How 136 People Became 7 Million Illegal File-Sharers

Barence writes "The British government's official figures on the level of illegal file sharing in the UK come from questionable research commissioned by the music industry. The Radio 4 show named More or Less examined the government's claim that 7m people in Britain are engaged in illegal file sharing. The 7m figure actually came from a report written about music industry losses for Forrester subsidiary Jupiter Research. The report was privately commissioned by none other than the UK's music trade body, the BPI. The 7m figure had been rounded up from an actual figure of 6.7m, gleaned from a 2008 survey of 1,176 net-connected households, 11.6% of which admitted to having used file-sharing software — in other words, only 136 people. That 11.6% was adjusted upwards to 16.3% 'to reflect the assumption that fewer people admit to file sharing than actually do it.' The 6.7m figure was then calculated based on an estimated number of internet users that disagreed with the government's own estimate. The wholly unsubstantiated 7m figure was then released as an official statistic."

Comments

Story meaning?

[sopssa]

I actually had several feelings about this summery, because:

1) Usually pro-filesharers try to make it sound like filesharing is usual activity and try go for most or 70-90% user share
2) The summary tries to paint this study bad because it "downsides" the amount of filesharers
3) The rant about examining only 1,176 people for the study - in which case the same kind of tv viewer statistics and other studies are made in what case.

So could someone please explain *why* is it a questionable research. It is like every other study where you study small amount of people and make estimates based on it to reflect whole population. Usually this amount of people also gives somewhat correct results on the whole population. Theres some error margin, but its close enough.

So what is the point of this story? That statistics researches use only minor subset or people to do their research instead of asking from everyone? They always have.

Re:Story meaning?

[Lemmy Caution]

Because statistics are hard and outrage is easy.

Re:Story meaning?

[bertoelcon]

I think this could be summarized under lies. damn lies, and statistics.

Re:Story meaning?

[wizardforce]

So could someone please explain *why* is it a questionable research.

1. the same size is small.. probably too small to make the claims they did. 2. they altered the numbers on an estimate of how many people fileshare on the assumption that the number was under-reported. 3. conflict of interest... it's like the tobacco industry sponsoring studies claiming that smoking doesn't have anything to do with lung cancer... there is significant reason to believe that the study carries significant bias in favor of their conclusion and must at the least be repeated by other sources.

So what is the point of this story? That statistics researches use only minor subset or people to do their research instead of asking from everyone? They always have.

N. real statistics researchers know that this study has numerable crippling flaws and should not be held as gospel by anyone. Even a first year stats student can see it. The reason this story is important is that it may influence governmental policy and it's flawed... That's dangerous.

Re:Story meaning?

[nethenson]

Because it is a study that is based on asking a few people the question "I am doing an study for the BPI, are you downloading files illegaly?" and then deciding that some of the answers (an arbitrary number of them) are lies.

Re:Story meaning?

[Loomismeister]

The point isn't that they surveyed a small group of people and therefore the statistics aren't significant. If you RTFA you would see that they based the 7m number on the false statistic that 40m some people were using the internet that year when there was really only like 33.9m. They also bumped up the percentage of filesharing people based on the assumption that some people lied about whether they had programs like that or not. Really the lesson here is to read the featured articles because the slashdot summaries as a general rule are misleading.

Re:Story meaning?

[caffeinemessiah]

Argh, where to begin?

The summary tries to paint this study bad because it "downsides" the amount of filesharers

I presume by "downsides" you mean "reduces"? Well the summary says "That 11.6% was adjusted upwards to 16.3% 'to reflect the assumption that fewer people admit to file sharing than actually do it.'" So they actually UPPED the number of filesharers. This is objection #1 to "good research":
1. You do a survey to objectively measure the support of your hypothesis
2. The survey of a tiny sample indicates that filesharers are a pretty low percentage
3. You "adjust" this number -- otherwise known as "fudging the data" -- to better reflect your own hypothesis.

The same tactics in any scientific endeavor would get your papers retracted, your funding canceled, some sort of disciplinary action initiated, etc.

The second objection, and this applies to other studies too that try to make grand claims from small samples, is that it's A SMALL SAMPLE. For your survey to be representative, your sample has to be representative. It's also difficult to choose people independently at random, and without that assumption, all your basic statistics fall apart. Perhaps they went through a list of BT subscribers and pulled names at random -- but what if downloaders are overrepresented amongst BT subscribers? What if they only polled home internet users, but then used the "total number of internet users" -- which includes corporate subscribers -- to come up with their 11mil number? There are other possible, non-numerical issues too. What if the respondents confused downloading from bittorent with downloading from iTunes?

If you want many other examples of "bad science", read Ben Goldacre's blog

Re:Story meaning?

[Trepidity]

It doesn't really make sense to claim "sample size is small" for an 1,100-person sample. If the sampling was done in a random, unbiased manner, that size sample gives a margin of error of +/- 3%. If there are flaws in the sampling method, that's another thing, but the sample size alone doesn't seem problematic, unless you need accuracy better than +/- 3%.

Re:Story meaning?

[Mikkeles]

In addition, sharing files is not in and of itself illegal.

Re:Story meaning?

[BikeHelmet]

"I am doing an study for the BPI, are you downloading files illegaly?"

11.6% of which admitted to having used file-sharing software

^
Where is the downloading illegally bit? I torrent linux distros fairly often. I know people that have limewire, and use it to get music from many decades back. Depending on where you are, this could be 100% legal.

Also, the outrage is more over the upwards estimation technique. 4m is significantly less than 7m.

Re:Story meaning?

[wizardforce]

Oh I forgot to note this... anyway it addition to other potential flaws TFA says

11.6% of which admitted to having used file-sharing software

emphasis mine. They admitted to using file sharing software not pirating goods via said software... The study is effectively making the assumption that filesharing = copyright infringement. Also from TFA:

The 6.7m figure was then calculated based on the estimated number of people with internet access in the UK. However, Jupiter research was working on the assumption that there were 40m people online in the UK in 2008, whereas the Government's own Office of National Statistics claimed there were only 33.9m people online during that year.

Even if the study did get the sample size correct the conclusion would still be nearly 30% wrong owing to their false assumption of the number of people with net access. neglecting the distinction between filesharing and copyright infringement TFA estimates that the actual number is between ~30 and ~50% lower than the study claims.

Re:Story meaning?

[Anonymous Coward]

1. the same size is small.. probably too small to make the claims they did.

First statistics lesson I ever had, first thing the professor did was make an estimate based on 10 people about the whole population. He was correct, by the way. He went on to rant that anything that uses large amounts of people (by which he meant more than at most a few dozen) was not proper statistics. If you simply count everybody, it should be called "counting", you see, not statistics.

2. they altered the numbers on an estimate of how many people fileshare on the assumption that the number was under-reported.

And since they are right that the number turned out to be bigger in other studies, slightly. It seems a reasonable adaptation. It's easy to say it's unreasonable, of course. But they are absolutely correct that the number is most likely smaller. So how much should they adjust it ? Like I said, it seems a reasonable adjustment. Not absurdly high, not absurdly low.

3. conflict of interest... it's like the tobacco industry sponsoring studies claiming that smoking doesn't have anything to do with lung cancer... there is significant reason to believe that the study carries significant bias in favor of their conclusion and must at the least be repeated by other sources.

There don't exist studies that have no bias. Either research is funded by companies, or it's funded by government. Both have serious axes to grind, mostly pertaining to political ideology. If business intrest groups would not fund research we'd never have even the semblance of unbiased research that we have.

By the way, who should pay for studies ? Obviously the government has a vested interest in more legislation. The ifpi (us dept) has a vested interest in creating legal instruments to counteract filesharing. And the filesharers have a vested intrest in more "privacy", and legal instruments against ISPs (for the same reason a thief wants privacy, obviously, let's please not start the "what about those who only share openbsd", we all know that's not the filesharers being talked about).

How about we do the sane thing, and let all of them fund studies. Then read them all, and see what we believe to be true.

Just because people are biased, by the way, does not mean the truth can be biased. We are simply limited to imperfect instruments for reading the truth. Truth is absolute, and the number of filesharers is just a single number, not 2, not 5. And yes, we'll probably need a better definition and classification than "filesharers". The effects of filesharing are negative for artists (certainly for pop artists), and especially for the "music industry". There can be little doubt about that. How much damage is done, is anyone's guess. But by criticizing their observations, AND listening to them criticize our observations, we can hope to get closer to the real truth.

Re:Story meaning?

[mdwh2]

Usually pro-filesharers try to make it sound like filesharing is usual activity and try go for most or 70-90% user share

Indeed - but equally the Government has also been trying to over-inflate the extent, from an window-breaking-fallacy angle "It's costing the economy billions of pounds!"

So could someone please explain *why* is it a questionable research.

Because of the way that 11.6% was replaced with 16.3%, and an incorrect value of the number of Internet users was used, as clearly stated in TFS. If you RTFA:

"If the BPI-commissioned Jupiter research had used the Government's online population figures, the total number of file sharers would be 5.6m. If the researchers hadn't adjusted their figures upwards, the total number of file sharers would be only 3.9m - or just over half the figure being bandied about by the Government."

Yes, you are right that saying "only 136" is silly, as 1,176 is more than enough for a valid sample as long as it is random, but there are other concerns that are clearly spelled out in TFA and TFA.

Theres some error margin, but its close enough.

Is an error of 79% "close enough" for you?

Re:Story meaning?

[Demonantis]

From TFA;

"The 6.7m figure was then calculated based on the estimated number of people with internet access in the UK. However, Jupiter research was working on the assumption that there were 40m people online in the UK in 2008, whereas the Government's own Office of National Statistics claimed there were only 33.9m people online during that year."

I think this demonstrates the point better then the rest of the article. They made significant numbers up in their head to inflate their claim. This is not questionable research. It is a bunch of opinions made by the researchers and stated as fact.

I remember hearing a while ago that there are huge penalties for fudging bids on government contracts(US). Maybe they should do the same for reports that are presented to the government.

Re:Story meaning?

[Volante3192]

And since they are right that the number turned out to be bigger in other studies, slightly. It seems a reasonable adaptation. It's easy to say it's unreasonable, of course. But they are absolutely correct that the number is most likely smaller. So how much should they adjust it ? Like I said, it seems a reasonable adjustment. Not absurdly high, not absurdly low.

Here's where I find a major problem. You do not fudge your data. Period. These other studies may show higher numbers, but do we have proof they weren't fudged as well?

There's too many stories about companies performing pharmecutical trials and then throwing the data away because it didn't present a positive light.

If you're going to adjust numbers, you better have a damn sound reasoning for it rather than "we have a hunch people lied, so..."

Re:Story meaning?

[jrumney]

What does an "error of 3%" mean? Does it perhaps mean there is only a 50% chance (assuming normal distribution) that the proportion of filesharers in the total population is somewhere between 8.6% and 14.6%?

