Could it be that they drink more because they are unsuccessfull instead of the inverse?
because the correlation just means 3 things:
1) they are unrelated
2) more drinking => bad scientist
3) bad scientist => more drinking
4) Unidentified common cause => bad scientist AND more drinking
In the UK beer drinking would be correlated with class and gender, for example, either of which might be correlated with scientific success (wine drinking would have the reverse correlations). I have no idea whether that would also be true in the Czech Republic.
My guess is (1), though. I can't find the paper the RA describes, but it's a common statistical blunder to look for correlations in existing data and assume that they mean there is a meaning to the correlation. There are so many possible correlations between two arbitrary factors in the world that statistics says that some -- many -- will turn out to be correlated by sheer chance (http://en.wikipedia.org/wiki/Data_dredging). Unless the paper is describing a properly controlled trial, the most that this correlation can be is an interesting area for future research.
Python - go learn why strong-typing doesn't mean a lot in practice (except for headaches). In what way? After all, Python is strongly typed. Did you mean "go learn why early binding doesn't mean a lot in practice (except for headaches)"?
Lisp / Haskell / whatever - go learn ' a different way' of programming things. Trust me, it rocks. Well worth a try, but hard to find anything that teaches you how to scale them into real applications. People do it, but I'm blowed if I can find out how they learned.
Ruby / ECMAScript - it's the future baby. Oh, I do hope not! I just can't find a redeeming feature of Ruby. Except the humour: I still haven't stopped laughing at that bit in the Pragmatic Programmer's Guide about "Ruby, unlike less flexible languages, lets you alter the value of a constant".
I assume the other half:
- Do it but don't admit it
- Or don't it but are way less productive than their peers
I don't know how it is for the rest of the slashdot crowd but almost everywhere I've worked it's impossible to be (decently) productive using only authorized software. Quite. I remember being employed to do software development when there were no programming languages included in the approved software, because the people who drew up the approved software list had never bothered to ask the business areas what they did with their computers. I never did get any languages approved, but I did get them to lift my authorisation level so I could run executables that weren't on their heavily locked-down desktop, which was all it took. The company bought the C++ compiler I asked for, and I installed and used it -- unauthorised.
Hmm, lose my child or risk them seeing a targeted ad...lose my child or risk them seeing a targeted ad...sure is a tough call. After all, there's no risk of the mobile phone provider delivering targeted ads based on which cell the phone is in, is there?
Infant mortality rates are a little better now than they were 50000 years ago -- or even 100 years ago. For the survival of humanity, no, we don't need this. For the survival of specific kids it could be a big help.
Imagine the damage that can be inflicted if a stalker could get that feed. Just because I'm paranoid doesn't mean they aren't out to get me.:-) Despite media hysteria, I think the chance of a child getting innocently lost and needing help is significantly higher than a stalker hacking into the child's tracking device. It's risk/benefit, and sure, it has risks, but the benefits vastly outweigh them. After all, all the stalker would know is, say, "the child is in Jubilee park", not "the child is unattended in Jubilee park".
This sounds like the police proposing completely outlandish things, which the citizenry immediately shouts down, but it desensitizes them to things like tracking their children with GPS units, which they voluntarily buy, without the government even telling them they have to. I'm a fierce advocate of civil liberties, but I would have bought such a system had it been available when my kids were younger. Not to spy on them, but what parent is not worried when their kids first start walking to school, first start travelling on their own on buses and trains, and so on? More: my son has learning difficulties, and is having to learn to cope with independent travel. If he goes wrong (if there are diversions to his usual route, for instance), he can phone us up but could have trouble reading the station signs where he has ended up. With technology like this we'd know where he was and be able to tell him what to do. Technology isn't always evil, you know!
For some, it's the slippery-slope:
First they collected DNA from sex criminals.
Then they collected DNA from felons.
Then they collected DNA from all criminals.
Then they collected DNA from people who get speeding tickets.
Then they collected DNA from people who drive.
Then they collected DNA from everyone else. We're between steps 3 and 4 on your list here in the UK, which is nearly the end of the list. The police collect and retain DNA if you are arrested, even if not subsequently charged. Remember that guy recently who was arrested for being in posession of a mobile phone in a public place? Because the police say that somebody thought it was a gun? If they can arrest for that and get DNA, they can already get any adult's DNA they want to. The new thing here is that they want kids DNA too.
