The first amendment only restricts what the US federal government and all of its subsidiary governments can do, and nothing else. Brush up on your country's laws.
Private corporations can do whatever they want in terms of providing or not services to the general public. That is why making disability a protected status resulted in decades of business changes. Political belief is not a protected status.
These kids have spent too much time in their bedrooms fake shooting people to understand the moral and ethical obligation to destroy Nazism and all of its sympathizers.
You mean calling out biased invested parties? One can not assume any value inherent in a fan post without disclosures. They are most likely in it for their own financial benefits and at the cost of your own, so they are not only useless but detrimental to everyone.
Musk routinely violates the rights of stockholders and investors by combining his personal ventures to reduce his personal risk while increasing that born by all others.
No. I'm afraid you are an idiot. The problem with CO2 and other greenhouse gases is that they absorb infrared energy, store it in the atmosphere, and prevent that energy from being reflected back into space as it would otherwise. This increases the energy in that large dynamical system we call the Earth, with predicable results that mean massive die offs in humans and all the animals and plants we depend on for survival.
CFCs do destroy the ozone. Their elevated release due to manufacture and use in consumer products was responsible for the predicable chemistry destroying ozone (O3, particularly in the stratosphere). Since 1987 CFC reduction has been responsible for the predictable chemistry that resulted in the reduction of the very same ozone hole. Understandably this is history for you, but not for those who lived through it and have worked in related industries.
Sure.... How do you find it, who pays for channel distribution (broadcast or hosting for electronic media), who pays for storage and display (physical media)? How does the artist eat, pay rent, afford school fees for their kids, etc.? Money is required to function in the economy, and artists are not exempted from that requirement.
Google doesn't profit exclusively from it, and doesn't curate those links based on their internal activity. They do face anti-competitive activity charges though.
In short all this data means and all it can mean: with a convenience sample collected before and after phases of a US election, the proportion of politically related phrases (see attached list 1) with racist or abusive phrasing (see attached list 2) increased within this data, on the basis of our weighting (see procedures section), for the censored (lost input values, adjustments ignored) data set of google searches, between groups so divided (with no effort to model autocorrelation in any variables). This has no definite relationship to a geographic area but only an IP range, or to people using that range. This frequency has no definite relationship to people (includes automated entries), and has no definite relationship to Americans or any other national groups (no stratified or any other sampling designs were used). Data collection was limited to freely available sources with no attempt to represent any smaller markets or bracketed to eligible voters by age, convict status, etc. And includes unidentified input from individuals with diagnosed psychological and medical conditions.
The problem is you don't understand what sampling is, what it is sued for. Doing your own research and independent study is the only way to fix the defects in your knowledge. Decent conversation, you do have some skill but it needs to be harnessed with reality and the significant body of statistical knowledge we have already built up. Don't get side tracked with machine learning which just duplicates statistical theory, spend your time learning the real stuff. The book is so old you can find a PDF for free online easily, although it is that older version that is not as well organized as the second edition. Try it.
The TLDR equivalent response is that the applications to which this data is being applied now need to have been considered from the beginning. The precise population identified must then by sampled in a way that the resulting sample is actually representative, so that the probability models can be applied validly. For a study with any behavioral modeling intent the data collection should have been form more than just a google search, and requires a well developed survey design and unbiased questionnaire, as well as follow up contacts for non-responses. As stated originally this study is a bad convenience sample, composed of only those terms submitted to a google search, so it can't represent any behavior.
Well for a supposed professional you ignored all aspects of the design of experiments. Instead you focused on what is taught in primary school, ignoring the actual work required for an experimenter in any real setting. We are expected to exclude those things that waste money, based on the existing knowledge base. You know, the parts that funding sources are most interested in due to the fact that it determines thee cost of research and experiments? Look I know this is a pissing match for you but I don't give a fuck. You're just a god damn idiot like the rest, trying to wave a fake flag around. Advice is just this: fix your defects and don't be such an idiot.
That just means you are an illiterate retard. Check for the dismissal of Nazi crimes as described. That shows they are sympathizers who idolize the fools.
Because there is no "full" data set. The available data is simply a very badly collected sample because it fails to capture any of the actual generating process. This produces inherent errors including bias (persistent error, since you are confused). The examples were not specifics but the simple product of a few minutes of thought from a trained statistician. There are an unlimited number, because the frame for any of the analysis done wasn't defined before the data was collected and didn't inform the process in any way. You should read that book by Deming.
No. Sampling theory is also the foundation of inferential statistics. See W.E. Deming "Some Theory of Sampling" (1950). Error in statistical estimation is fundamentally either bias or stochastic. A sampling defect introduces a systemic data collection flaw which means all statistics calculated from it do not estimate the population parameters they are being used to imitate.