Re:Story meaning?

[Trepidity]

Basically, except that the confidence level for the interval is 95%, not 50%. Should've quoted that, but 95% is the usual assumed one.

Re:Story meaning?

[vikstar]

Every politician should undergo a statistics examination as a prerequisite.

Re:Story meaning?

[Narpak]

1. the same size is small.. probably too small to make the claims they did.

Agreed. I would say that for a better and more precise statistic one would have to take samples of say 15.000 people (or more) and to take several samples from varied parts of the UK. 1,176 people from a population of of about 61 million spread across varied geographical, economical, political, social, cultural and religious demographics seems far far far too small to make any sort of estimation one way or another.

Re:Story meaning?

[Mozk]

First statistics lesson I ever had, first thing the professor did was make an estimate based on 10 people about the whole population. He was correct, by the way. He went on to rant that anything that uses large amounts of people (by which he meant more than at most a few dozen) was not proper statistics. If you simply count everybody, it should be called "counting", you see, not statistics.

How is using 1.67^-7 percent of the population (10 out of 60 000 000) better than using 1.96^-5 percent (1176 out of 60 000 000)? Not to set up a strawman here, but are you saying that this estimation would be more accurate if they had used only 10 people?

It'd be absurd to make estimations about a population as large as this using 10 people. 1176 is still small enough to not be "counting", yet would provide much more accurate estimates.

Re:Story meaning?

[Anonymous Coward]

A margin of error of +/- 3% is the Maximum margin of error for a random sample of 1100 drawn from a large enough population at the 95% significance level (actually its really +/-2.95%), i.e this is the margin of error when the observed % is 50% , The margin of error is less when the observed % approaches 0 or 100%.

In the case of an observed % of 11.6 the margin of error is +/-1.9% so it is 95% likely that the population figure is between 9.8% and 13.5%

Re:Story meaning?

[mysidia]

From what the article says, they equate file sharing software users to 'illegal file-sharers'; so someone using BitTorrent to download Linux ISOs would put them in 'file sharer' territory, and therefore count towards the illegal file sharer statistics?

The statistical inferences about the internet using population are only valid if the 1,176 net-connected households are a representative sample of the population they are attempting to draw statistical inferences about.

Meaning the characteristics of the people they randomly chose to be in their group are similar to those of the community they will make inferences about (the internet community).

E.g. if X% of the internet population are catholic people between ages 16-18, and Y% of people in the internet population are ages 22-28 and atheist, X% and Y% of the people in their sample should have those respective characteristics, for the sample to be representative.

If there are differences in the composition of their sample, or if the composition of internet users is not fully understood, then the inferences are fundamentally flawwed.

It's almost impossible to choose a usable sample of internet users as random; asking internet users to pick theirselves results in a biased (self-selecting) sample. There's not a book of all internet users they can use; they can't force all internet users to partake in their study, internet users with certain characteristics may be more likely than others to even refuse to participate.

Here are some of the issues:

That 11.6% of respondents who admitted to file sharing was adjusted upwards to 16.3% "to reflect the assumption that fewer people admit to file sharing than actually do it." ... The report's author told the BBC that the adjustment "wasn't just pulled out of thin air" but based on unspecified evidence

Without specifying the evidence, there is no way to substantiate the validity of this adjustment, and the resulting conclusion is highly questionable.

Jupiter research was working on the assumption that there were 40m people online in the UK in 2008, whereas the Government's own Office of National Statistics claimed there were only 33.9m people online during that year.

In other words, the population something being inferred about was smaller than the report assumed.

Re:Story meaning?

[shutdown -p now]

So basically it's not 7 million, but more like 4?

Lies it may well be, but does it really make any difference with regard to conclusions drawn from that number (which obviously vary widely depending on which side you're on)? The order of magnitude is still the same.

Re:Story meaning?

[CarpetShark]

this study bad because it "downsides" the amount of filesharers

That's "downsizes"

Re:Story meaning?

[mysidia]

In this case, 3% is >1.17 million people.

Re:Story meaning?

[Atario]

it's A SMALL SAMPLE

No, it's not.

http://www.raosoft.com/samplesize.html

About 60 million people in the UK, sample size of 1,176, confidence interval of 96% gives a margin of error of 2.99%. So, it's 96% likely that they got within 2.99% of the right answer (to the question of how many people admit to it).

I hate seeing this "that's too small a sample size" objection to every single study, from people who clearly don't know enough about how sample sizes work.

Re:Story meaning?

[Runaway1956]

1 The 7m figure had actually been rounded up from an actual figure of 6.7m
2 It gets worse. That 11.6% of respondents who admitted to file sharing was adjusted upwards to 16.3% "to reflect the assumption"
3 The 6.7m figure was then calculated based on the estimated number of people with internet access in the UK.

TFA is pretty clearly challenging those figures based on assumptions made, faulty estimates, and rounding up. The original "research" was clearly engineered to give a high number.

Is there anything else I can help you with?

Re:Story meaning?

[commodore64_love]

>>>e rant about examining only 1,176 people for the study - in which case the same kind of tv viewer statistics

Yes but those TV stats are produced by carefully selecting the homes to reflect each city's ethnic makeup (at least that's true with Nielsen in the States). In contrast the 1176 people were an uncontrolled survey of people who *volunteered* to take the poll, and therefore represents... essentially nothing. It's unscientific.

Re:Story meaning?

[Goaway]

Yes, I'm sure there's a large enough fraction of people who sit around torrenting Linux distros to significantly affect the result.

Re:Story meaning?

[Goaway]

So your problem with it is that they claimed 7 million when it was actually more like 4 or 5 million?

Re:Story meaning?

[Anonymous Coward]

The study is effectively making the assumption that filesharing = copyright infringement.

I have a very hard time believing that the vast majority of people that use any filesharing application do so exclusivley for legit and non-copyright infringing purposes.
Given the vast quantity of content, I seriously doubt that very many people go through any sort of hassle to determine what is legit and what is not, which results in virtually everyone obtaining material that is copyrighted, regardless whether they know (or care). Given that, I think its a fair guess on their part that yes, most people that claim they are using file-sharing software do so to obtain material illegally.

I just don't understand the stance that most people on this board seem to take regarding this issue. How can everyone be so supportive of what very obviously amounts to theft? It appears to me that somehow people think it is their "right" to obtain copyrighted material for free. I just don't buy for a second that people who claim to only use file-sharing apps for legitimate purposes only actually do so.

If you do indeed use all file-sharing applications for 100% legit purposes, please educate me what you use these services for that makes them so very essential to cause these very emotional posts here.

Re:Story meaning?

[sopssa]

1176 people compared to 10 comes a lot closer to real truth. Close enough to have just a few % error margin, so in statistics it works good.

Re:Story meaning?

[Runaway1956]

Yes, it makes a difference. When the lobbyists stand in front of lawmakers, those lawmakers want to know the real size of the problem. If the industry's lobbyists have to say, "We think we are losing almost a million pounds each and every year to piracy", lawmakers are going to be mildly concerned. However, if they lie, and claim that they are losing BILLIONS of pounds, those lawmakers realize that the tax collectors are losing a huge sum of money.

When you want action, you always exaggerate your losses and/or the governments benefit.

I think that claims in the us are 42 billion dollars lost annually. I followed THOSE studies back once, to find where the figures came from. That number is totally unsubstantiated as well, almost entirely based on guesses, estimates, and even false assumptions. One study after another cites the previous study, and almost no one knows where that 42 billion dollar figure came from, but it's impressive, so everyone continues to quote it.

Re:Story meaning?

[vivaelamor]

Nice calculator, I think the GP's main point though was that there is no evidence of a properly selected sample. You would be right in saying that the sample size has very little to do with anything compared to whether the sample is biased or not.

Re:Story meaning?

[treeves]

Thank you.
What was real strange was someone else objecting that this sample size was too *large*, and that TEN should be enough for any kind of statistical analysis.

Re:Story meaning?

[vivaelamor]

To me, the number is meaningless in itself. The fact that government agencies have been using the number is the issue. Either they knew that the number was wrong or they didn't bother checking it. Both possibilities can point to incompetence or malice and reflect very badly on the people responsible.

You might be happy with government by making shit up and gut feelings but for the rest of us this is a good example of why government has no respect.

Re:Story meaning?

[vivaelamor]

I really don't get why people are bothered about which way the number is pointing, it is completely irrelevant to whether piracy is good or bad. Unless of course you use the number to provide evidence that the piracy is having a harmful effect. I think the statistical likelihood of anyone using the number in that way is pretty insignificant though.

Re:Story meaning?

[wizardforce]

I just don't understand the stance that most people on this board seem to take regarding this issue. How can everyone be so supportive of what very obviously amounts to theft?

not everyone does obviously... most reasonable slashdotters advocate for reformed copyright pertly because of the unenforceable nature of longer copyright terms. many such as myself support the concept of a shorter more reasonable copyright term that does what the constitution requires: encourage the advancement of the arts.

If you do indeed use all file-sharing applications for 100% legit purposes, please educate me what you use these services for that makes them so very essential to cause these very emotional posts here.

most of the anger is directed toward the music/movie industry's response to piracy- weaken/destroy fair use, demonize all p2p [possibly restricting its use in the future out of fear] suing people as a scare tactic, excessive/un-constitutional fines, DRMed media etc...

Re:Story meaning?

[stuartdb]

I'm currently doing a stats paper at the moment (still basic stuff). I never thought I would actually use anything from it in the real world but meh....

Correct me if I'm wrong, but the calculations for sample size are only correct if it is actually a true SRS (Simple Random Sample) I couldn't find a link to the actual paper in the article but I think it would be safe to assume but the sample taken would not be a true SRS. It is more likely that this would resemble a self selecting sample, if that is the case, the calculations don't apply.

I'm not saying the data is incorrect, but without more detailed information (perhaps this information is in the actual paper) it is hard to make a conclusion. What is clear though is that the linked article is pretty much FUD.

Re:Story meaning?

[memristance]

One study after another cites the previous study, and almost no one knows where that 42 billion dollar figure came from, but it's impressive, so everyone continues to quote it.

Ars did an article on attempting to trace one such oft-quoted figure of losses, and slashdot discussed it. This $42 billion figure is probably from the same source.

tl;dr: big-dollar piracy numbers probably come from some unsubstantiated source from 1993 or earlier.

Re:Story meaning?

[hot soldering iron]

If they're going to "adjust" the numbers, why did they even bother doing the research at all? Why not just come out and say,"We didn't like what the numbers said, so we threw them away and we're making a WAG with some bullshit we're pulling out of our ass." I understand that they're a research (read "marketing") company, and so are constitutionally incapable of telling the plain truth because they could burst into flames, but it would be a new experience. And fun to watch!

Re:Story meaning?