It's already anonymous if you want it to be. You can buy an Oyster card over the counter for cash without giving any personal details. You can optionally register the card, so you can top up the pre-pay online and so on, in which case it ceases to be anonymous, but the default is anonymous.
Of course, if you really have something to hide, you buy individual tickets, which would only be traceable with a lot of work correlating the CCTV images (no change from the present). Ok, it's £4 per Zone 1 journey instead of £1.50, but I bet the terrorists can afford it. In other word, this isn't a measure against the terrorists -- it's too easily circumvented: it's just more monitoring of the ordinary reasonably law-abiding citizen.
One of the two patents wasn't granted until Tuesday (although the application was made in 1999). Presumably they didn't think "hmmm, we have a patent on that" because they didn't. The other patent is a bit older, but not that much. Still, reading the RA makes it so much harder to think up sarcastic comments, which is probably why it's so unpopular here.
At its worst, yes, but it would take a double-blind trial to find out whether it's actually saying anything. I've done well-regarded psychometric testing, supposedly scientifically validated, where I could say just the same about the results.
Do the personality claims pre-date western schooling? I don't know. Do they pre-date the possible correlation with the weather, and corresponding environmental effects on the baby and infant that have been suggested? Almost certainly not, but if they're causal factors we should see geographic variation, and ideas that emerged from Europe should be scrambled in the tropics and reversed in the southern temperate zone. Have any studies been done into that? Buggered if I know.
If astrology were to show a reasonable success rate (I think it might on personality profiles, I don't think it does on forecasts) then it would be justifiable to accept other results from astrology in the same area, pending better studies. Life is like that; we don't always have complete information, and sometimes we have to go on the best information we have at the time, accepting that it might later turn out to be dud.
If you want to know about how schoolchildren react to being in school, you should be listening to people who are making scientific studies of schoolchildren's actual behaviour, not to astrologers' descriptions of the personalities they claim are associated with star signs. And how do the scientists decide what to investigate? The real way that science works is that scientists hear some anecdotal evidence, they think "Hmm, that's interesting, I wonder if it's right?" (this step is most likely to happen if it's a funding body that they hear it from) and they then investigate the anecdotal evidence and determine whether it's supported or not (ie, whether or not there's a statistically significant correlation in tests independent of the original anecdotal evidence). If it turns out to be supported they go on to look at possible causes. My suggestion is that astrology is as good a source of anecdotal evidence as any other (astrologers tend to be shrewd observers of human nature -- I might suggest that they have to be in order to sucker their marks); it falls down on the tests of significance being independent of the original anecdotal evidence and on the identification of credible causal factors.
It's not the age of astrology that matters, it's the age of those particular personality "predictions" -- but I don't know the answer either way.
By the way, has anybody mentioned Lalla Ward in this thread? I remember her being interviewed in her single days, and she came over as a complete new-age astrology-head. It doesn't seem to have put Richard Dawkins off -- but I suppose if anybody was going to re-educate her, he had a better chance than most. Or maybe she'd already changed her views.
Also, I find it incredibly sexy when women look up an astrological chart of me or something, of course it's having their attention so focused on me that I find sexy, I take the astrology stuff as seriously as an AD&D game. Well, yes. It did occur to me that the correct answer to whether scientists should date people who believe in astrology" might be "as long as they use contraception, so the astrology gene doesn't persist".
Slashdot Mirror
because the correlation just means 3 things:
1) they are unrelated
2) more drinking => bad scientist
3) bad scientist => more drinking
4) Unidentified common cause => bad scientist AND more drinking
In the UK beer drinking would be correlated with class and gender, for example, either of which might be correlated with scientific success (wine drinking would have the reverse correlations). I have no idea whether that would also be true in the Czech Republic.
My guess is (1), though. I can't find the paper the RA describes, but it's a common statistical blunder to look for correlations in existing data and assume that they mean there is a meaning to the correlation. There are so many possible correlations between two arbitrary factors in the world that statistics says that some -- many -- will turn out to be correlated by sheer chance (http://en.wikipedia.org/wiki/Data_dredging). Unless the paper is describing a properly controlled trial, the most that this correlation can be is an interesting area for future research.
In what way? After all, Python is strongly typed. Did you mean "go learn why early binding doesn't mean a lot in practice (except for headaches)"? Lisp / Haskell / whatever - go learn ' a different way' of programming things. Trust me, it rocks. Well worth a try, but hard to find anything that teaches you how to scale them into real applications. People do it, but I'm blowed if I can find out how they learned. Ruby / ECMAScript - it's the future baby.
