July 24, 2012
Random Science Trivia
Don't pretend this hasn't been keeping you up nights.
2. Does this mean that Republicans were right to threaten to shut down the government over the debt ceiling hike?
Next time you are locked in a fierce negotiation, whether it's over a salary, a house sale or a business deal, it may not be a bad idea to make a few subtle, or even overt, threats, new research suggests. When negotiating, you're more likely to get what you want with a threat than by getting angry. ... both anger and threats led to concessions. Threats, however, were more likely to lead to a concession, though, because the research found that a perceived poise came with people who were making them. Threats, such as walking away from the negotiation, were seen as being most effective late in the negotiation process.
Of course calmn resolve may not matter much if the media characterize everything you do as unreasonable or motivated by anger.
3. If it doesn't work, maybe you're doing it wrong:
Even if all of the premises are true in a statement, inductive reasoning allows for the conclusion to be false. Here’s an example: Seventy-five percent of humans have brown eyes. John is a human. Therefore, John has brown eyes. That logic doesn’t work in the scientific method because it would be false 25 percent of the time.
A better "Therefore" conclusion would be, "There's about a 75% chance that a representative human selected at random would have brown eyes. But if John is a real person, the chance that he has brown eyes has more to do with his genetic heritage than with the frequency of brown eyed men in the general population."
I could tighten up the wording even more, but I'm on my lunch break.
4. Which doll does your 6 year old daughter identify with most?
Across-the-board, girls chose the "sexy" doll most often. The results were significant in two categories: 68 percent of the girls said the doll looked how she wanted to look, and 72 percent said she was more popular than the non-sexy doll.
Women whose firstborn infants have a high birth weight may have more than double the chances of having breast cancer decades later in life, a new study suggests. Researchers found that women with big babies — those weighing in the top fifth of babies on a growth chart, or more than 8.25 pounds — had a risk of breast cancer that was 2.5 times higher than that of women with smaller infants. "We were surprised at how strong this effect was," said study researcher Dr. Radek Bukowski, a professor of obstetrics and gynecology at the University of Texas Medical Branch in Galveston. "We were not expecting a large baby to be that strong a predictor of breast cancer."
6. The Blog Princess must be a cheap friend.
Posted by Cassandra at July 24, 2012 01:02 PM
TrackBack URL for this entry:
John and the brown eyes: yes, a lot of confusion about probability theory comes from trying to apply it to things that have already happened. Events that have already occurred, for instance, are neither likely nor unlikely. Individuals are real; conclusions about individuals based on the collective characteristics of the groups they belong to are probabilistic and speculative, subject to confirmation by reality.
Until you check John's eyes for yourself, you can use probability to guide your expectations and perhaps make plans for what color of tie to buy to match his eyes. Once you've checked his eyes, though, your expectations become meaningless no matter how carefully you amassed your data and performed your calculations. I think the essence of bigotry is that it refuses to allow facts to refute expectations -- just as the essence of junk science is the refusal to let facts refute theories.
Posted by: Texan99 at July 24, 2012 02:09 PM
What you're proposing as part of the solution is actually the introduction to one of the hardest problems in physics. It's a pretty big problem for philosophers, too. Enjoy! :)
Posted by: Grim at July 24, 2012 02:32 PM
Well, if statistics (inductive reasoning) has no place in Science does that mean we can dispense with the whole Climate Science thing, which is a wholly (hole-y/Holy?) statistical based endevour? Apparently, the whole "Settled Science" thing isn't even science.
Of course, there goes most of our medical research too, but, oh well, sacrifices must be made.
Posted by: Yu-Ain Gonnano at July 24, 2012 02:32 PM
#1. Can the stats be trusted? Well, I'm not holding my breath.
#2 With few exceptions, I've never negotiated anything that I was unwilling to walk away from if not satisfied.
#3. The first three comments are all over this one.
#4. Makes my heart heavy...
#5. I was in the 11 pound range. Belated apology mom.
#6. Meaningful friendships do not come with ledgers. And some of the best gifts do not cost a penny, fortunately.