[Pax681]

98.79% of statistics are made up on the spot - Vic Reeves

:P

Re:Story meaning?

[RKThoadan]

Although I agree with your analysis I'm not certain that I agree the effects of filesharing are negative for artists or even for the music industry. Some artists have run some of the numbers and don't mind. They make more money from concerts so wide exposure is highly beneficial. The music industry is more likely to be harmed, but even that isn't certain.

Re:Story meaning?

[HiThere]

How did they standardize their sample? I'd guess they picked either people it was convenient to pick, or people they thought would prove their point.

And I also think the sample size was too small, even given an unbiased sample. We used lots larger numbers to guess at journey-to-work times, and we KNOW that some of our predictions were wrong. We tested.

Re:Story meaning?

[kailSD]

Those "adjustments" are probably standard formula added to such.. erm.. polls. An obvious way to manipulate these surveys would be to carefully select the geographical location and demographic into a group. And then randomly select from that. That has a higher probability of getting the answer you wanted. If not, select (or deselect) more test subjects until the desired answer is obtained.

Re:Story meaning?

[dlthomas]

It does not "obviously" amount to theft. It *is* illicit, and it may be immoral (see Free Rider Problem), but it is not theft. If I steal 10 M&Ms from you, you have 10 fewer M&Ms - not the case if I download your song, in which case you have less than you otherwise would have *if and only if* I would otherwise have paid for it. This clearly is not the case for, say, college students with tens of thousands of dollars "worth" of media on their hard drive.

As for legal uses of "file-sharing" technologies, well - how about the entire world-wide web? We're sharing files...

Specifically P2P file-sharing technologies? Linux ISOs and WoW updates, to name two common legal uses.

Finally, I for one have an emotional reaction to assertions that technology should be restricted unless I can make you understand what it is for - and I don't even personally use any P2P software at the moment.

Re:Story meaning?

[Bigjeff5]

The second objection, and this applies to other studies too that try to make grand claims from small samples, is that it's A SMALL SAMPLE. For your survey to be representative, your sample has to be representative. It's also difficult to choose people independently at random, and without that assumption, all your basic statistics fall apart. Perhaps they went through a list of BT subscribers and pulled names at random -- but what if downloaders are overrepresented amongst BT subscribers?

You don't seem to understand the way good polling and statistics work. If you already have solid data on the demographic makeup of your population, it does not take a very large sample size at all to get accurate results. A sample size of 1000+ is more than enough to come within 3% accuracy (plus or minus) for any given study provided you already have good demographic information. To be accurate with a small sample size, you do NOT want to choose your survey takers at random, at least not completely. Specifically who takes the survey is random, but where they come from, what income level they fall under, how many computers they own, etc. should not be random at all. That's how you make a small sample size representative of the population, and can therefore get accurate results.

For example, if a census (which has a near 100% sample rate) 5 years ago told you that 75% of the population owns a computer, and 75% of computer owners use the internet, and 50% of internet users have broadband, you can get very accurate results with a sample size a fraction of a percent of the size of the total population by simply making certain that your smaller sample breakdown matches the larger survey. 100% of people surveyed should own a computer, since the survey would need to be 30% larger to include those who don't have a computer and still get the same accuracy (accuracy would be slightly better, but almost certainly not worth the expense). 75% of those people should have internet (you could start here instead with still very high accuracy), and 50% of those people should have broadband.

A result of 10% of people share files from the study that followed demographics and only used 1,000 people is going to be exponentially more accurate than a survey of 10,000 people chosen completely at random. To get any kind of accuracy with a pure random sampling you would need to sample a very large percentage of the total population. This is impractical and idiotic and not very useful.

Statistics done well are reliable, it's who's using the statistics, what they are saying about them, and what they aren't telling you about them that make statistics untrustworthy.

It's not the statistician who is the liar, it's the lawyer, or marketer, or politician who is the liar. It's their fault that 60% of statistics can be made to say whatever the hell you want them to say. That said, I don't trust any numbers given by the MPAA, especially when they arbitrarily adjust them up. More than likely the number should have been adjusted up, but the 5% figure seems rather pulled from thin air and unjustified. 2% or 3% would be more conservative, boosting the number of filesharer's by 50% just 'cause screams of desperation.

Re:Story meaning?

[jrumney]

Thanks, I'm now starting to count the time since I studied statistics in decades rather than years, but its starting to come back to me. So the norm is to assume 3 standard deviations, not 1 for the estimate of error.

Re:Story meaning?

[Bigjeff5]

How did they standardize their sample? I'd guess they picked either people it was convenient to pick, or people they thought would prove their point.

Had they done that, instead of using demographics (which is critical to good statistics and has a very very solid mathematical foundation), then a 3rd grader could pick apart the survey in their sleep and it would never see the light of day.

Since all I've seen are complaints from people who obviously don't know how statistics work, this is obviously not the case.

We used lots larger numbers to guess at journey-to-work times, and we KNOW that some of our predictions were wrong. We tested.

Did you use completely random samples? If so, it's no wonder your results were off. In that case, I'd expect them to be wildly inaccurate, since you obviously don't know what goes into an accurate survey.

Seriously, instead of just handing out fliers or putting a survey box on your website, you should do a little research on how statistics work, and maybe then you could design an accurate survey. Because doing it at random you won't get any reasonable accuracy until you're sampling 75%+ of the population. At that point, as another poster said, you aren't doing statistics, you're just counting.

Re:Story meaning?

[Bigjeff5]

Worse, 60% of statistics can be made to say whatever you want. - That one guy from that DirecTV commercial

Re:Story meaning?

[HermMunster]

This story piqued my interest because I used to work writing programs to collect statistics for analysis. I am also someone that watches the press regarding piracy claims.

When I read the summary I wasn't concerned about the 136 people being their basis. The concern I had the most with was that the article stated that there were approximately 136 people out of roughly 1100 that admitted to using file sharing programs, yet neither this summary nor the article stated whether these individuals used them for the purpose of copyright infringement. And finally, I don't see it stated anywhere where they advise how the sampling was broken out (i.e., age and profession).

Re:Story meaning?

[Bigjeff5]

When the lobbyists stand in front of lawmakers, those lawmakers want to know the real size of the problem.

Wow, that has to be one of the most naive statements about politics I've ever heard.

First, lobbyists don't "stand in front of lawmakers", they take them out to dinner, or a skiing trip, or chat up their aids, and hang out at the same exclusive restaurants and clubs. It's more of a "saddling up to" rather than a "stand before".

Second, the lawmaker who gives a shit about the actual size of a given problem is rare. What they want to know is a.) will their constituants go along with it b.) will it improve their ability appease their constituants (thereby ensuring their tenure in office) and c.) do they get anything out of it - fun stuff like food or trips, or more clout and influence among their peers.

On the issues where the politician actually cares, they are usually idiologically biased in one extreme direction or another, and so the actual size of the problem and actual truth of what is happening is meaningless anyway.

Big numbers and wild statistics are used to sway the constituants, not the lawmakers.

Re:Story meaning?

[Anonymous Coward]

>>most reasonable slashdotters advocate for reformed copyright pertly because of the unenforceable nature of longer copyright terms. many such as myself support the concept of a shorter more reasonable copyright term

>Why would the solution to something that is not easily enforceable be to make it legal? At the end of the day, the people that produced the content should be entitled to reap any benefits from it. If the market decides that it does not like the content, it will not pay for it.

So you're arguing that the people using p2p are paying? Or you're saying they're not paying, but in a way that's completely different from what you meant by not paying?

>> encourage the advancement of the arts.

>I don't see how this relates.

Because we're talking about p2p.

>>most of the anger is directed toward the music/movie industry's response to piracy- weaken/destroy fair use, demonize all p2p [possibly restricting its use in the future out of fear] suing people as a scare tactic, excessive/un-constitutional fines, DRMed media etc...

>As someone who makes a living creating copyrighted content, I don't see why these tactics are unreasonable.

As someone whoe makes a living creating copyrighted content, I do see why these tactics are unreasonable. I don't think you've *really* thought about this very hard. I didn't bother quoting the rest of your statement, as it operates under a mistaken assumption. Cheers.

Re:Story meaning?

[B3ryllium]

In statistics, a "population" doesn't necessarily mean "the population of a country".

It could just as easily mean the "population" of the classroom.

Re:Story meaning?

[Runaway1956]

"stand in front of lawmakers" Figure of speech - duhh.

As you point out yourself, the size of the problem is important. The amount of money involved is more important than anything - NAFTA was passed despite what the voters wished, because there was so much money involved. After the size of payoffs, the voters perceived size of a problem are considered.

Naive, huh? Right......

Re:Story meaning?

[Raistlin77]

most of the anger is directed toward the music/movie industry's response to piracy- weaken/destroy fair use, demonize all p2p [possibly restricting its use in the future out of fear] suing people as a scare tactic, excessive/un-constitutional fines, DRMed media etc...

...I don't see why these tactics are unreasonable...

So, just so that you can protect your "copyrighted content" from being stolen by someone other than me, you believe that it is "reasonable" to use bogus or flawed "research" to fool the government into a) taking away my legal rights (fair use); b) criminalizing software that can be and is used for legal purposes (P2P); c) abuse our legal system (suing people as scare tactic/impose excessive/unconstitutional fines); and d) crippling your "copyrighted content" so that I cannot exercise my right of fair use after I have purchased your "copyrighted content" (DRM/refer back to a) )?

It is even more difficult to attach a value to the legitimate uses of file-sharing networks, but if you can point me at examples of how file-sharing systems have a positive economic impact on anyone, please let me know.

Really? So you don't see value in a content provider being able to reduce operating expenses by distributing their content via P2P? Just because you are too lazy to do a simple search using any common search engine doesn't mean such examples don't exist. And why exactly does it have to have a positive economic impact on anyone - why does it have to have any economic impact at all? There are many things that have neither a positive economic impact nor any economic impact whatsoever, should those be illegal too?

Re:Story meaning?

[Trogre]

I don't know what you mean, but I'm outraged.

Outraged, I said.

Re:Story meaning?

[FrozenGeek]

It is not a question of supporting theft. The RIAA, et al, use these stats, and similar data, to try to influence governments to ban file sharing of all sorts as well as to implement other laws that many tech users find oppressive and which do punish those who are not guilty. The **AA groups have proved to be total jackasses in their attempts to defend their interests, alienating large segments of their customer base as well as people who are neither part of their customer base nor among those harming their interests. Such people will respond negatively to anything that might be perceived as supporting the **AA groups. As far as these services being essential, "essential" is irrelevant. If I chose to share data, which I am legally entitled to share, it is neither your business nor your right to tell me how I may share said data. If I am not legally entitled to share data, the method by which I share it is irrelevant - I am breaking the law regardless of how I share it.

Re:Story meaning?

[Bigjeff5]

No, you missed the point of that post.

