242: Unknown Quantities

[The Random One]

STOP PICKING ON LAUREN GUYS

Seriously, though, the idea is that science, as opposed to, say, religion or folklore, is not based on irrefutable dogma. This, to a layman, may sound as a point against science - why would I want to believe in something if even the people who live out of it don't do? But she defends that this lack of irrefutable belief is exactly what makes science trustworthy.

Her examples of scientific theories that were abandoned may have been poor, (and indeed, thousands of years ago mankind already had the sailing technology that was advanced enough that not taking into account the earth curvature might land you in a Lost episode) but there are plenty of good examples. I think it was Aristotles who, not long after essentially creating the prototype of what we call the scientific method, went on to declare that living things were born out of nothing, showing as proof the fact that alligators would 'spawn' out of the mud on the Nile. It looks laughable now, but it was only some two hundred years ago that it was definitively ruled out as a theory, and it took some insane experiments to do so. A similar thing happened to the theory of the aether, which is lumped with hollow earth theories now but was only proven wrong in the early 20th century.

Her example of placebo may have been concise, but keep in mind that placebo is one of the weirdest things in medicine right now. Some other fun facts about placebo: it is much more effective if it's a double-blind test, i.e. the doctors/nurses don't know if they're giving placebo or real medicine. (So a placebo would be more likely to be effective on you if the person giving it to you thought it was real medicine... without any input from yourself.) And I remember a major study that essentially discovered the effectivity of placebo knee surgery is not that much lower that actual knee surgery. If you think that a few centuries ago medicine was still thinking the best way to treat the ill was to remove their blood and replace it with mercury, what might that say about modern medicine?

Martian Methane and Dark Flow more or less speak of how amazing are things we know so little about. About the latter, I'm not the first to make a connection between megastructures such as galaxies and microstructures such as atoms.

OoBE are one of those interesting things in which no scientific progress can be made because scientists say it's bullshit and believers think scientists are evil people who are trying to disguise the TRUTH from the public. The general theory is that people can remember what happened while they were still conscious and their brain 'retcons' it so that they think it happened while they were clinically dead. I remember a scientist's experiment suggestion, that in a hospital they had a TV screen that displayed a different image, changing every minute. This way, if a pacient could relate what the screen was showing during a time period in which his brain shouldn't be able to receive input, it would certainly be an unexplained phenomenum. If not, case closed.

Lauren

Lauren

I still Admire you

 

[ucciolord1]

A brilliant and insightful look into the scientific process.
Of course, however, all our knowledge is rendered moot when confronted with Space Cuthulhu.

 

[gigastrike]

Now, in the dark flow part, when they say "structures" they mean...nebulas? Solid objects? Some space-time thing we can't comprehend? ...Cthulhu?

 

[rddj623]

Exceedingly interesting and informative article.

 

[Awexsome]

I enjoyed reading it. I hope to learn more about the dark rift/space Kthulu stuff in my career but alas that will have to wait until I'm farther than a Freshman in college.

 

[Jordi]

We're arguing semantics here, mostly of what "proof" and "Fact" and "knowledge" and "truth" mean, but Lauren is more or less right in the statement that nothing can be ascertained via the scientific method. All scientific experiments are evaluated using the null/alternative hypothesis testing method. The null is always that no effect exists, and the alternate is that SOME effect exists.

A test statistic is calculated, a function of desired variance and error variance, and if the ratio of "good" to "bad" variance is acceptable based on a (usually arbitrary) predetermined critical value (usually equivalent to a 5% chance of exceeding such a difference value in the absence of the phenomenon in question) then we "reject" the null as being implausible at the percent confidence in question. We never say "the alternative hypothesis is true," we speak in terms of "there is less than a 5% chance that a difference this extreme or more extreme would be detected in a representative, properly obtained sample this size, given that no effect were actually in existence." That 5%? Completely arbitrary. Nothing magical happens at a 5% likelihood. It doesn't represent much of anything except (yes, really) the likelihood of a bunch of tea-drinking fops sipping chamomile in an English garden in the 1920s guessing wrong as to whether their tea was poured into milk, or vice versa.

On top of this, recent ventures into the study of power analysis has proven that a disturbing proportion of tests lack statistical power (defined as 1-the likelihood of failing to detect an present effect) to back up the conclusions they purport to have reached. The (equally arbitrary) cutoff of acceptably low power is .8, and the average study had a power around .3, even if the study turned out to be statistically significant. Sample size is often anemically low such as to render meaningless seminal findings published in respected journals.