Posted by: bthun at July 24, 2012 03:18 PM
Sure, deductive reasoning gets all the attention and credit for being the most logical way to solve for X, but it requires that you begin with at least two true statements. I mean, if you've already got "All A are B" and "All B are C," unless you're a moron, how hard is it really to get to "All A are C?" It's staring at you in the face! Inductive reasoning,by contrast, has no such stodgy requirements. With inductive reasoning, you can begin with premises that are utter garbage and yet still arrive at the correct answer. For example "All A are B; All Bees are Seas; therefore, all A's play baseball for Oakland." That's logic!
Posted by: spd rdr at July 24, 2012 03:46 PM
Gimme a minute to catch my breath from laughing so hard...
Posted by: Yu-Ain Gonnano at July 24, 2012 04:26 PM
....some of the best gifts do not cost a penny, fortunately.
Friendship is the best gift of all :)
Posted by: Cass at July 24, 2012 05:00 PM
Threats, whether overt (like walking away), or more subtle ("That's not good enough": the "do more or else" is left unstated because the other person can often think of worse consequence than you would have) have been taught in most negotiation classes for a very long time.
My other favorites is using cash and "If I give...then take". I bought a car at nearly $2,500 off Kelly Blue Book that way. The seller had too many toys and too many bills. He had already lowered his list price $1k because it hadn't sold after 30 days and got another $1,500 by telling him I could write a check that day. I knew I was asking for a lot, but he didn't have to wait for me to get approved for a loan or go to a lawyers office to close or change my mind while waiting for that to happen. And I told him upfront that if he didn't feel the trade of time for money was the right deal for *him* that I understood and wished him luck. Turns out, he did need the time more than the money and as I had cash on hand, time was no object to me.
Posted by: Yu-Ain Gonnano at July 24, 2012 05:18 PM
With inductive reasoning, you can begin with premises that are utter garbage and yet still arrive at the correct answer.
Years ago I took a test that supposedly measured my inductive and deductive reasoning skills. Can't recall which one I was better at, but I suspect it was inductive.
I think that's why they told me I should be a lawyer!
Posted by: Cassandra at July 24, 2012 05:35 PM
It's never a good idea to try to apply the principles of quantum mechanics to very large, slow objects, such as human beings. I don't pretend to understand how zillions of tiny probabilistic blurs resolve themselves into concrete events at the human scale, but I do know that talking about the odds of a coin toss that we've already completed is an abuse of both language and thought. What we call the laws of probability are a model with a lot of assumptions built in. You can get crazy results if the assumptions are unstated, misunderstood, and applied to situations they don't fit. That's how you get people talking about the probability of the next coin toss as if it were affected by the results of the last. You have to look at the problem one way if you're thinking about the two coin tosses before they've happened, and another if you start considering them halfway through, or your theory will start giving answers that don't match the real world.
Posted by: Texan99 at July 24, 2012 05:49 PM
...a lot of confusion about probability theory....
A classmate of my wife's told their statistics professor, in all seriousness, that her (the classmate) having kids and her (classmate) mother having kids were independent events.
Of course, the counter to this is bthun's procedure of not engaging in the negotiation unless you can ignore the threat. Stay focused. On the other hand, if you have to listen to the threat, you're not negotiating, you're begging.
Posted by: E Hines at July 24, 2012 05:53 PM
"I think that's why they told me I should be a lawyer!"
And here I thought the one reason it was because you couldn't keep your opinions to yourself.
And loved to argue....the two reasons why you should have been a lawyer.
Or control your inner Alpha Bitch Goddess...the three reasons.....
Posted by: DL Sly at July 24, 2012 05:55 PM
...you get people talking about the probability of the next coin toss as if it were affected by the results of the last.
But it is. There are only a finite number of coin tosses available in the Universe. Each realized toss subtracts from that population and alters, irrevocably, the balance of the options remaining.
Posted by: E Hines at July 24, 2012 05:57 PM
Please tell me you are being absurd.
Posted by: Yu-Ain Gonnano at July 24, 2012 06:36 PM
...the odds of a coin toss that we've already completed...
The question in quantum mechanics is one about what it means to 'complete the toss.' As you'll discover by reading the article, the fact that it has already happened isn't actually what determines the outcome. The fact that it has happened already and the result was recorded by a machine doesn't collapse the waveform. It's only observation by a mind that collapses it.
That's really interesting. And since it is necessary to use probabilistic approaches in quantum mechanics, it means that problem of induction that you're talking about gives rise to some difficult conceptual problems.
Posted by: Grim at July 24, 2012 07:11 PM
Not even buying it a little bit, though I did read the article and have read many others much like it over the years.