The point was that the sample size has almost no bearing on the accuracy of the survey provided it is truly representative of the overall population.

If you can get a sample size of 10 that is representative of a population of 60,000,000 people, you'll have a pretty accurate survey. The reality is, that's not possible in most cases. You'll generally have more than 10 demographics of varying percentages of the total population, making 10 simply too small. 1000, however, is not too small unless you are looking for very, very small percentages of the population. I.e if you are expecting results of less than 2%, a sample size of 1000 is too small because the margin of error is around 3% - you could easily run the survey and get no positive hits at all. For that survey, you'd probably need to bump it up to around 10,000 to drop the margin of error low enough to get reliable results.

Since the results they got were 11.6%, and the margin of error was about 3%, you can very reliably say between 8.6% and 14.6% of people use file sharing software.

I don't like that they added 4.7% to their figure without anything to back that up, especially since that is nearly 50% of their results. They basically said 30% of file sharers lie about being file sharers, without any data to back that up. They also used 40 million as their figure for people on the internet, when the government survey states something like 33.5 million.

The numbers they should have used were 2.9 - 4.9 million people use file sharing software. That is accurate and can be backed up by statistics. It is probably more like 6 million due to people lying about using file sharing software, but that's still just a number I pulled out of my ass, and not statistically accurate.

Re:Story meaning?

[Attila Dimedici]

The effects of filesharing are negative for artists (certainly for pop artists), and especially for the "music industry". There can be little doubt about that.

Sorry, I think there can be significant doubt about that. There have been several studies that indicate that those who obtain music illegally (primarily by filesharing), also, buy more music than those from the same demographic who don't obtain music illegally.
Just from my personal experience this is supported. I have five or six friends who download a lot of copyrighted music without paying for it. They also buy quite a bit of music. I have never downloaded any music. I have also bought fewer than 10 CDs in the last 10 years.

Re:Story meaning?

[jedidiah]

This is EXACTLY why you can't trust the result of a survey like this.
There is always going to be some JACKASS that tries to equate ANY
filesharing with "piracy". Not everything available for download on
the web is OPP. A great deal of it is perfectly legal to download or
to share with friends.

Re:Story meaning?

[jamstar7]

3) The rant about examining only 1,176 people for the study - in which case the same kind of tv viewer statistics and other studies are made in what case.

So could someone please explain *why* is it a questionable research.

Because if there were actually 33 million people online in the UK and they only surveyed 1176 people, they only sampled 0.003563% of the people online. Anybody who has even slight exposure to statistics can tell you, a sample size of .003563% of the whole is statistically irrelevant. When you start looking at 1% or greater, you start getting relevance. 1% of 33 million is 330,000 people.

Re:Story meaning?

[Colonel Korn]

It doesn't really make sense to claim "sample size is small" for an 1,100-person sample. If the sampling was done in a random, unbiased manner, that size sample gives a margin of error of +/- 3%. If there are flaws in the sampling method, that's another thing, but the sample size alone doesn't seem problematic, unless you need accuracy better than +/- 3%.

If the sampling was done in a random, unbiased matter? That's one of the main problems. Another is rounding up by 50% to account for under-reporting. Preposterous.

Also, as a rule of thumb, never trust root N statistics on survey data. Systematic error is always huge and there's no reason to be confident that a normal distribution really applies. Notice how election polls with 10000 participants will show their root N error (+/- 1%) and disagree with similar polls with the same number of participants done by a different source by 8%. All the time.

Re:Story meaning?

[MeNeXT]

He was correct, by the way. He went on to rant that anything that uses large amounts of people (by which he meant more than at most a few dozen) was not proper statistics.

10 people could in no way make any reasonable assumption of any population. He was wrong. First and foremost 10 randomly selected people of any population would not represent any complete demographics of any nation. Rich, poor, middle class, professional, artist, unemployed, computer literate, music fan, and student does not come close to representing the whole segment of any nation. I am positive that you could fill in hundreds more. Some are more prone to "Download than others". The combinations and permutations may affect the results. The sample size needs to be large enough and random to arrive at any reasonable accuracy.

Re:Story meaning?

[Colonel Korn]

1176 people compared to 10 comes a lot closer to real truth. Close enough to have just a few % error margin, so in statistics it works good.

You assume a normal distribution, which is debatable, and no systematic error.

Re:Story meaning?

[wizardforce]

Why would the solution to something that is not easily enforceable be to make it legal?

because it doesn't work? why are our police resources being used to enforce extended copyright law when it is neither enforceable nor in the public interest to do so?

With this particular issue, it simply became trivial for virtually anyone to obtain copyrighted material illegally

hence the law is unenforceable- that is to say that it can't be enforced without far more draconian measures that violate other rights.

Nobody is going to stop the advancement of the arts if it is made more difficult to share copyrighted content

all it has to do is discourage the advancement of the arts relative to an alternative solution. In that case the copyright system as it is would be unconstitutional in the US.

As someone who makes a living creating copyrighted content, I don't see why these tactics are unreasonable.

those tactics are often illegal, rights violating and unconstitutional. suing people for 10,000 x damages is a violation of the 8th amendment. various practices by the RIAA/MPAA are illegal including but not limited to violating the DMCA, abuse of the legal system, fraud and entrapment...

but if you can point me at examples of how file-sharing systems have a positive economic impact on anyone, please let me know.

live cds, distribution of software patches, advertising which ADV films uses P2P to distribute advertising clips for their anime media, distribution of creative commons licensed materials etc...

ALL CD stores but one have been driven out of business, and virtually everyone I know has stopped purchasing CD's

I'm sure that had nothing to do with single tracks being sold on Itunes, the poor state of CDs released today or the recession.

Re:Story meaning?

[Loomismeister]

Well we don't really know what the actual numbers are, but no I don't have a problem with anything. The music industry has a problem!

Re:Story meaning?

[blueg3]

Just as easily as a random sample can accurately reflect a population as a whole, it can equally be skewed to be a completely inaccurate representation of the real world.

If by "just as easily" you mean "with an enormously lower probability", then yes. But then, that's what a statement of margin of error says.

Statistics isn't all that complicated, and what a statistical measure means can be both demonstrated and proven. You don't need to get all faux existential about how "it's all just a bunch of crap, man". You don't know what you're talking about.

Also, entropy? No such thing as random? Really? Don't inject physical phenomena you clearly don't understand in a discussion about pure mathematics.

Re:Story meaning?

[Dan541]

100% of internet users use file sharing software. I dont see how you can avoid it.

Re:Story meaning?

[tsm_sf]

Please stop posting to /.

Re:Story meaning?

[koxkoxkox]

Just as easily as a random sample can accurately reflect a population as a whole, it can equally be skewed to be a completely inaccurate representation of the real world.

Of course, but not more than 5% of the time. Please read a bit more about statistics and maybe listen to real statisticians instead of journalists before spewing so much hate. Statistics don't try to prove anything, they are just tools we can use to make decisions.

Re:Story meaning?

[julesh]

It doesn't really make sense to claim "sample size is small" for an 1,100-person sample. If the sampling was done in a random, unbiased manner, that size sample gives a margin of error of +/- 3%.

Yes, it does, because the size of the result is too small. A 3% error on an 11.6% result is over 30% of the result. When scaled up to the whole population and rounded (as done with the figure when it was used) the error bars on the result become "between 5 million and 9 million people" which is somewhat different to the 7 million claimed.

Re:Story meaning?

[Karhgath]

ok, 6.7m from 7m, bad rounding.
11.6% fudged to 16.3%. Bad.
The estimate of internet access is a valid statistic tho and doesn't impact the study (but do read below!)
If I did my calculation right, they estimated 40m have internet access on 60m total pop.

So, in the end, lets use the 11.6% of 40m. instead of 16.3%.

If you do the math, the 6.7m falls to around 4.6m.

6.7m to 4.6m. Ok, they fudged the data, true. But 6.7m and 4.6m is pretty much the same order of magnitude for me is it? Far from the "136" people from TFS.

In fact, unless they only surveyed people WITH internet access, the sample is big enough to compare to the WHOLE population. That means they downsized the resulting pool even more by using the estimate number of people with internet access instead of the total population.

If you don't understand, let me explain: you have a population of 100, and 60 have access to the net. If you survey 10 random sample, about 6 of them will have internet access, so a *maximum* of 6 will admit to filesharing (since the 4 other don't have access). The % of internet access is already represented in the sample, no need to reduce it even further.

So lets take 11.6% of 60m instead. Ohhhh... 6.96m!! Thats pretty much 7m.

Now, the ONLY criticism is that filesharing doesn't automaticall equal piracy. However, you have to admit, I doubt 50% of people who fileshare do it totally legally. Event then, we'd go from 7m to 3.5m, that's still the same order of magnitude to me.

The other assumption is that filesharing implies internet access. Unless most of them fileshare at their grandparents or friends' house, I doubt this is a very big issue.

Hey, I hate the MPAA and BPI as much as the next guy, but the stats are there. You cannot dismiss the stats juste because you don't like the result.

Now, however... this is the most important part: Does the result means anything beside a number of filesharer?!?! that's something else entirely! You can make that result say anything you want, it's just a stat afterall (albeit accurate). Some will say "that's ONLY 7m" other "wow, 7m, that's huge!", etc. Like you say, that's how you lie with stats. I'm pretty confident that the stat is accurate, but what does it means? That's the thing we should talk about.

Re:Story meaning?

[Merls the Sneaky]

Also...... "That 11.6% was adjusted upwards to 16.3% 'to reflect the assumption that fewer people admit to file sharing than actually do it."

How do they know the amount of people that "fewer people admit to file sharing than actually do it" involves inflating the numbers by exactly 4.7%. Where and how did they get that data?

Re:Story meaning?

[Merls the Sneaky]

11 million world of warcraft players regularly use file sharing as a means to receive regular game updates. They could easily have picked up a bunch of those players and lumped them in with the "illegal" file sharing crowd. The study is bogus because it doesn't account for those situations.
 

Re:Story meaning?

[julesh]

About 60 million people in the UK, sample size of 1,176, confidence interval of 96% gives a margin of error of 2.99%. So, it's 96% likely that they got within 2.99% of the right answer (to the question of how many people admit to it).

Yes, but 11.6% +- 2.99% covers quite a wide range of values. Using the same process to inflate the figures and rounding up as in the original reporting, you get a figure of between 5 million and 9 million. Without guessing at how many people are lying and using the more accurate figures for number of Internet users the range becomes 2.9 million to 4.9 million, which is somewhat different to 7 million.

Re:Story meaning?

[julesh]

To me, the number is meaningless in itself. The fact that government agencies have been using the number is the issue. Either they knew that the number was wrong or they didn't bother checking it. Both possibilities can point to incompetence or malice and reflect very badly on the people responsible.

