic field, and it really is producing an effective voltage in free space, and so you can think of the time-domain signal as a function of time and space, where E is electric field strength: E(x,y,z,t). At a single point (x,y,z), you can simply consider E(t), the electric field fluctuation with time. Represented this way, light is exactly like a radio signal (except at higher frequency); so just like picking up an AM broadcast with a radio antenna, you could pick up the electric fluctuations of the light. (Technical details about building an antenna and amplifier are non-trivial - but let's ignore that issue for now).
 * So now you have a well-defined signal: E(t). It has values for all time t. Just as the mathematics describes, you can perform the time-integral of this signal, and calculate its fourier transform (or approximate its fourier transform for a short sampling window, since it is impossible to measure E(t) for all t out to ± ∞). Let's define E(ω), where ω is the frequency: E(ω) ≡ FOURIER{ E(t) } = α ∫ ( E(t) e-2πiωtdt . You know E(t); and let's assume you're an expert at calculus and/or Mathematica software - so you can calculate the integral.
 * You need to determine the value of α based on your sampling window (all this does is normalize the values to the correct amplitude; in other words, "calibrate" for the units of energy, based on how long you collected energy for E(t)). There are some standard values, standard formulae, based on "standard" integral limits. This is common if you know the frequency content you are interested in, and can easily define a sampling time based on that - see Nyquist sampling theorem. For now, we'll just assume α doesn't matter. (It really doesn't make a huge difference; if you're an engineer or applied scientist, you probably read this value off of the instrument calibration spec anyway, which accounts for other practical details we don't care about here). There are also issues about whether you are using a digital system with discrete sampling (signal processing); this forces you to use a discrete fourier transform, which is an approximation of the real, full-blown continuous version.  Most of the time, you won't care about the difference, because the NMR machine has been designed to do this approximation very accurately (by sampling fast enough for its needs, and computing the answer very quickly, using the FFT algorithm).
 * So now you have E(ω) - the amplitude of the electric field as a function of frequency, and you can describe the entire light signal in terms of its frequency content. (You can easily convert frequency to wavelength, if you prefer - that's a fairly trivial unit change, and it scales and flips the axis, because wavelength λ = 1/ω). You asked about what white light looks like in this domain. Well, it depends on how precise we're being in the definition of "white". Sunlight is usually the "golden standard" of white light - but its spectrum is actually blackbody radiation, plus a few resonances from nuclear fusion, noise, and other processes - technically, not really white noise: see a standard solar frequency spectrum). But in a small range around the visible spectrum, the amplitudes are reasonably flat - "white enough."  And if you're looking at a real (non-theoretical) source, like a measurement off a spectrophotometer, even a really good white light source with a perfectly flat spectrum has a noise component - see Welch's method and spectral density estimation for some conceptual and mathematical analysis about estimating noise of a noise-spectrum. (This can be a little abstract).
 * So what does it mean when we have a frequency spectrum? Well, it means that we have represented the same signal (the light we measured, with our antenna apparatus), but we represented it in a different way. Now we can analyze the signal in terms of the frequencies that make it up. A lot of times, like if we are looking at a light source from a chemical reaction, or if we are looking at terahertz or IR spectra from an Nuclear Magnetic Resonance machine, we will see specific frequencies popping out of the noise. Since physicists know a lot about simple harmonic oscillation, if you tell them a frequency, they can estimate the mass (or some other parameter) that would correspond to that frequency. This really helps, if you care about mass (and in NMR, you usually are using mass as a proxy for chemical or functional-group identification). So, what we did is convert the signal into a different format (in fact, a different domain, the frequency domain), where the physical process that created the signal shows up very cleanly. Nimur (talk) 08:32, 20 March 2010 (UTC)