Probability is a model we apply when we don't understand a mechanism, but we do have some statistical information about how apparently similar events turn out on average. At extreme sizes and speeds we run into the additional problem that we can't even discern an outcome without affecting the event, as well as the possibility that what we normally mean by "happened" or "didn't happen" doesn't even apply in that context. That's not true at macro levels, and it's a different issue from the absurdity of treating events that have already been observed to occur as if we could apply probability theory to them retroactively.
My husband's contribution to this exchange is that all he knows about probability is that the dice hate him. I can only reply that "God does not play at dice," and therefore that God loves him. Q.E.D.
Posted by: Texan99 at July 24, 2012 07:35 PM
We often apply probability when we understand the mechanism perfectly, and run into the same issues. For example, roulette is a very simple physical model, involving balls of known weight and wheels of known size. The odds are so predictable that it's very difficult to make money off of it without scaling your bets (which is why there are table limits in casinos, by the way -- not because they aren't willing to take more of your money, but to keep you from scaling bets on repeated spins of the wheel. It's the one predictable way to beat the house, if they don't screen for it).
By the way, the problems of relativity that you're citing here actually map pretty well without any need for probability. You can describe them with geometric forms that are precise; you don't need waveforms that collapse, and there's no "jump." It's different at the micro level, and it's really interesting as to why it is.
Posted by: Grim at July 24, 2012 10:06 PM
Einstein was wrong. God does play dice. And I think He sometimes has a loaded set. But to me, the creepiest evidence that this quantum physics stuff isn't bunk is the classic double-slit experiment. Rather than try to ham-fistedly explain this, I'll let this cute animated video explain for me:
Posted by: MikeD at July 25, 2012 08:24 AM
Not only does God play dice with the universe, he throws them underneath the couch where you can't see the result.
Posted by: Yu-Ain Gonnano at July 25, 2012 09:10 AM
"Not only does God play dice with the universe, he throws them underneath the couch where you can't see the result."
Posted by: Al E. at July 25, 2012 09:31 AM
Niels Bohr was supposed to have answered: "Quit telling God what to do, Albert."
Posted by: Texan99 at July 25, 2012 02:30 PM
Flip all the coins you wish, and roll the dice as many times as you like, entropy always wins.
I think... well maybe...
Posted by: Allen at July 25, 2012 03:41 PM
Three Statisticians go hunting and find a deer. The first takes a shot and misses a foot to the right. The second takes a shot and misses a foot to the left. The third jumps up triumphantly and shouts "We got him!".
Posted by: Yu-Ain Gonnano at July 25, 2012 04:58 PM
Statisticians do it with Confidence, Frequency, and Variability.
Posted by: Yu-Ain Gonnano at July 25, 2012 04:59 PM
Speaking of cool stuff that I totally do not understand:
In theory, quantum computers should be able to perform certain kinds of complex calculations much faster than conventional computers, and quantum-based communication could be invulnerable to eavesdropping. But producing quantum components for real-world devices has proved to be fraught with daunting challenges.
Now, a team of researchers at MIT and Harvard University has achieved a crucial long-term goal of such efforts: the ability to convert a laser beam into a stream of single photons, or particles of light, in a controlled way. The successful demonstration of this achievement is detailed in a paper published this week in the journal Nature by MIT doctoral student Thibault Peyronel and colleagues.
If this is for real (and why shouldn't it be?) then we on the threshold of an amazing leap (dare I say "quantum") in mankind's ability to update its Facebook pages across an infinite number of universes while simultaneously ripping every season of "Happy Days" onto our Sony Watchmans - in 4D!!! Truly, what hath God wrought?
Posted by: spd rdr at July 25, 2012 05:30 PM
"Statisticians do it with Confidence, Frequency, and Variability."
Von schmall problem dere. Zee outlier to zee exscheption to zee rhule is dat constants aren't and variables don't.
Und dat plays zee dickens vit de vormulas... Ase von might exschpect.
Posted by: Al E. at July 25, 2012 05:38 PM
"then we on the threshold of an amazing leap (dare I say "quantum") in mankind's ability to update its Facebook pages across an infinite number of universes while simultaneously ripping every season of "Happy Days" onto our Sony Watchmans - in 4D!!!"