Not only that, it's actually a violation of government policy, which is to only use/publish statistics that have been approved by the Office for National Statistics.

Re:Story meaning?

[jamstar7]

It is not a question of supporting theft.

It's copyright infringement, not theft. If I hotwire your car and drive it off, that's theft as I've deprived you the use of the car. If I make a copy of your notes and leave you your notes, I haven't stolen them. If they're copyrighted Cliff notes, I've infringed on their copyright. The difference is, I have your car, but you still have your notes, and I have a copy of them.

Re:Story meaning?

[stephanruby]

So could someone please explain *why* is it a questionable research. It is like every other study where you study small amount of people and make estimates based on it to reflect whole population. Usually this amount of people also gives somewhat correct results on the whole population. Theres some error margin, but its close enough.

It's the fact there is a conflict of interest here. It's the same reason you don't let your own sales people do the final tally of their own sales, you have your own CFO/accountant do it. And it's the same reason you don't rely on your ad agency to tell you how successful their campaign was, you hire an independent third party usually Nielsen Ratings (which actually does way more than just television ratings) to tell you if the money you spent on the ad agency was even worth it to begin with (or not).

It's a question of proper financial controls and divisions of duties, because if the people paying for the study have a financial incentive to lie to you -- they *are* going to lie to you -- that much you can be certain of.

Re:Story meaning?

[CAIMLAS]

Their sample size was outrageously small, and then they padded their figures based on their presumptions/desired result.

Example: out of myself and three of my friends, all of us smoke. None of us have lung cancer. So, by extension, you can imply that smoking does not cause lung cancer in 100% of those studied, and using the metrics of the topic's study, we can imply that there is no lung cancer in America, because 100% of smokers do not have lung cancer.

It's weak, but it's the same basic thing. It's rife with logical fallacy.

Re:Story meaning?

[siloko]

So you swallow without concern the presumptions of an industry sponsored report yet you

don't buy for a second that people who claim to only use file-sharing apps for legitimate purposes

are actually telling the truth. Well I guess it's a good job you aren't doing any research as your objectivity is seriously askew! Some presumptions you seem happy with:

• Using file sharing software = knowlingly downloading copyrighted material
• Abitrarily adjusting the reported 11.6% upwards to 16.3% because people lie when asked questions
• Extrapolating from a survey of 1100 to make presumptions about a net-connected population of 40 million (despite the government's own figures suggesting only 33 million peope in the UK have access to the internet

Given that the industry in question has a poor track record in producing objective, fair analysis of the current crisis in their business model I find your naive acceptance of their publicity touching. From the article:

If the BPI-commissioned Jupiter research had used the Government's online population figures, the total number of file sharers would be 5.6m. If the researchers hadn't adjusted their figures upwards, the total number of file sharers would be only 3.9m - or just over half the figure being bandied about by the Government.

and a nice comment was left stating that it's good to see the governments target of reducing illegal filesharing by 60 per cent has almost been achieved - touche!

Re:Story meaning?

[BlackCreek]

The study has bias showing all over (as the research group was paid by an interested party). There are obvious procedural flaws being made: (1) file sharers == priracy; (2) the hand-waving adjusting of the numbers; equating "have used" to "use it frequently".

As the research firm has clear obvious bias, and such a lack of conformity to proper statistical procedures, assuming in good faith that the sample was done right is downright stupid.

Re:Story meaning?

[gilgongo]

I just don't understand the stance that most people on this board seem to take regarding this issue. How can everyone be so supportive of what very obviously amounts to theft

I suggest you research the issues around all this then. Illegal file sharing is a symptom of something much, much bigger. Personally, I don't "support" it any more than I support the taking of illegal drugs - but I oppose the Drugs War. It's on that level we are talking.

Incidentally, I have my own (possibly twisted) morality here. I believe that copyright should last for 14 years renewable on application for another 14, and not in perpetuity as they (in effect) last today. I therefore freely admit to imposing this view on rightsholders by not downloading anything created after 1995. This is therefore another facet of the issue: I download because I can. Rightholders need to wake up to this fact and adjust their business models accordingly.

Re:Story meaning?

[iJusten]

1) Usually pro-filesharers try to make it sound like filesharing is usual activity and try go for most or 70-90% user share

It probably is 70-90% in certain age groups. The biggest file-sharers would be people without money and with good computer skills. This means mostly people under 25, as older people have less need to obtain their entertainment illegally. After all, to get the population to 90% file-sharers, it means that grannies and tottlers both would have to take part!

As far as the usability of the report goes... even as we can assume that 1176 people is big enough group (apparently with around 5% margin of error), they basically tweaked every possible figure upwards. Instead of 33 million users of Internet, let's say 40 million (I'm sure there's official records for this), people will lie, so let's adjust to that (why would people lie in anonymous poll?) and lastly, YOU DON'T SHORTHAND THE +/- PART INTO POSITIVE AND ADD TO THE RESULTS.

Not to mention that some of the sourced research was reported wrong. The same author, but different funding and parametres.

Re:Story meaning?

[dna_(c)(tm)(r)]

The study is effectively making the assumption that filesharing = copyright infringement.

I have a very hard time believing that the vast majority of people that use any filesharing application do so exclusivley for legit and non-copyright infringing purposes. [...]

I have used bittorent for downloading Mandriva about 6 years ago, am I a copyright infringer now? What if you use FTP, WebDAV, etc?

It's a devious question, they should ask "Do you illegally download copyrighted material regularly?" and "How often do you illegally download copyrighted material?".

As the question is formulated, you can argue with the same level of truth:

"Filesharing in UK cost industry only 1 248 637 GBP"

(6.7 million, round down to 6, 50% overestimated, 4m, 30% can't actually use the software 2.8m, 18% of downloads are unusable, 2.296m, 10% downloads too slow, 2.06m, 1 song 99 USD cents, 1 GBP = 0.61 USD)

Re:Story meaning?

[ericlondaits]

There are ways to measure lies in a survey. I'm not an expert, but when I took a statistics class it was explained, and as well as I remember:

- Instead of the 1.000 surveyed people separate a smaller group, say 100.
- Assume the last digit of their telephone numbers are uniformly distributed
- Ask them something along the lines of: "If you lied about file sharing or if the last digit of your phone number is an even number say YES, if not say NO"

If nobody lied you should get something close to 50% YES and 50% NO... but if the result is different you can assume it's because some of them lied (although you don't know which). ... Doing this kind of test you can approximate the percentage of people that would lie to the question.

Re:Story meaning?

[namgge]

They admitted to using file sharing software not pirating goods via said software... The study is effectively making the assumption that filesharing = copyright infringement.

This does seem important to me. At work, everybody uses network drives as a method of 'filesharing' (spreadsheets of the latest paperclip audit, the coffee club accounts, etc., but not music thanks to the BOFH's watchful scripts.). They also do this from home using the vpn. I'm sure if you asked any of them if they 'used filesharing software' they would reply 'yes'. However, the majority would have no idea what peer2peer sharing was.

namgge

Re:Story meaning?

[coaxial]

1. the same size is small.. probably too small to make the claims they did.

Probably? So you don't actually know, but instead are just going to rely on the old saw of "Well they didn't ask me! I don't know anyone they asked! Those muckity muck scientists aren't so smart!"

The population of the UK is approximately 60,943,912. Assuming your draws (i.e. choosing who to ask and who answered) are uniformly random, and independent and identically distributed (i.e. the answers of any one person does not effect the answers of anyone else, nor who you ask) the size of your sample for being 95% sure (i.e. your "confidence level") that your results are within 4 percentage points (i.e. the "confidence interval" (i.e. the +- number seen at the bottom of polls)), then you need to a mere 600 random individuals. If you want to be 99% sure that number of file sharers are within 4 percentage points of you counted, then you need to ask 1,040.

Assuming the sample is unbiased, asking 1,176 individuals from a population of 60,943,912 with a confidence level of 95% in your 11.6% result gives you an error bar (i.e. confidence interval) of 1.83%. Going up to 99%, your error bar increases to 2.41%.

2. they altered the numbers on an estimate of how many people fileshare on the assumption that the number was under-reported

This is known as controlling for sampling bias. Not only is it an long established and mathematically proven method, but to not control for sampling bias is sign of shoddy work. Work so shoddy, that you would fail an undergrad statistics course.

3. conflict of interest... it's like the tobacco industry sponsoring studies claiming that smoking doesn't have anything to do with lung cancer... there is significant reason to believe that the study carries significant bias in favor of their conclusion and must at the least be repeated by other sources.

Same can be said about your unsupported allegations of bias, especially given your lack understanding of long established statistical sampling techniques.

N. real statistics researchers know that this study has numerable crippling flaws and should not be held as gospel by anyone. Even a first year stats student can see it. The reason this story is important is that it may influence governmental policy and it's flawed... That's danger

At least the irony of you making this statement isn't lost on one of us.

In short. Go to school and take a stats class.

Re:Story meaning?

[BlackPignouf]

Actually, entropy is just that : pure mathematics & statistics.
It can be applied to thermodynamics, though.

Re:Story meaning?

[dna_(c)(tm)(r)]

• 1 x standard deviation = ca. 70%
• 2 x standard deviation = ca. 90%
• 3 x standard deviation = ca. 95%

Re:Story meaning?

[1u3hr]

So could someone please explain *why* is it a questionable research.

Someone did. RTFA, not just the summary. (Actually, the whole problem is people not bothering to read where information comes from.)

Aside from the source being a survey commissioned by a partisan body, with no auditing, though presented with the authority of the government; even if it were totally accurate, several "assumptions" and "adjustments" were then made that served to double the figure originally found.

If the BPI-commissioned Jupiter research had used the Government's online population figures, the total number of file sharers would be 5.6m. If the researchers hadn't adjusted their figures upwards, the total number of file sharers would be only 3.9m - or just over half the figure being bandied about by the Government.

Re:Story meaning?

[1u3hr]

How can everyone be so supportive of what very obviously amounts to theft?

I don't know about "everyone", but IN FACT, copyright infringement IS NOT THEFT. Despite all the "You wouldn't steal a car..." crap that is pushed to try to equate the concepts.

And yes, copyright infringenment is illegal, but it does not "obviously" amount to theft. Because NOTHING IS STOLEN for a start.

Re:Story meaning?

[selven]

Most people don't even know about copyright or that file sharing copyrighted stuff is infringement, and they don't know about free music/books/software.

Re:Story meaning?

[Tony Hoyle]

They were volunteers?

That makes the results essentially random. There's no way you can compare a self selecting group of volunteers with the entire internet using population.

Re:Story meaning?

[Tony Hoyle]

100% of computer users use file sharing software. Windows, Linux and OSX can all share files.

Re:Story meaning?

[4D6963]

Excellent Summery

Statistics are hard, and so is gramer.

Re:Story meaning?