 * Your time-domain function f(t)=sin(t) is a sinusoidal waveform (with no phase shift i.e. it goes through the origin f(0)=sin(0)=0). Note that it is periodic i.e. it repeats forever in both positive and negative directions of t. Its FT (Fourier transform) is simply a spike at the real part of frequency 2*pi/t. Now some variations: Change the phase of the sine wave i.e. f(t)=sin(t+P) and the FT spike spreads to both real and imaginary parts of the frequency. However the power Re²+Im² is unchanged. The phase shift just rotates the vector in the complex plane without changing its length (magnitude). But real waveforms don't go on forever. Taking the FT of the sine wave over a limited time causes the spike to spread and look like the Sinc function. That has distracting side lobes but they can be reduced by applying a window function explained in the article Spectral leakage with examples here. Real waveforms need not be sine waves. Often the time domain samples are treated as real values by an FT because we ignore (set to zero) their imaginary components. Note however that both time-domain samples into the FT and frequency samples out of the FT are complex i.e. numbers with two variables. Therefore the same number of variables enters and leaves the FT. All the information is preserved and the time-domain wave can be recovered by doing an inverse FT on the complex frequency samples. For light a glass prism can act as a simple analog spectrometer doing what an FT does digitally. Cuddlyable3 (talk) 15:45, 20 March 2010 (UTC)

Phase
The Wiki's Phase article says, "The phase of an oscillation or wave is the fraction of a complete cycle corresponding to an offset in the displacement from a specified reference point at time t = 0."

The definition makes no sense at all. I'm learning Fourier Transform and NMR and I'm trying to understand how phase is related to FT. Any resources that don't assume high-level math/spectroscopy would be appreciated. —Preceding unsigned comment added by 70.68.120.162 (talk) 08:23, 20 March 2010 (UTC)


 * Remember that a Fourier transform is describing a signal as an equivalent set of sinusoidal functions. You already know from basic math (or you can trust me here) that a sine wave can be totally defined by three parameters:
 * A*sin(ωt+φ), where A is amplitude, ω is frequency, and φ is phase
 * So, when you represent an arbitrary function f(t) in Fourier domain, you generate those three parameters. For every frequency ω, there is an amplitude A and a phase φ.  Many applications of spectrum analysis ignore the phase; but it contains information which is required to totally describe the original signal f(t).  In conventional NMR, you probably care more about amplitude, because your source is incoherent (probably), so your phase signal ends up being mostly stochastic noise.  Detailed analysis can extract information from that phase signal, but most physical chemists who use NMR only care about the amplitude spectrum because that can be used to easily identify chemical function groups or individual chemical elements.  Nimur (talk) 08:41, 20 March 2010 (UTC)

The definition is poorly worded. Here is another shot: ''The phase of an oscillation at a given moment is the fraction of a cycle that has been traversed at that moment, after taking away all complete cycles that have been traversed. The concept of phase requires that some point in the cycle be designated as the beginning of the cycle.'' That's still a bit awkward -- really it's hard to be both clear and precise without resorting to equations. Looie496 (talk) 17:29, 20 March 2010 (UTC)

Essential oils
In extracting the essential oils from a plant, how much oil, roughly, could you expect for a given weight? Ballpark figures are fine. Thanks. 173.179.59.66 (talk) 08:31, 20 March 2010 (UTC)


 * Juniper Oil Yield, Terpenoid Concentration, and Antimicrobial Effects on Deer  (1980), indicates the process. Depending on what you consider "raw material", (i.e., full plant; just the leaves; or a prepared substrate mash), your answer will vary wildly.  Once you get a mulched-up leaf pulp, fractional distillation yields about 1-2 microliters of juniper oil per milliliter of plant mash.  Nimur (talk) 09:10, 20 March 2010 (UTC)
 * And it seems that if you use a captive deer to digest the plant, (instead of a laboratory and a distilling machine), the yield can double or triple. Unfortunately, this requires "stomach pumping a captive deer" to extract "rumen inoculum".  Before Wikipedia, I suspect that insomniac scientists had far too few options for intellectual stimulation... Nimur (talk) 09:14, 20 March 2010 (UTC)


 * It really is impossible to say, depending on the plant itself and the method of extraction. This site claims that "200 kg (440 lb) of fresh lavender flowers, between 2 and 5 metric tonnes of rose petals and 3,000 lemons are needed to produce 1 kg (2 1/4 lb) of essential oils of lavender, rose and lemon respectively". --TammyMoet (talk) 09:16, 20 March 2010 (UTC)

That's a bit unfortunate, but thanks. 173.179.59.66 (talk) 22:59, 20 March 2010 (UTC)