Any book on statistics or research methodology STRONGLY advises against using any such words as "proven" to describe any scientific study, and encourages theory-driven repeated testing of any phenomenon, with adequate sample size appropriately taken from the population in question (as opposed to a bunch of college freshmen who happen to be in the same class), before any strong conclusions can be reached.

In a publish-or-perish world, these technicalities are often overlooked at the expense of getting articles published, but we are right to question the validity and "truthiness" of any of a number of scientific claims, even those purported to have been conducted under strict, rigid experimental control.

I was going to say something similar. Science can in some cases be a bit fragile, especially if researchers build on each others' work without really checking it. However, I'd say most of the well-known theories are fairly robust, because they have been tested very often.

Anyway, I enjoyed the article. I would like to add the quest to understand what makes us intelligent (and perhaps replicate it) to the list.

Also, I was wondering about the part of the article that mentions that we cannot see X far into space, because light hasn't traveled there yet. Now bear in mind that I only know some high school physics, but isn't that inconsistent with the Big Bang Theory and/or the idea that nothing can travel faster than light? I mean, if everything originated in the same place, than how can some structures (for instance the ones supposedly causing Dark Flow) outrun light?

 

[Bobzer77]

We're arguing semantics here, mostly of what "proof" and "Fact" and "knowledge" and "truth" mean, but Lauren is more or less right in the statement that nothing can be ascertained via the scientific method. All scientific experiments are evaluated using the null/alternative hypothesis testing method. The null is always that no effect exists, and the alternate is that SOME effect exists.

A test statistic is calculated, a function of desired variance and error variance, and if the ratio of "good" to "bad" variance is acceptable based on a (usually arbitrary) predetermined critical value (usually equivalent to a 5% chance of exceeding such a difference value in the absence of the phenomenon in question) then we "reject" the null as being implausible at the percent confidence in question. We never say "the alternative hypothesis is true," we speak in terms of "there is less than a 5% chance that a difference this extreme or more extreme would be detected in a representative, properly obtained sample this size, given that no effect were actually in existence." That 5%? Completely arbitrary. Nothing magical happens at a 5% likelihood. It doesn't represent much of anything except (yes, really) the likelihood of a bunch of tea-drinking fops sipping chamomile in an English garden in the 1920s guessing wrong as to whether their tea was poured into milk, or vice versa.

On top of this, recent ventures into the study of power analysis has proven that a disturbing proportion of tests lack statistical power (defined as 1-the likelihood of failing to detect an present effect) to back up the conclusions they purport to have reached. The (equally arbitrary) cutoff of acceptably low power is .8, and the average study had a power around .3, even if the study turned out to be statistically significant. Sample size is often anemically low such as to render meaningless seminal findings published in respected journals.

Any book on statistics or research methodology STRONGLY advises against using any such words as "proven" to describe any scientific study, and encourages theory-driven repeated testing of any phenomenon, with adequate sample size appropriately taken from the population in question (as opposed to a bunch of college freshmen who happen to be in the same class), before any strong conclusions can be reached.

In a publish-or-perish world, these technicalities are often overlooked at the expense of getting articles published, but we are right to question the validity and "truthiness" of any of a number of scientific claims, even those purported to have been conducted under strict, rigid experimental control.

I was going to say something similar. Science can in some cases be a bit fragile, especially if researchers build on each others' work without really checking it. However, I'd say most of the well-known theories are fairly robust, because they have been tested very often.

Anyway, I enjoyed the article. I would like to add the quest to understand what makes us intelligent (and perhaps replicate it) to the list.

Also, I was wondering about the part of the article that mentions that we cannot see X far into space, because light hasn't traveled there yet. Now bear in mind that I only know some high school physics, but isn't that inconsistent with the Big Bang Theory and/or the idea that nothing can travel faster than light? I mean, if everything originated in the same place, than how can some structures (for instance the ones supposedly causing Dark Flow) outrun light?

Well someone can definitely offer a better explanation to this than I can as I too am still in secondary school and am not even doing physics = P. But while light may have reached those darkest depths of the universe it still has to return to us. Light has been there, (un-less for some reason I don't know the universe can expand faster than light) but it still has to travel the billions of light years back to earth.