[4D6963]

Regarding 1 and 2, all polls use that sample size, and pretty much all polls are weighted somehow. It doesn't necessarily make them less accurate. So you're really talking out of your arse.

The reason this story is important is that it may influence governmental policy and it's flawed... That's dangerous.

Not really. This poll and its weighting gives an accurate idea of the order of magnitude of the file sharing phenomenon. It might very well be 30% off, it doesn't matter much, it doesn't change the picture a lot. Besides, I believe it's in the interests of file sharers to appear as large a movement and as ubiquitous as possible. Governments tend to think twice before coming down hard on something that one out of 6 citizens do.

Re:Story meaning?

[vivaelamor]

Since all I've seen are complaints from people who obviously don't know how statistics work, this is obviously not the case

What the hell? You're accusing others of being idiots and you come out with a statement like that? How does lots of people talking about something they don't know about validate the study in any way?

Re:Story meaning?

[vivaelamor]

It sounds like the original publication was merely for some commercial interest. They should not have been used by the government in any way whatsoever. The government has a whole department devoted to statistics, there should be no excuse for this sort of thing and the sooner politicians come up with some sort of rulebook for verifying this sort of thing the better.

Re:Story meaning?

[walshy007]

At the end of the day, the people that produced the content should be entitled to reap any benefits from it.

Just how will they reap the benefits of it 75 years after they are dead?

Its not like artists are actively prevented from creating content.

But.. they are, if practically eternal copyright had existed before, disney would not have been able to make most of it's movies. Just as people today cannot take those stories and be creative with them making their own adaptations.

This may surprise you, but very little of what we make is actually new, it is merely different combinations of already existing ideas, when you disallow the partial use of existing ideas, you are stifling creativity.

The idea of copyright is to give people incentive to create new works, what incentive is there to create new works when you can sit on already completed works for longer than a lifetime?

Re:Story meaning?

[vivaelamor]

I don't think that's how the science works. When you are reporting in percentages, the difference between 1000 and 10000 is insignificant.

Re:Story meaning?

[Runaway1956]

You seem to be working from an assumption of some sort. How about we consider the taxman coming to your place of business to conduct an audit. He finds that you owe 7 million dollars in taxes, instead of the 4 million dollars that you claimed. Can you argue that the numbers aren't off by an order of magnitude? "But, sir, my numbers are only wrong by about 75%, this isn't fraud!" Or, we can work those same numbers backward - "But, sir, my numbers are only off by about 50% - of course it's not fraud!" Good luck with that, huh?

The numbers are fraudulent, plain and simple. As others have pointed out, anyone in the scientific field(s) would be laughed out of academia for submitting such flawed numbers and such flawed reasoning.

"In fact, unless they only surveyed people WITH internet access,"
BTW - TFA specifically says that everyone in the survey had internet access.

There is no line of work on planet earth where people are permitted to do such obviously fraudulent math. If similarly flawed mathematics were applied to a construction job by a bunch of backwoods hill billies, they would soon be out of business.

You simply cannot justify the numbers with any sort of logic. Attempting to do so is an exercise in fraud.

Re:Story meaning?

[commodore64_love]

>>"Please stop posting to /."

Frak off Censor.

Re:Story meaning?

[vivaelamor]

I think the big point is that there isn't an actual paper to link to. The BBC radio guys phoned up the people who did the original survey and were told some of the figures over the phone along with an assurance that it was done properly. While that may be good enough for a commercial company who just want to quote a number a nice sounding number from a statistician, government should be a lot more careful about where they get their numbers.

I would recommend you listen to the BBC iPlayer link on TFA. When the government has its own statistics department (which was set up by this current government, not that long ago) they better damn well use it. If they did use it and these statistics are actually approved then the department need to be hit over the head with a cluebat for having no quality standards.

Re:Story meaning?

[Clairvoyant]

Did you RTFA? They've rounded up several figures during their calculations. You're using one single point (out of, what, 5?) a straw man fallacy.

Re:Story meaning?

[DarKnyht]

I cannot speak for anyone else, but I use filesharing for downloading and distributing Ubuntu Linux (torrents) and to download legal content such as the World of Goo soundtrack. I believe that downloading copyrighted material that is not freely distributed is wrong and refuse to participate in it even if I believe the laws behind copyright are seriously flawed.

So it is wrong to believe that a file-sharing application users is automatically a copyright infringer.

Re:Story meaning?

[roguetrick]

Superb retort.

Re:Story meaning?

[kbrasee]

Gramer != Speling

Re:Story meaning?

[Narpak]

I don't think that's how the science works. When you are reporting in percentages, the difference between 1000 and 10000 is insignificant.

Indeed. Which is why I suggested that to arrive at a more comprehensive and accurate estimation one would have to take larger samples and to ensure that those samples are taken from more than one district of the region one wish to cover. Taking a sample of 1000 people and then try to make an estimation about a population of about 61 million is, as I said, far too insignificant a number to allow for accurate predictions.

Re:Story meaning?

[blueg3]

No, it's a thermodynamic concept that has been extended to information theory. Prior to major developments in statistical mechanics, entropy was a loss of energy associated with physical transformations. This significantly predates both statistical mechanics and information theory. Stat mech formulated the modern definition of entropy, and Shannon applied it to information theory.

Re:Story meaning?

[SL Baur]

So could someone please explain *why* is it a questionable research.

It's not, or at least not explained in TFA. The sample size, if it is random should provide pretty reliable statistics, but ... I use file sharing software to download Linux ISOs and with World of Warcraft, never music or movies, so they probably asked the wrong question. In which case, the number is totally meaningless.

In terms of a car analogy, it's like asking if you ever killed something while driving and not distinguishing between road kill and your neighbor's children and deciding that everyone who ever drove a car and admitted to killing something was involved in a hit & run.

Re:Story meaning?

[vivaelamor]

Taking a sample of 1000 people and then try to make an estimation about a population of about 61 million is, as I said, far too insignificant a number to allow for accurate predictions.

If you have a margin of error of 1% and a confidence level of 99% then 16000 would be about right for 61 million. However, firstly the story is about internet users not total population and secondly there is no indication that they were trying to be that accurate. 1072 is perfectly adequate sample size for a confidence level of 95% and an error margin of 3% with a population of 61 million.

Re:Story meaning?

[SL Baur]

I have a very hard time believing that the vast majority of people that use any filesharing application do so exclusivley for legit and non-copyright infringing purposes.

It is a mistake to claim to that *everyone* who uses file sharing software does so for nefarious purposes. I do not claim to be in the majority, but I can state with certainty that such people exist. I have only to look in the mirror.

Re:Story meaning?

[SL Baur]

More like lies and FUD. The sample size is high enough to get very good statistics if it's a random sample.

The issue is that are all people who use file sharing software doing it illegally? Everyone who subscribes to World of Warcraft uses file sharing software. That's certainly not an illegal use.

It's a horrible article, written by an idiot and I would have more expected this coming from kdawson rather than from soulskill who usually has a clue.

All uses of file sharing software are not "illegal".

Re:Story meaning?

[4D6963]

Ah yeah, good point.

Re:Story meaning?

[magus_melchior]

The logical fallacy is the argument "{All} file sharing is copyright infringement and/or theft" (known as a "fallacy of accident"). What you and other advocates for the status quo are saying is that there is no other purpose for file sharing software than copyright infringement. (One may argue that "file sharing = copyright infringement or theft" means "SOME file sharing...", but a categorical assumption such as this one will generally be taken as "ALL file sharing...")

Although it's true that the lion's share of file sharing activity infringes copyrights (which is an entirely different argument than the one I posited above), it is not the case that this is the exclusive use of file sharing software.

Ask anyone who obtained a Linux DVD electronically. Ask anyone who used BitTorrent to grab OpenOffice.org.

Re:Story meaning?

[JAlexoi]

The problem is that the government is taking numbers from a statistical research paper, that was commissioned by a biased entity.
We all know, that statistics tend to sway towards the point of view of the paying party.
And then we get a problem with Wikipedia Source paradox. The state takes numbers from a biased corporate study and present's them as fact, then that entity will say that "According to the government..." and present that study as official government sponsored study.

Re:Story meaning?

[Bigjeff5]

Ask anyone who obtained a Linux DVD electronically. Ask anyone who used BitTorrent to grab OpenOffice.org.

Or anybody who plays World of Warcraft, or Guild Wars, or a dozen other new games. I'm not sure if it's true for WoW, as I don't play it, but I know it is possible to purchase Guild Wars online, download a small client from their website, and get the entire game via a torrent P2P protocol.

At least 11+ million people use P2P filesharing for legitimate reasons, and that is a quantifiable fact. That they do not realize they are using filesharing software does not really matter, the fact is they are, and it is stunning proof of its legitimacy. There are a number of companies, like FOX, Lionsgate, Paramount, ComedyCentral, Sega, Warner Bros, Mtv, Starz Media, G4, Kadokawa, and others who all use P2P to reduce their distribution load.

As popular as "pirating" videos and software is, p2p is becoming more and more of a business, and these idiots at the RIAA and MPAA just seem intent on staying behind the curve. Pretty soon they will be stepping on their own toes, making it harder for their own companies to make any money in the competitive market. You can already see this with their inability to produce anything similar to iTunes or Amazon's music store, inspite of flagging CD sales. All they know to do is point their finger at a small fraction of the market who, in reality, would likely not have purchased anywhere near that amount of media anyway. They may think they lost $1 billion plus in sales, but the truth is probably closer to $100 million, if that. If the people pirating software were not able to do so, they probably would not have purchased much of it at all.

Frankly, I'm too lazy, and my internet is too expensive to do any downloading, so I just go without. And it doesn't bother me.

Re:Story meaning?

[jmerlin]

In statistics, the smaller your sample size in relation to the whole population, the less accurate your predictions will be (generally speaking). Your professor was right -- if you poll EVERYONE in the population, you no longer need statistics as you aren't making any guesses, you're just reporting numbers. But if you only poll 0.1% of the population, your estimations are almost certainly going to be horribly inaccurate. If your professor used a contrived example from sample data he produced or gathered somewhere about which he knows certain information, he can most certainly make a very accurate prediction for "see, this is my point" purposes. It does not make a case for real statistics. If he had 10 people in class answer questions, then he was able to make an estimate about the population (your class), then he's got a 20% to 50% sample size (assuming your class is between 20 and 50 students). At that size, an estimate is much more accurate.

I do believe the 'stats' you're taking is some kind of engineering stats or freshman/sopho? Let's just say if you had taken a mathematics stats class where you have to prove things like why the assumptions behind your models must be accurate, why certain data transformations work, etc etc etc, or one where the professor actually gave you a data set he knew NOTHING about and worked with it as you did -- true statistics teaching btw -- so that you can see how you can gain information from a sample and what's valid etc, you wouldn't be 1) posting as anonymous coward [and you did because we all know you have no education in statistics] or 2) making claims as ludicrous as the ones above. You sound like just another conspiracy theorist nutjob. GTFO slashdot please.

