Slashdot Mirror

Yes, they did. From what I understand, they had two random samples of children: One group that was given a Playstation, and another that didn't.

Sit down my son.

What do we mean when we say random? We mean there is an equal chance any given child would receive a Playstation or not. Any given child. The best way to ensure this is to use a good quality random number generator. Leaving aside this--very interesting--topic, I'm going to assume the team had access to such a generator. Now, supposing the team had this generator, it might stand to reason that the two children would indeed have an equal chance of getting a playstation. That is, if you thought about this for less than twenty seconds.

Lets take two children. Little Johnny who has a rather sheltered and strict upbringing with an emphasis on self betterment and education. And little Jimmy, who lives with two working parents who barely have the time to kiss him goodnight in the evening, let alone worry about his upbringing. Now, our ever exacting scientists have an equal chance of offering either little Johnny or little Jimmy a playstation. But, do you really think that little Johnny and Jimmy have an equal chance of receiving a playstation? I think not.

Of course, you might counter that the whole study was done scientifically by professional, blah, blah, etc, etc. But let me tell you something. Most psychologists have about as much scientific rigour as a fifth century alchemist, as much political motivation and funding as a republican congressional candidate, and about as much understanding of statistics as a primary school student. I'm sure their r=0.3 shows a statistically significant correlation in their eyes, but you and I know a little better.

Just because you've put on a lab coat, run a random number generator, and crunched some statistical calculation, that does not make you a scientists or indeed make your work good science.

I'll tell you what else jumps out. The near immaculate periodic nature of the graphs. It's really regular. Something tell me this study has a lot more to do with the school cycle than it does with the season babies are born in. So I'm going to go ahead and guess that a lot of these births are nine months after proms or spring breaks or whatnot.

And how may I ask does the month your mother gave birth to you lead to a lifelong plight? If ever their was a classic junk study showing the usual correlation-causation woolly thinking, this is it. Apparently, a lot of unmarried, less educated mothers have more unprotected sex in May (or less in January). Why would this lead you to conclude that being born in winter disadvantages someone. I was born in winter and my mother was married, educated and employed. Has my life been deprived somehow? Do I need extra money or protection or something? Yeah sure, chuck me some money. I'll consider it an idiot tax; like the lottery.

Correlation is NOT Causation. Correlation proves nothing. Saturn is correlated to the S&P 500 with r=0.88. And don't think there a correlation so profoundly stupid that someone won't publish a "scientific" paper on it.

This research is junk. Correlation studies need to die.

Take this particular news story: There are no specific claims of any sort that I can see in any of the article links on either side. There aren't any specific correlations being asserted or presented between anything and anything else that I can tell, just a bunch of bitching on both sides.

What do you think the arguments of the UK government are based on? What evidence to you think they will present for their claims? They will trot out correlation studies like the dead horse they are, to be beaten in front of the media once again. This is always the case when they need any flimsy excuse to demonise video games.

The "correlationisnotcausation" whine-fest is completely beside the point, like a mass hallucination.

"Correlationisnotcausation" is the point. These ads would never have been aired without people accepting the validity of dubious and misleading correlation studies. Correlation is not and never will be causation, and in case you have doubts about this, I'll trot this link out again about Saturn's correlation with the S&P 500.

Correlation studies and other misapplications of statistics are behind this claim, and the injustice it represents. It's right to point that out.

There is a direct correlation between piracy and lost sales, I've seen it. Grow up.

Correllation does not imply Causation.

It sucks, but ya, people steal music on the internet, sales drop for the karaoke labels, we get less karaoke.

Having seen some of the karaoke subs produced by anime fansubbers, I'm willing to bet that fan made karaoke videos will produce higher quality content than any professional label. In the face of ubiquitous video editing software, your industry has simply succumbed to its own irrelevance.

That in 80% of cases some remnant, some energy of that person was left behind? Just because it happens frequently doesn't mean it is *not* supernatural in nature.