 * As an aside, it's also possible to produce rose oil from benzene and ethylene, and oil of geranium from isoamyl alcohol or even isopentane (and you could get 80-90% or better yield in both cases). FWiW 24.23.197.43 (talk) 05:32, 22 March 2010 (UTC)

Relationship between Bragg planes, atoms, unit cells in crystallography
How are Bragg planes related to individual atoms in a lattice, for the purpose of X-ray crystallography? Is each Bragg plane horizontally lined by atoms? —Preceding unsigned comment added by 70.68.120.162 (talk) 08:57, 20 March 2010 (UTC)

.
 * Have you seen Miller index? This is the quantitative way to identify how a crystal plane relates to the actual atomic positions.  The diagram at right shows two variants.  The Bragg diffraction always occurs with respect to some planar alignment of atoms; depending on the geometry of the atoms in the lattice, that plane may be oriented in any particular way.  Nimur (talk) 09:02, 20 March 2010 (UTC)

determining a person's age
I've heard one can tell how old a horse is by looking at its teeth. For a tree, you can count the rings, though you have to cut it down. Is there a medical way to tell how old a human is, if their age is disputed? Some invasiveness (e.g. an X-ray) is ok, but you can't decapitate the person or anything like that. Inspiration for asking: Abduwali Abdukhadir Muse. 66.127.52.47 (talk) 19:22, 20 March 2010 (UTC)
 * Not precisely. As with horses, the younger they are, the more accurately you can tell their age. I don't think anyone could accurately determine if someone is less than a year younger than 18 or less than a year older. There are all kinds of changes happening to the body at that time, but they happen at slightly different times in different people. --Tango (talk) 19:40, 20 March 2010 (UTC)
 * You might find the Wikipedia article Bone age informative. Also . Disorders can often be diagnosed by retardation  in development: Osteo- archaeologist can also disern years of famine from bone growth rings. But I don’t have a reference for this.--Aspro (talk) 21:05, 20 March 2010 (UTC)


 * In forensic science the most common are what they call morphological techniques—estimating based on tooth or bone growth. The problem is the margin of error with adults in particular is quite large, some plus or minus 10 years. There are some new methods being investigated (but not yet widely used, I don't think) that show promise at getting more precise estimates. One is by using teleomeres (little bits on the end of the DNA in your cells that get smaller every time the cell duplicates); another is by radiocarbon dating of teeth (which only works if you were born after nuclear testing took place, and you have to know which hemisphere the person was born in; another is to look at the amount of aspartic acid in human dentin, which apparently correlates with age. The latter two are somewhat destructive—they require teeth and you can't get them back, I don't think. --Mr.98 (talk) 21:40, 20 March 2010 (UTC)


 * Well, you can tell a lot about a person by having a conversation with them, or reading their online profile, etc. I bet an accurate scheme could be created to estimate age through natural language processing or metadata analysis of a personal website.  In this controversial experiment from MIT, a computer system accurately estimated a person's sexual orientation based on metadata in their Facebook profile, using a statistical database of keywords, friend/social-network patterns, and other public data.  If sexual orientation can be uniquely identified, I wouldn't be surprised if an equivalent system could be trained to accurately estimate a person's age based on publically available social-network metadata - keywords, preferences for TV shows, etc.  But the OP was probably more interested in biological tests....  Nimur (talk) 00:20, 21 March 2010 (UTC)


 * That's all assuming they put in accurate data, though. The example that inspired the original poster is someone whose age is under dispute, who is not on Facebook, and has apparently either lied about or does not their own age. (I have to say, I'm not impressed by the MIT experiment. Half of my gay friends on Facebook have pictures of them running around Fire Island-like environments in speedos, and the other half are heavily involved in groups like "Repeal Don't Ask Don't Tell" and have "The Birdcage" as their favorite movie. This isn't rocket science...) --Mr.98 (talk) 01:28, 21 March 2010 (UTC)
 * Sure, but you could probably use "favorite TV show" as a highly correlated indicator of age-group; and narrow down based on other details... and as far as "fake" data - there's no such thing! Provided that your system trains on a good set, "inaccurate" data is a statistical (albeit, noisy) signal which can be observed and used.  Nimur (talk) 18:43, 21 March 2010 (UTC)
 * I am just not sure your system could distinguish between a heterosexual 16 year old girl and a homosexual 50 year old man pretending to be a heterosexual 16 year old girl. The latter case definitely does exist on these here internets! --Mr.98 (talk) 19:23, 21 March 2010 (UTC)
 * I don't see evidence he either lied about or does not know his own age. According to the article, his mother gave one date, his father another and the US government yet another. We don't have any evidence of what he stated his age was, if anything nor if what he stated was wrong Nil Einne (talk) 15:19, 21 March 2010 (UTC)
 * I was referring to this part: "Assistant United States Attorney Brendan McGuire informed U.S. District Court Judge Andrew J. Peck at a hearing to determine Wali Muse's age that he had told Americans he was 16, 18, 19, and 26 years old." --Mr.98 (talk) 15:27, 21 March 2010 (UTC)
 * Apologies, I missed that part Nil Einne (talk) 16:22, 21 March 2010 (UTC)