Yah, und not to mention having zee ability to haf youhr iLaz photon emitter accidentally slice zee tail off zee Virgin Airlines Near Earth Orbiter...
A schlight malvunction vich vil be vixed in a vuture virmvare oopdate.
Posted by: Al E. at July 25, 2012 05:50 PM
Statisticians do it with Confidence, Frequency, and Variability.
Quantum physicists have the superposition, though. As long as nobody looks.
Posted by: E Hines at July 25, 2012 05:53 PM
@Grim. Dude, that link was less than satisfying. Took far too long to basically describe 'Schroedinger's Cat'---all possible events exist until measurement occurs and then all states must collapse into one experienced state---- and resorts to 'but we don't know for sure, therefore it's true!'-kinda-sorta reasoning.
There's a reason why it's called faith and not fact, Good Sir. Lack of proof: It's a test. It's easy to believe in the Almighty when you know for a fact he's there and you'll be punished, and that isn't proof of your love of God or your fellow man as it's far more a test of your own fear of punishment(test design, it's a BIG FURRY DEAL). Far harder to hold to The Way if you are susceptible to being told it's a hoax and might get away with it(and I say this as someone who has and does stray quite often).
#4 pisses me off. And I think it's parents without courage that's allowing this to happen (vice causing it---that's a whole other and distinct barrell of @55holes). Your child does not have to dress up like a club girl. you let her, and in too many instances I've seen lately support financially doing so.
Posted by: ry at July 25, 2012 06:41 PM
There've been different opinions about that question, Ry. Kant thought as you did, at least in his earlier work: that there were walls that kept reason from getting to a proof of God, perhaps precisely to require faith.
I tend to take a view that is somewhat older, and was held by Christians like Aquinas, Jewish philosophers like Maimonides, and Islamic ones like Avicenna. They were all keenly aware of the problem of knowing God, given what we can reasonably infer about God's nature and our own limitations. But you could come to know some things about God, they thought, by knowing his works. If God is the first principle of creation -- that which defines the rules by which the world operates -- then learning about the rules of the world brings you closer to understanding God.
For them, the issue wasn't a test of loyalty, but of reason. If you wanted to get closer to God, you should try to understand the order of the world.
Posted by: Grim at July 25, 2012 09:21 PM
...perhaps precisely to require faith.
Hmm, I seem to remember Douglas Adams having something to say about that:
"The argument goes something like this: 'I refuse to prove that I exist,' says God, 'for proof denies faith, and without faith I am nothing.'
"'But,' says Man, 'the Babel fish is a dead giveaway, isn't it? It could not have evolved by chance. It proves you exist, and so therefore, by your own arguments, you don't. QED.'
"'Oh dear,' says God, 'I hadn't thought of that,' and promptly vanishes in a puff of logic.
"'Oh, that was easy,' says Man, and for an encore goes on to prove that black is white and gets himself killed on the next pedestrian crossing.
Posted by: Yu-Ain Gonnano at July 26, 2012 08:26 AM
Knowing and understanding God is one thing, having evidence via scientific proof is another, Sir.
It's a Naturalist view, not that that's bad. Trying to understand Sky Boss isn't a bad thing either. But, resorting to my Jesuit upbringing, you're trying to fathom the unfathomable. He can change the Rules whenever he wants. The rules aren't as important as the why he made them the rules and left/leaves them that way.
That's how I see it, leastways. The more the merrier, I guess. However one get's to Damascus isn't as important as getting there....
Posted by: ry at July 28, 2012 01:27 PM
Knowing and understanding God is one thing, having evidence via scientific proof is another, Sir.
You might as well seek a legal proof of God as a scientific one. Even if you achieved one -- easy enough in Iran -- it wouldn't be demonstrative of the thing you were really seeking to demonstrate, because it is too low a form of inquiry. The law is subordinate to political philosophy, which is the realm theorizing about voluntary actions: we have the laws we have because we adhere to a particular philosophy about what it means to live a good life and have a just society.
That, in turn, is overlain by a higher order of meaning -- what it means to exist, what good and evil are, and why the world we encountered is ordered as it is. All you can learn from science is how it is ordered, not why -- except in the limited sense of "why" that is equivalent to "how." Learning "how" is not irrelevant to the higher question of "why," though. It provides much of the material for the inquiry.
Posted by: Grim at July 29, 2012 10:30 PM