Absolutely. Why, in 46% of documented cases, energy inductance drain has been detected in the vicinity of dead bodies, decaying exponentially with time and oscillating about a void karma mean. And in 67.2% of such cases, inductance eddies were suggested by gathered data as having occurred before the obituarial event. Couple this evidence with well known ESP studies and psionic-harmonics studies, and the case for ghosts, and particularly poltergeists becomes more than compelling, it's practically irrefutable.

You can link these findings with the proven influence of Saturn, on general supernatural phenomena, especially those involving the recently deceased. True, r is only 0.13 in the case of 80% energy remnants(measured on a Kasparov scale), but the results ARE statistically significant.

It doesn't guarantee that the other option won't be found to be possible at some point.

Or has been found already, and is just being ignored my small minded skeptics. The truth is out there. Keep the faith!

So nearly all species in the animal kingdom have inherent behavioral differences between males and females - except humans? You really believe that?

What do these behavioral differences have to do with professional work in the field of computer science, or any science for that matter? What is it about behavior in male and female monkeys that effects the differences in their computer programs?

Some people, talk a lot of shit. Shit about "female brains" and "male brains" and "spatial reasoning" and "behavioral differences" and some other tripe. But I have yet to see any solid proof, outside of neuroscience and psychology junk studies, that any of the shit they talk has anything to do with people's ability to do a job, or have an interest in a field. If all you can offer me is statistical hocus pocus, then you can't offer me anything at all.

Go back in time 50 years and make a replacement in most of the arguments you see in this thread. Specifically, change "men" and "women" to "whites" and "blacks". My intention is not to make a rhetorical point, but to show you just how invalid most of the arguments for the lack of women in science really are. They keep the same company as old ideas and arguments about racial superiority/inferiority, so they cannot honestly be regarded as objective sources of truth.

Changing workforce trends over the last 50 years in western society have taught us one thing. The huge majority of reasons for the past lack of women in highly skilled labour was due to cultural and social factors. Therefore, if you ask me why there is a lack of women in computer science at this point in time, I can with some confidence state that it is because of social and cultural reasons.

Of course, I could be wrong. There could be some testosterone/estrogen activated/deactivated brain region or gland that promotes/inhibits emotional/spatial thinking when someone is exposed to dolls/trucks/fixed width fonts. Maybe. We could all pull all manner of such tripe out of our asses all day, and are indeed doing so in this thread. But you know what? I'll bet not one single experiment, a real experiment (blah, blah, wiki-trip, blah), will ever confirm a single word of it.

It's cultural. So stop talking shit, and acknowledge it.

(ie A->B, B->A or C->A and C->B, these are the ONLY explanations for statistically significant correlation). ....

I'd like to see you try to refute this (without resorting to insults or rhetoric), particularly if you can think of a way for variables to be correlated without some form of causal relationship as I've described.

I have decided to respond to your post with the refutation you seek. In short, the Right Ascension of the planet Saturn in the night sky is correllated with the S & P 500 stock market index. The correlation coefficient is 0.88.

True, my example is somewhat facetious, but just how far removed is it from the junk studies posted on Slashdot nearly every day? Why is this (apparently quite strong) correlation to be rejected, yet even weaker one accepted in other fields. What is the essential difference here? Do we only accept correlations between variables that our "common sense" tells us should be related? Is that science?

No. You cannot infer or imply causation from correlation. It is a false premise from the outset. Only an experiment, a real experiment can establish causation. Anything else is second or third rate material which cannot and should not be accepted by anyone without further proof.

Correlation is not causation. It does not imply causation. It is a statistical technique that has been consistently abused and misapplied by charlatans for decades and proves exactly nothing. In the hands of many, if not most, who use it correlation is isomorphic to numerology, and should be soundly rejected by any serious scientist.

Therefore, our mantras are just fine, and I see no reason to alter them.