And you don't have to cut trees down in order to count their rings (we do not have an article on tree coring, but we do on ice cores and core drilling, which work in a similar manner.  DRosenbach  ( Talk 18:09, 21 March 2010 (UTC)

Solar power satellite
can solar power satellite affect ozone layer? --Siddharth9936 (talk) 20:03, 20 March 2010 (UTC)


 * Don't think so. I can't imagine any mechanism by which one would.  -- Jayron  32  20:05, 20 March 2010 (UTC)


 * No, I don't see how they could. Satellites orbit high above the ozone layer and they don't generally emit much pollutants once they are up there so they don't have a direct effect.  The amount of sunlight blocked by their solar panels is utterly negligable - so there is no effect there.  Probably the only possible effect would be from the exhaust gasses produced by the rocket while getting it up there - and that's the same no matter whether the satellite is solar powered, runs on batteries or something else.  SteveBaker (talk) 00:45, 21 March 2010 (UTC)
 * I think they are asking about the microwave beam from a theoretical solar power satellite and whether the beam could damage the ozone layer. 75.41.110.200 (talk) 01:41, 21 March 2010 (UTC)
 * I agree. And the answer is still no. Dauto (talk) 02:34, 21 March 2010 (UTC)


 * If the silicon chips for the photovoltaics are washed in chlorofluorocarbons, then that chemical can cause ozone depletion. Nowadays I think they use other fluorocarbons for washing which are greenhouse gases instead.

raw data on smoking deaths
I would like to know where I can find raw data, or something closer to it, on "deaths caused by smoking". I am particularly interested in how deaths by cancer are separated into cancer caused by smoking and cancer caused by something else, since I don't think we know a lot about the specific mechanisms that cause cancer in general.

I am not interested in debating whether smoking is harmful to health, or whether it causes cancer, etc. I believe that it is, and that it does. I don't smoke, I've never smoked, I don't want to smoke, and being around cigarette smoke irritates me (though not as much as some people, evidently). But I do wonder whether the stats are "puffed" a bit (if you'll pardon the expression), since there is so much moral overtone in a lot of anti-smoking material.

I did a Google search and went through several pages of references to stats; they are pretty uniform in what they report, but nowhere did I find a reference to, say, a literature review on the science behind the summary numbers. I am always hearing things like "450,000 deaths in the US each year caused by smoking." So from what scientific evidence do they get this number?

Ralphcook (talk) 22:24, 20 March 2010 (UTC)


 * You will find more than you probably want to know in this survey paper. It contains a lot of different sources of data and different methods of analysis. --  k a i n a w &trade; 22:51, 20 March 2010 (UTC)


 * Thanks, that is the kind of information I'm looking for, but only a small part of it. It is an 11-page paper from 1956, summarizing data from a survey of 40,000 British doctors sent out in 1951.  It compares the number of deaths in broad groups of smokers -- heavy, light, non -- and draws the not-unreasonable conclusion that people who smoke, especially cigarettes, have to expect a greater chance of lung cancer.


 * But it's a long jump from there to the quoted figures of "numbers of deaths caused by smoking", especially in a different country and culture and outside the population of this study. Does anyone have anything on the source of those figures, especially anything more recent than 45-50 years ago?