Re:Story meaning?

[clive_p]

I have used file-sharing software and I'm in the UK. I downloaded the BBC I-player, and have used it a few times. That may well account for most of the 11% of users who admit to using file-sharing software

Re:Story meaning?

[FrankieBaby1986]

Because 1,176 people is a miniscule (0.0168 %) amount compared to 7 million, so there is room for a LARGE margin of error, in either direction. The sample size is too small for the number of people they are trying to represent.

Re:Story meaning?

[digitig]

More like lies and FUD. The sample size is high enough to get very good statistics if it's a random sample.

The issue is that are all people who use file sharing software doing it illegally? Everyone who subscribes to World of Warcraft uses file sharing software.

That's a point that I flagged to the program's discussion page. The main points made by the program were that the data didn't come from where the government said it did (an independent research organisation) but rather came from an unpublished study by an organisation with a vested interest, and that all of the assumptions made pushed the data in the same direction -- the direction wanted by the organisation who funded the study.

Re:Story meaning?

[Metalloy]

In response to your (very naive) *why* is it a questionable research : In statistics, there is a something called the "confidence interval" which relates the "confidence" in an extrapolated average to the size of the sample taken. Mathematically, the greater the sample size, the greater the "confidence". Now for our case, where 1,200 samples were taken and working the arithmetic back : 7M = 16.3% of all the net users, therefore these are approx. 43M people, so the sample size is only a meager 0.002%, which can bring down the confidence level to the floor.

Re:Story meaning?

[HiThere]

For our purposes we needed samples from every census block-face to get decent error control. This would have been too expensive, and was impossible anyway, since for many kind of data the census didn't release information at that level of detail. So we used census tracts, and occasionally aggregates of census tracts.

It made things more difficult that between two censuses many of the census blocks were split, or otherwise redrawn. (Sometimes this was inevitable, say when a freeway went through.)

We also collected employment data, but that came at a different level of detail. We eventually worked with map zones, most of which were aggregates of census tracts, but sometimes we needed to split census tracts partially into one map zone and partially into another.

We verified this procedure with a smaller telephone survey to make certain, among other things, that our data handling hadn't munged anything. And to collect sample information about where people in certain places tended to work.

At that point we made multiple models of how this would work out, fit the models to the data, and then checked which models made a best projection.

And we STILL didn't have the gall to say that our guess was the truth. Merely that it was the best guess that we could make. We provided a range of models as our best guesses as to what would be the result of potential choices (where to build which freeway, what to put the bridge tolls at, etc.). After twenty years I'm pleased that our final model was approximately true. But we didn't say that we knew it would be.

These people strike me as arrogant and reckless with their claims on small amounts of data.

Re:Story meaning?

[easyTree]

Why is flamebait -1 rather than +1 ? Huh?

Re:Story meaning?

[easyTree]

The study is effectively making the assumption that filesharing = copyright infringement.

Are we really to believe that out of every one of those thousands of times we've all seen it implied in print that 'file sharing == illegal', the authour has merely been sloppy or less precise than they might ? OR is there actually a campaign to convince people that sharing files is somehow wrong. If so, why might that be?

Re:Story meaning?

[easyTree]

I just don't buy for a second that people who claim to only use file-sharing apps for legitimate purposes only actually do so.

So, if you're not paying, I hope you're able to live with the onslaught from your conscience.

Re:Story meaning?

[easyTree]

most of the anger is directed toward the music/movie industry's response to piracy- weaken/destroy fair use, demonize all p2p [possibly restricting its use in the future out of fear] suing people as a scare tactic, excessive/un-constitutional fines, DRMed media etc...

They want to prevent ANY filesharing because if we're sharing files (media, audio, etc..) which we created, it reduces our dependence on THEM. By recovering our heritage, that of free artistic expression amongst all individuals, which has been with us all since the dawn of time, we remove the unwanted middleman and their ability to tax the artistic expression - this is their issue with filesharing in general.

The way they have it at the moment, anyone who wants a career in music must do business with them, because they have all the distribution channels under their control - apart from the 'net. (disclaimer: so I hear - IANAM)

Re:Story meaning?

[Dekker3D]

100% chance of rage and fury?

you're right though, and win at summarizing.

Re:Story meaning?

[sopssa]

The study is bogus because it doesn't account for those situations.

And your comment is bogus because almost no one of the WoW players know its filesharing when the updates come. Filesharing by name has become something you share copyrighted files with, aka warez. "Filesharing" has happened since the internet and also http we're developed. Slashdot shares html and image files to you too. But filesharing has got the meaning of sharing warez.

Re:Story meaning?

[turgid]

Given the vast quantity of content, I seriously doubt that very many people go through any sort of hassle to determine what is legit and what is not, which results in virtually everyone obtaining material that is copyrighted, regardless whether they know (or care).

Copyrighted != illegal to download.

The license determines whether it is legal to download/upload/share etc. not the copyright.

Can we please start being more clear about this. Then we can have a sensible discussion.

Re:Story meaning?

[ajlisows]

Reminds me of a company I worked for. They were trying to work with their accounting system and did a bunch of time studies and such. In the end, the President looked at the ACTUAL NUMBERS, decided he didn't like them, and changed them. Which of course, made me wonder why people spent a bunch of time and money doing the studies in the first place.

I think a lot of times, people have an idea of what types of numbers they want and do a study to try to validate them. If they numbers match or exceed their expectations, great! If the numbers don't quite come out how they wanted, they'll push in both directions (their idea of what it should be and the actual) to meet in the middle. If someone asks them about the numbers, they can say a study was done. Slightly "Adjusted" numbers sound a lot better than "I just made this up."

Re:Story meaning?

[ultranova]

I just don't understand the stance that most people on this board seem to take regarding this issue. How can everyone be so supportive of what very obviously amounts to theft? It appears to me that somehow people think it is their "right" to obtain copyrighted material for free. I just don't buy for a second that people who claim to only use file-sharing apps for legitimate purposes only actually do so.

As long as the Mickey Mouse Protection Acts and the resulting infinite copyright and any form of legal protection for any form of DRM remain in place, I shall refuse to acknowledge the validity of any form of copyright law, and will copy/share ("steal") to my heart's content, as well as help others do the same.

Fuck you MAFIAA/Teosto. You should had listened to Lord Macaulay, but you didn't, and now it's far too late. Burn in Hell, where you belong, and cease hindering the progress of the human species.

Just in case you didn't get it yet, FUCK YOU .

Re:Story meaning?

[billybacs]

While I second much of what you stated, the one point I will nitpick from you is that the xxAA are engaged in entrapment. They are a private entity, and specifically because they AREN'T police, they can't entrap. As irresponsible some of their tactics are, they aren't entrapment. It also comes up from time to time that police are entrapping filesharers and other people. It's only entrapment (in the US anyway) if they get you to commit a crime you would not have committed already...like asking you if you want a program so you can get a new album illegally, and then arresting you for it.

Re:Story meaning?

[treeves]

If you can get a sample size of 10 that is representative of a population of 60,000,000 people, you'll have a pretty accurate survey. The reality is, that's not possible in most cases.

No, it's not possible in *any* case. If you have a sample of ten people out of a sixty million, and they are *representative* of the sixty million, you *did not* get them by randomly selecting them. You knew about the population to begin with, and chose your sample very carefully to "represent" it, and it will still be a coarse measure. The sampling standard deviation is proportional to the square root of the sample size. No way around that.

Re:Story meaning?

[roguetrick]

I didn't say it because I preferred my noncommittal attitude in the posts, but why is everyone spelling Grammar as Gramer while making fun of my spelling?

Re:Story meaning?

[Redwing]

They don't understand that you and sopssa thought the original article reminded you of summertime.
In which case, you might include a comma, i.e., "Excellent, summery."

Re:Story meaning?

[4D6963]

Haha, nice one!

Re:Story meaning?

[RivieraKid]

Given the vast quantity of content, I seriously doubt that very many people go through any sort of hassle to determine what is legit and what is not, which results in virtually everyone obtaining material that is copyrighted, regardless whether they know (or care). Given that, I think its a fair guess on their part that yes, most people that claim they are using file-sharing software do so to obtain material illegally.

"do so to obtain material illegally" requires intent. First you say people may obtain unauthorised copies of copyrighted material without knowing, then you accuse them of wilfully seeking out copyrighted material to which they have no authorisation to obtain.

 

Either it's unknowingly and therefore unintentional, or they are specifically seeking it out - it's your choice, but it isn't both, live with it.

 

It matters, because copyright is a civil issue and civil issues generally require intent, rather than something actually occuring.

Re:Story meaning?

[hrimhari]

This is therefore another facet of the issue: I download because I can. Rightholders need to wake up to this fact and adjust their business models accordingly.

But they *are* adjusting! They are suing people like you and finally winning : )

Re:Story meaning?

[Acaeris]

Actually, a large portion of that player base were around when the downloads were available via the 'Official Blizzard Updater' or as direct torrent files for use in uTorrent, etc.

They removed the direct torrent files fairly early but they could still be extracted from the updater and were regularly found on gaming related websites as alternative patch options. Up until a recent update of the launcher the Blizzard Downloader had a details page that pretty much screamed bittorrent. I'd say that at least 20% of the official players who can use the updater know it is updated by bittorrent (which isn't an insignificant number) and I'm pretty certain most of those who play behind a university firewall have been told that the updater uses bittorrent, either by the administrator or by a friend who's suggested sharing the updater over the internal network so as not to annoy the higher ups.

Also, if you ever have a problem updating, one of the first things you get pointed to by a tech support member is a guide to allowing bittorrent traffic through your firewall. If you have problems after a few attempts they just point you to your ISP and tell you they are possibly restricting your connection because of the weight of traffic and it may be worth informing them you are updating the game.

Re:If it's bogus, it's probably too low.

[SoupGuru]

That's what I was thinking. The summary makes it seem that estimating the number that high is outrageous. I certainly wouldn't wager any money that it's significantly higher than actual piracy.

How 1 article became 2

[geekoid]

there is a question that needs answering

Fantastic programme ...

[Alain Williams]

I am a ''loyal'' listener[**], they take current numbers in the news and put them under the microscope. I wish they were part of the main newsrooms - it could result in some really interesting questions being put to some of the politicians who spout numbers without any justification.

[**] You need to be a loyal listener to understand the choice of phrase.

What's the confidence interval?

[Anonymous Coward]

Whenever you estimate a statistic like that, you should also indicate the level of uncertainty surrounding the estimate. Why are they not reporting the upper and lower bounds of the confidence interval surrounding that estimate?

Re:What's the confidence interval?

[caffeinemessiah]

Whenever you estimate a statistic like that, you should also indicate the level of uncertainty surrounding the estimate. Why are they not reporting the upper and lower bounds of the confidence interval surrounding that estimate?