 

[PlasticTree]

Nothing new for me, but still a nice read.

 

[Sayvara]

You make some decent points, but I will defend the author on one point, science can't actually prove. To prove something, you have to show that it is correct in all possible conditions. Since an infinite number of conditions exist, it is impossible to test anything under all conditions. The most we can do is to gather evidence that support hypotheses and theories, or disprove them by finding conditions in which they don't work.

Oh I'll go even further than what you are saying and claim that nothing can be proved, ever, because tomorrow, for unknown reasons, all conditions may change.

Much of our science relies on the laws of nature being constant and unchanging. This is an assumption that cannot be validated as true for the future, because we havn't been there to look yet (as far as we know).

Ponder hyopthetically that we live in a The Matrix like simulation of things. then things such as the constant of gravity is but a couple of clicks away. Just go to "Control Panel" -> "Physical constants" -> "Gravity" and punch in a new value.

But as James Randi points out in his lectures: arguing like that pretty much moot because we have to draw a line somewhere and accept some assumption, or it becomes impossible to live your life, or do research for that matter.

So science doesn't say "this is verified as correct"... it says "Under the following conditions... the theory has been verified".

I should add one thing... that science does indeed make predictions about what will happen when we try something under new conditions. That is what science is all about. Noone knows for sure how long it will take for my cell phone to fall to the floor once I drop it from my desk. But the scientific theory about gravity makes a prediction, to name a simple example.

/S

 

[Jared]

If you think about it. There is so much about our own world we don't understand it's scary at times...

 

[AdamG3691]

Dont be silly about space Ct'hulhu
everyone knows that the galactic nuclear terror is actually Azathoth...
(anyway, dead ct'thulhu waits at rl'yeh dreaming, not in space )

Ct'hulhu Fta'ghn everyone

 

[Quiet Stranger]

Okay so Placebo effects are mind over matter (your mind plays tricks on you) and out of body experiences? More like very real dreams and hallucinations

 

[Tyrannowalefish_Rex]

Friggin awesome stuff. News about dark matter is so much fun.

 

[Leaper]

I think Mass Effect 3 should have something to do with those massive superstructures that are (thearetically)pulling matter to them

 

[Lordtommy]

Great Article! It makes me happy to see that The Escapist likes to publish intellectual pieces on Modern Science instead of merely focusing on video games .

Thank you Ms. Admire for spending the time and energy to look into these various scientific subjects and bring them to light for the rest of us.

I'm a tad wary of the placebo effect example in this article however. From the description, it sound as if there was a possibility the reduced pain was in the last treatment may not have come from the placebo effect, but from excess morphine present in the body from previous treatments. That said, however, the placebo effect is a very interesting subject for study, and does seem to hold practical uses, especially in the medical field.

I hadn't heard of this "Dark Flow" before either. Wouldn't it be cool if it was so dense it messed with space-time, allowing someone to use it to travel through time? That is, given it doesn't smash you into a human crepe.

..which would then be eaten by space chtulhu.

 

[ItsAPaul]

Dark flow = reapers. They're already the space cthulus of mass effect.

 

[Zerbye]

I'm a tad wary of the placebo effect example in this article however. From the description, it sound as if there was a possibility the reduced pain was in the last treatment may not have come from the placebo effect, but from excess morphine present in the body from previous treatments. That said, however, the placebo effect is a very interesting subject for study, and does seem to hold practical uses, especially in the medical field.

Though the placebo study described in the article is a review paper that is not available to read for free, the neurochemical link between the body's natural painkillers (endogenous opioids) and the placebo effect has been demonstrated in many papers. The 1999 article below is one example from the same authors who wrote the aforementioned review (this one you can access for free). You might find it interesting, as it contains details that address your concerns.

http://www.jneurosci.org/cgi/content/full/19/9/3639?maxtoshow=&hits=10&RESULTFORMAT=1&author1=Benedetti&andorexacttitle=and&andorexacttitleabs=and&andorexactfulltext=and&searchid=1&FIRSTINDEX=0&sortspec=relevance&resourcetype=HWCIT

 

[Dhatz]

I have been knowing for many years that we are far from knowing all we would need to construct matrix even if we had the hardware.

 

[Dhatz]

I don't get why would we know what is there, or why do you think of specific things