In this case that means the data collected from google searches isn't indicative in any way of what Americans as a whole (or any other group for that matter) think, believe, or have influencing their behavior. Data not entered into google searches includes why terms are entered, the terms thought but not entered, the terms modified by google's previews of popular search terms, the indication that a user entered them rather than a machine, etc.
One final point before you go home to study some real statistics. Assumptions without validation make everything useless.
Can you figure out what (100% - 60%=) 40% of 1.3 Billion is? Hint: > 500 Million. That means you are stating even more of a difference exists than the figure which caused the original shallow outrage in this thread. This is basic numeracy.
While this is only partially damning to the US compared to the thousands of other things it is failing at, this method of data collection ignores the need for sampling. Even taking a census of collected data is nothing but a biased sample due to the sheer quantity of data that is never entered into a google search. At best the changes in frequencies may show the behavior of whatever subset of the area targeted participates, but it remains a convenience sample with limited use in larger inference.
Thanks for the anecdote! Unfortunately even removing AC wouldn't provide any real benefit to uncovering or determining the possibility of or extent of any shill/propagandist style activity. Puppet accounts are easy to create. How about some kind of country of origin geoip location with filtering for tor exits/vpns? AC can be retained, but I expect all sorts of remote compromises and bouncers to make this idea worthless, too.
The only solution is to create innumerable named accounts, use each randomly and throw each away after a random time. Better than posting AC, this destroys the entire idea of a coherent group with implicit trust which is the mechanism abused for these kinds of manipulations. That applies not just to slashdot but to every forum. The Internet is dysfunctional, and is being used to mount threats to physical security. Deprive the paid fools of an audience by making everything as fake as they are. Deprive Russia and its pawns of all power and observe peace through strength greater than anything they can piss on.
From those with actual military training and/or combat experience. Killing Russian invaders is what I spent the better part of 20 years for in the Bundeswehr. Partisans and nationalists in Ukraine are less skilled but more up to date with recency of combat experience. Goodnight you weak pitiful runt, you are lucky you are not just be a stain in your fathers jockstrap.
God damn you are an idiot, learn to read paragraphs. Russia under Putin is evil and everywhere it is fought deserves reinforcement. Russia must be reformed and Putin must be killed.
Slashdot Mirror
The first amendment only restricts what the US federal government and all of its subsidiary governments can do, and nothing else. Brush up on your country's laws.
Private corporations can do whatever they want in terms of providing or not services to the general public. That is why making disability a protected status resulted in decades of business changes. Political belief is not a protected status.
It is not protected, and your thinking is wrong from first premises.
These kids have spent too much time in their bedrooms fake shooting people to understand the moral and ethical obligation to destroy Nazism and all of its sympathizers.
You mean calling out biased invested parties? One can not assume any value inherent in a fan post without disclosures. They are most likely in it for their own financial benefits and at the cost of your own, so they are not only useless but detrimental to everyone.
Musk routinely violates the rights of stockholders and investors by combining his personal ventures to reduce his personal risk while increasing that born by all others.
No. I'm afraid you are an idiot. The problem with CO2 and other greenhouse gases is that they absorb infrared energy, store it in the atmosphere, and prevent that energy from being reflected back into space as it would otherwise. This increases the energy in that large dynamical system we call the Earth, with predicable results that mean massive die offs in humans and all the animals and plants we depend on for survival.
CFCs do destroy the ozone. Their elevated release due to manufacture and use in consumer products was responsible for the predicable chemistry destroying ozone (O3, particularly in the stratosphere). Since 1987 CFC reduction has been responsible for the predictable chemistry that resulted in the reduction of the very same ozone hole. Understandably this is history for you, but not for those who lived through it and have worked in related industries.
Sure.... How do you find it, who pays for channel distribution (broadcast or hosting for electronic media), who pays for storage and display (physical media)? How does the artist eat, pay rent, afford school fees for their kids, etc.? Money is required to function in the economy, and artists are not exempted from that requirement.
Google doesn't profit exclusively from it, and doesn't curate those links based on their internal activity. They do face anti-competitive activity charges though.
In short all this data means and all it can mean: with a convenience sample collected before and after phases of a US election, the proportion of politically related phrases (see attached list 1) with racist or abusive phrasing (see attached list 2) increased within this data, on the basis of our weighting (see procedures section), for the censored (lost input values, adjustments ignored) data set of google searches, between groups so divided (with no effort to model autocorrelation in any variables). This has no definite relationship to a geographic area but only an IP range, or to people using that range. This frequency has no definite relationship to people (includes automated entries), and has no definite relationship to Americans or any other national groups (no stratified or any other sampling designs were used). Data collection was limited to freely available sources with no attempt to represent any smaller markets or bracketed to eligible voters by age, convict status, etc. And includes unidentified input from individuals with diagnosed psychological and medical conditions.