 * Ralphcook (talk) 15:52, 21 March 2010 (UTC)

Testosterone and intelligence
I heard someone make the claim that high levels of testosterone correlate with generally lower intelligence. Is this true? ScienceApe (talk) 23:03, 20 March 2010 (UTC)


 * It's not a simple matter - I suggest you read Testosterone. SteveBaker (talk) 00:39, 21 March 2010 (UTC)


 * Reading the article linked to by Steve Baker, it seems that it would be truer to say the reverse. I hate these stereotypes of masculinity. 92.29.149.119 (talk) 17:11, 21 March 2010 (UTC)


 * Part of the problem with that is the (belated) understanding that male and female brains are different. As that section of the article says, men have bigger brains but women's are better interconnected - as a direct consequence of testosterone levels.  In a computer system, both size and interconnectedness produce higher processing speeds - but some ways of implementing algorithms are better for a bigger computer while others are better for a more interconnected computer.  Hence, we find that women are better at multitasking - while men are better at 3D visualisation.  The difficulty then is how we measure intelligence and (at the heart of the problem) what that word actually means. SteveBaker (talk) 19:26, 21 March 2010 (UTC)


 * Woman being better at multitasking is a prime example of a stereotype. I do multitasking all the time - I'm a male. Although studies have shown men to be better at spatial stuff, this may be just the effect of a lifetime of practice. Nobody can really say what's innate and what is due to a lifetime of conditioning from gender roles. 92.29.149.119 (talk) 20:00, 21 March 2010 (UTC)


 * I'm Chinese, yet I'm living in Canada. I suppose that must mean most Chinese people live in Canada, right?  --99.237.234.104 (talk) 00:51, 22 March 2010 (UTC)


 * Indeed - and nobody said that men can't multitask - only that women are better at it. SteveBaker (talk) 02:06, 22 March 2010 (UTC)


 * I noticed that the claim in our article that women's brains are better interconnected is unsourced. I couldn't find a source that claims testosterone is responsible for a decreased number of dendritic connections.  --99.237.234.104 (talk) 01:03, 22 March 2010 (UTC)


 * Indeed. When my girlfriend was doing her psychology degree she told me that the only known sexual dimorphism in the brain is that women have a thicker corpus callosum, which is a) disputed and b) has no known effect on anything, despite the assumptions about multitasking that people love to leap to (due to widespread fascination with guessing at what "the difference between men and women" is, probably because thinking about it is sexy). 81.131.36.119 (talk) 19:36, 23 March 2010 (UTC)


 * Although no differences in performance at certain tasks can be proven outright; it would be foolish to suggest that the physiological differences in the brains of the genders would have no impact on their ability at tasks.


 * Why's that then? Why would that be foolish? 86.21.204.137 (talk) 14:22, 24 March 2010 (UTC)


 * Not all the differences between male and female brains are simply the result of presence or absence of T. However, back to the original question: has anyone ever done a research study in which people took an IQ test and had their T levels measured? I suspect it has been done, and it would not shock me if they found a mild negative correlation, but if so, all further interpretations founder on all the factors mentioned by the previous respondents plus a few others: the questionable ability of an IQ test to measure all forms of intelligence, and the relatively poor ability of T assays to accurately measure T production or effect. alteripse (talk) 19:38, 21 March 2010 (UTC)


 * We have no solid definition of intelligence beyond "the ability to score well in an IQ test" - and no solid grounds for saying that IQ tests measure intelligence since that is a circular argument. How could we be sure that we were measuring the right thing?  We know that male and female brains are qualitatively different - and it seems quite likely that differences in IQ test results could be more to do with the test itself measuring properties that are different in male and female brains that have little to do with whatever we regard as "intelligence" on a day-to-day, practical basis.  You simply can't make a proper scientific study of intelligence versus testosterone levels until/unless you have a proper scientific definition of what we mean by "intelligence"...and we don't.  So any results you might come up with would be debatable to say the least.  SteveBaker (talk) 02:06, 22 March 2010 (UTC)


 * Let me address the perception that testosterone lowers intelligence (rather than if it actually does). If two men have a disagreement, it's probably true that those with high testosterone levels are more likely to end up fighting, while those with a moderate level might reason out a compromise.  Thus, the "more intelligent solution" is associated with the lower levels.  However, this doesn't mean the high testosterone subjects were actually less intelligent, just that they didn't use their intelligence in this situation. StuRat (talk) 02:37, 22 March 2010 (UTC)