Oh, I do hope not! I just can't find a redeeming feature of Ruby. Except the humour: I still haven't stopped laughing at that bit in the Pragmatic Programmer's Guide about "Ruby, unlike less flexible languages, lets you alter the value of a constant".
- Do it but don't admit it
- Or don't it but are way less productive than their peers
I don't know how it is for the rest of the slashdot crowd but almost everywhere I've worked it's impossible to be (decently) productive using only authorized software. Quite. I remember being employed to do software development when there were no programming languages included in the approved software, because the people who drew up the approved software list had never bothered to ask the business areas what they did with their computers. I never did get any languages approved, but I did get them to lift my authorisation level so I could run executables that weren't on their heavily locked-down desktop, which was all it took. The company bought the C++ compiler I asked for, and I installed and used it -- unauthorised.
Hmm, lose my child or risk them seeing a targeted ad...lose my child or risk them seeing a targeted ad...sure is a tough call. After all, there's no risk of the mobile phone provider delivering targeted ads based on which cell the phone is in, is there?
Infant mortality rates are a little better now than they were 50000 years ago -- or even 100 years ago. For the survival of humanity, no, we don't need this. For the survival of specific kids it could be a big help.
Er -- isn't that what I said?
Most of that 60% were smart enough to realise they were screwed however they voted. But I expect some of that 60% were too stupid to vote.
First they collected DNA from sex criminals.
Then they collected DNA from felons.
Then they collected DNA from all criminals.
Then they collected DNA from people who get speeding tickets.
Then they collected DNA from people who drive.
Then they collected DNA from everyone else.
We're between steps 3 and 4 on your list here in the UK, which is nearly the end of the list. The police collect and retain DNA if you are arrested, even if not subsequently charged. Remember that guy recently who was arrested for being in posession of a mobile phone in a public place? Because the police say that somebody thought it was a gun? If they can arrest for that and get DNA, they can already get any adult's DNA they want to. The new thing here is that they want kids DNA too.
So far, it's only "the police want". I want a hot date with Keira Knightley. What we want ain't always what we get.
It's already anonymous if you want it to be. You can buy an Oyster card over the counter for cash without giving any personal details. You can optionally register the card, so you can top up the pre-pay online and so on, in which case it ceases to be anonymous, but the default is anonymous.
Of course, if you really have something to hide, you buy individual tickets, which would only be traceable with a lot of work correlating the CCTV images (no change from the present). Ok, it's £4 per Zone 1 journey instead of £1.50, but I bet the terrorists can afford it. In other word, this isn't a measure against the terrorists -- it's too easily circumvented: it's just more monitoring of the ordinary reasonably law-abiding citizen.
No, it makes you conservative. The capital C makes a lot of difference.
Even INTERCAL would be more readable than normal legislation.
Well, the patent was granted on Tuesday and they filed on Wednesday -- something tells me that they had the papers already prepared.
Prior to 1999?
One of the two patents wasn't granted until Tuesday (although the application was made in 1999). Presumably they didn't think "hmmm, we have a patent on that" because they didn't. The other patent is a bit older, but not that much. Still, reading the RA makes it so much harder to think up sarcastic comments, which is probably why it's so unpopular here.
Because the security guards are really hot?
At its worst, yes, but it would take a double-blind trial to find out whether it's actually saying anything. I've done well-regarded psychometric testing, supposedly scientifically validated, where I could say just the same about the results.
Do the personality claims pre-date western schooling? I don't know. Do they pre-date the possible correlation with the weather, and corresponding environmental effects on the baby and infant that have been suggested? Almost certainly not, but if they're causal factors we should see geographic variation, and ideas that emerged from Europe should be scrambled in the tropics and reversed in the southern temperate zone. Have any studies been done into that? Buggered if I know.
If astrology were to show a reasonable success rate (I think it might on personality profiles, I don't think it does on forecasts) then it would be justifiable to accept other results from astrology in the same area, pending better studies. Life is like that; we don't always have complete information, and sometimes we have to go on the best information we have at the time, accepting that it might later turn out to be dud.
It's not the age of astrology that matters, it's the age of those particular personality "predictions" -- but I don't know the answer either way. By the way, has anybody mentioned Lalla Ward in this thread? I remember her being interviewed in her single days, and she came over as a complete new-age astrology-head. It doesn't seem to have put Richard Dawkins off -- but I suppose if anybody was going to re-educate her, he had a better chance than most. Or maybe she'd already changed her views.