The problem is you don't understand what sampling is, what it is sued for. Doing your own research and independent study is the only way to fix the defects in your knowledge. Decent conversation, you do have some skill but it needs to be harnessed with reality and the significant body of statistical knowledge we have already built up. Don't get side tracked with machine learning which just duplicates statistical theory, spend your time learning the real stuff. The book is so old you can find a PDF for free online easily, although it is that older version that is not as well organized as the second edition. Try it.
The TLDR equivalent response is that the applications to which this data is being applied now need to have been considered from the beginning. The precise population identified must then by sampled in a way that the resulting sample is actually representative, so that the probability models can be applied validly. For a study with any behavioral modeling intent the data collection should have been form more than just a google search, and requires a well developed survey design and unbiased questionnaire, as well as follow up contacts for non-responses. As stated originally this study is a bad convenience sample, composed of only those terms submitted to a google search, so it can't represent any behavior.
Well for a supposed professional you ignored all aspects of the design of experiments. Instead you focused on what is taught in primary school, ignoring the actual work required for an experimenter in any real setting. We are expected to exclude those things that waste money, based on the existing knowledge base. You know, the parts that funding sources are most interested in due to the fact that it determines thee cost of research and experiments? Look I know this is a pissing match for you but I don't give a fuck. You're just a god damn idiot like the rest, trying to wave a fake flag around. Advice is just this: fix your defects and don't be such an idiot.
That just means you are an illiterate retard. Check for the dismissal of Nazi crimes as described. That shows they are sympathizers who idolize the fools.
Because there is no "full" data set. The available data is simply a very badly collected sample because it fails to capture any of the actual generating process. This produces inherent errors including bias (persistent error, since you are confused). The examples were not specifics but the simple product of a few minutes of thought from a trained statistician. There are an unlimited number, because the frame for any of the analysis done wasn't defined before the data was collected and didn't inform the process in any way. You should read that book by Deming.
No. Sampling theory is also the foundation of inferential statistics. See W.E. Deming "Some Theory of Sampling" (1950). Error in statistical estimation is fundamentally either bias or stochastic. A sampling defect introduces a systemic data collection flaw which means all statistics calculated from it do not estimate the population parameters they are being used to imitate.
In this case that means the data collected from google searches isn't indicative in any way of what Americans as a whole (or any other group for that matter) think, believe, or have influencing their behavior. Data not entered into google searches includes why terms are entered, the terms thought but not entered, the terms modified by google's previews of popular search terms, the indication that a user entered them rather than a machine, etc.
One final point before you go home to study some real statistics. Assumptions without validation make everything useless.
Can you figure out what (100% - 60%=) 40% of 1.3 Billion is? Hint: > 500 Million. That means you are stating even more of a difference exists than the figure which caused the original shallow outrage in this thread. This is basic numeracy.
You have the understanding of a child just learning by theory. Useless for application, and you don't even understand why.
Read! It corrects your current stupidity.
Read more you fucking idiot. Or are you really just another illiterate illegitimate bastard?
While this is only partially damning to the US compared to the thousands of other things it is failing at, this method of data collection ignores the need for sampling. Even taking a census of collected data is nothing but a biased sample due to the sheer quantity of data that is never entered into a google search. At best the changes in frequencies may show the behavior of whatever subset of the area targeted participates, but it remains a convenience sample with limited use in larger inference.
Thanks for the anecdote! Unfortunately even removing AC wouldn't provide any real benefit to uncovering or determining the possibility of or extent of any shill/propagandist style activity. Puppet accounts are easy to create. How about some kind of country of origin geoip location with filtering for tor exits/vpns? AC can be retained, but I expect all sorts of remote compromises and bouncers to make this idea worthless, too.
The only solution is to create innumerable named accounts, use each randomly and throw each away after a random time. Better than posting AC, this destroys the entire idea of a coherent group with implicit trust which is the mechanism abused for these kinds of manipulations. That applies not just to slashdot but to every forum. The Internet is dysfunctional, and is being used to mount threats to physical security. Deprive the paid fools of an audience by making everything as fake as they are. Deprive Russia and its pawns of all power and observe peace through strength greater than anything they can piss on.
You are a sad stupid inept and powerless little boy who can not accomplish anything. Go away and hide in your stained pillow fort or something.
From those with actual military training and/or combat experience. Killing Russian invaders is what I spent the better part of 20 years for in the Bundeswehr. Partisans and nationalists in Ukraine are less skilled but more up to date with recency of combat experience. Goodnight you weak pitiful runt, you are lucky you are not just be a stain in your fathers jockstrap.
God damn you are an idiot, learn to read paragraphs. Russia under Putin is evil and everywhere it is fought deserves reinforcement. Russia must be reformed and Putin must be killed.