Perhaps because it's hard to come up with confidence intervals when you admit to fudging your own data by bumping the estimate up by almost five percentage points.

Re:What's the confidence interval?

[sexconker]

Because all confidence intervals are equally bullshit.

When you're conclusion is "probably probably within a range of these two probabilities...", and your subject matter involves people, you've done a lot of work and achieved nothing.

The only effort of any use ever in statistics involving people is to increase the sample size. But, you know, that requires actual effort.

Re:What's the confidence interval?

[pilgrim23]

Lies, Damned Lies, Statistics... -Mark Twain

Re:What's the confidence interval?

[ppanon]

Well, the other thing of course is that, since the statistical estimate was ~11%+/-3%, that 5% increase results in possibly over a 50% increase in the estimated number of file sharers in their "worst case" scenario (8%). Basically these people padded the estimates at every possible step instead of carrying through the error factor. Shoddy error analysis that would probably get a bare pass (if not an outright fail) for a first year paper in any scientific discipline.

The only possible interpretation for all those analytical errors from a professional organization is deliberate bias in favour of the organization that commissioned the report. Much more rigorous statistical analysis for things like civilian deaths in Iraq and climate change reports have been dismissed by the same type of politicians who champion this report. The problem is that around 90% of the population may have enough math for their taxes, but have no interest in the mathematical baggage to recognize the bullshit in reports like the BPI's. Heck, you have to explain to many of them that 90% means 9 out of 10. That same segment of the population has no interest in achieving that understanding of statistics either, so they are ripe for believing those manipulative lies.

Re:What's the confidence interval?

[toriver]

Benjamin Disraeli, as quoted by Mark Twain.

Wait, you believed them?

[girlintraining]

They think that a single copy of a song is worth over a hundred thousand dollars too. They claim to lose more in revenue each month than the GDP of most countries. All because of those dyyyeaaarrrn pirates. Enron looks positively boring in comparison to the accounting techniques the recording industry uses. None of this is news. About the only people that buy this crap are judges and legislators -- the rest of us are almost universally of the mindset that a bag of potato chips has more value than most of the recording industry's portfolio.

Re:Wait, you believed them?

[mdwh2]

Indeed, let's look at the maths - supposing each person only shares 24 mp3s. By US standards at least, that's a cost of $1.92 million. So with 7 million file sharers, that's $13.44 trillion.

Now let's check out http://en.wikipedia.org/wiki/List_of_countries_by_GDP_(nominal) - wow, these 7 million people are causing damage to the UK economy equal to almost 5 times the entire GDP of the UK...

Re:Wait, you believed them?

[blueg3]

No they don't. They think an appropriate penalty for being found guilty of sharing a song is on the order of 10-100k (it varies significantly). You do no one any favors by intentionally (or, I suppose, out of ignorance) confusing the concepts of "penalty" and "value".

Meaningless admission

[Trevin]

Using file-sharing software does not equate to sharing files illegally. I admit to using BitTorrent to download Fedora ISO's, and there's nothing illegal about that.

Re:Meaningless admission

[Blakey Rat]

I asked the British government, but unfortunately they told me you don't actually exist. Sorry.

Re:Meaningless admission

[Anonymous Coward]

They'd probably backtrack to only sharing music online... but some big names at the moment (Nine Inch Nails, Radiohead) have legally released their own music as free torrents.

The music industry is STILL stuck in the "downloading"="file sharing"="always illegal" mindset and they show no sign of shifting from that idiotic viewpoint.

Re:Meaningless admission

[Albert Sandberg]

... but, but... you're downloading linux. That means you don't pay for windows. You pirate.

Re:Meaningless admission

[4D6963]

Yeah, sure, because I'm sure there's SOOOO many people who use BitTorrent only to download free linux ISOs and never ever download movies, series, porn, games, books, music.

Translation : the number of people who only do that are insignificant. It would be dishonest or delusional to disagree.

Re:Meaningless admission

[Technician]

For me with ISP traffic shaping, it was either a 24 hour download or FTP from University of Oregon Open Source Lab in 15 minutes. In one I get to run up hosting costs and the other I get slow rates trying to off load a busy server. Bit torrent would be useful if it didn't crawl at 2.5 K bits. Dial up is faster for many users.

Re:Meaningless admission

[skeeto]

World of Warcraft uses BitTorrent to distribute its patches and other content. That's 11.5 million legal file-sharers right there. That's a lot of people.

the story title is kind of lame

[Trepidity]

Some of the estimation steps might be sketchy, but the basic practice of estimating a population proportion from a sample of that population is not particularly questionable. That's how almost all studies of populations work, because taking censuses of all people in a country is rarely feasible. We have century-old statistical theory on how to put bounds on the sampling error, too, assuming the sample was indeed random.

You could have a whole slew of these stories if you really objected to that basic methodology, e.g. nearly every estimate of N million people suffering from a disease or disorder is based on a sample.

Re:the story title is kind of lame

[Abcd1234]

Indeed. In fact, the 1,000 person sample size is a well-established sweet spot for statistical sampling of a population. Of course, in addition to the munging they did to the final results (rounding up to account for self-reporting errors), there may be issues with how the sample was taken (ie, was it truly a random sample of net-connected households). But on the face of it, I agree, I see very little wrong with their methodology.

Re:the story title is kind of lame

[Abreu]

Is it ok to change "11.6%" to "16.3%" based on a "hunch"?

I'm not a statistician, this is an honest question

Re:the story title is kind of lame

[Trepidity]

If there was some previous result that only 2/3 of filesharers admit it when asked, then an upwards revision by 1/3 in an estimate would be defensible. A "hunch" is not quite as good evidence. of course.

I was objecting mainly to the "how 136 people became 7 million" title, which to my ears reads mainly as a criticism of the sample size. But whatever the problems with this estimate, the sample size wasn't really among them.

Re:the story title is kind of lame

[caffeinemessiah]

Is it ok to change "11.6%" to "16.3%" based on a "hunch"? I'm not a statistician, this is an honest question

IAAS, and the answer is no. That goes for the GP as well -- no one is contesting estimation theory, just that the fundamental assumptions are so grossly unmet in this "study" as to render it meaningless. And as someone else already commented, it's dangerous here because it's going to dictate public policy.

If you're going to "adjust" your objective findings, based on some bizarre assumption that a certain percentage of people will lie about file sharing, then why do a survey at all if not to create mathematical/sciency-sounding smoke and mirrors?

Re:the story title is kind of lame

[ghrucla]

>Is it ok to change "11.6%" to "16.3%" based on a "hunch"? The short answer is no, but it's understandable. The 11.6% figure probably is a lowball figure precisely because people tend to underreport illicit or stigmatized behavior and so it was good of the researchers to report this. However they should have just said "this is a lowball figure" because putting a number on it implies more precision than you really have. It's possible they had more than a hunch but even if they were doing some kind of fancy adjustment the adjusted figure should not have been reified by appearing out of context. btw, the several people who have mentioned that inference from n=1000 is just fine are absolutely right. (I'm not a statistician per se, but I am a quant and I teach a graduate statistics course)

Re:the story title is kind of lame

[Erinnys Tisiphone]

Insightful - yes, that would be a valid reason to make that adjustment. "2/3 of filesharers admit it" would be a nearly impossible study, I would think. I agree though, reading the article, the sample size was not an issue to me. It was the arbitrary changing of figures for (what appear to be) unsubstantiated hypotheses that I don't think is valid statistics.

Re:the story title is kind of lame

[mdwh2]

I would hope it's no.

There's actually a clever way to try to account for this kind of thing - you ask them something like "Do you file-share, or is your birthday in January" (or perhaps something even more obscure that the questioner/Government wouldn't know). The point is that people are more willing to admit to it, because people can't know for sure if they really do file-share, or if they answered yes because of the second question.

But when it comes to the population as a whole, because you can estimate the proportion who fall into the second category, you can factor that out, and work out the true value.

But it doesn't look like they did anything like that here.

Re:the story title is kind of lame

[Atario]

some bizarre assumption that a certain percentage of people will lie about file sharing

So it's your contention that no one would lie when asked if they file-share? Really?

Re:the story title is kind of lame

[mysidia]

There may be some validity to it if you were trying to make an overestimate, to establish an upper bound on the number of people (and not a lower bound), and you had some valid basis for saying no more than 4.7% of people lied.

Then your study would find that "At most X people" would be in this category on average, within your confidence interval.

However, you couldn't say "No less than X people", or "X amount of people" are in this category within your confidence interval.

Re:the story title is kind of lame

[DarkOx]

Not but you need some basis if you are going to make such an adjustment. There are ways to determine the rate of sampling error for instance and then use that. In this case that might be to much effort or get you into legally murky waters so what an honest researcher would write something like this:

In my sample of XXXX, YY responded that they sometimes used p2p software in an illegal fashion. Based on this the number of extra legal file sharers in the total population would be ZZZZZZ. I would not expect a person who does not use p2p in an illegal way to respond to my survey in the affirmative while it is easier to image someone who does would respond in the negative; therefor the number may actually be greater than ZZZZZZ.

---
Do so would present the numbers as clearly as they can actually be known; states its assumptions and bias in a consice way.

Re:the story title is kind of lame

[oljanx]

A 4.7% adjustment makes sense. It provides that little extra "oomph" they desired, while not making it look blatently obvious that the survey is biased.

Re:the story title is kind of lame

[Mitreya]

If you're going to "adjust" your objective findings, based on some bizarre assumption that a certain percentage of people will lie about file sharing, then why do a survey at all if not to create mathematical/sciency-sounding smoke and mirrors?

If my memory serves, statisticians have invented a way to work around a "biased" (i.e. embarrassing) question where people are likely to lie. The system is very simple -- you ask the recipient to flip a coin and change their answer depending on the outcome. That allows you to extract the correct population mean (since the error is known to be 50%), but you can't tell what the true answer of the surveyed person is. If anyone in that company had taken a course of statistics (and wanted right answers), they could easily work around the "people lie about file-sharing" problem.

Re:the story title is kind of lame

[caffeinemessiah]

The system is very simple -- you ask the recipient to flip a coin and change their answer depending on the outcome.

You might be referring to the bootstrap or any number of other resampling techniques, and I think you'll find that its often more complicated that turning your dataset into noise.

I would ju like to note

[owyn999]

That 70%* of all statistics are wholly or in part made up

*This number created by taking a sampling of 300 statistical comparisons and then noting that most statistics are based entirely on fiction so multiplying that number by 6

welcome to the world of statistical inference

[Anonymous Coward]

Drawing inferences to the broader population from a sample of about a thousand is a totally accepted and scientific practice. There are innumerable ways you can screw it up (most having to do with the sampling procedure being biased) but in principle there's nothing wrong with saying that 12% of a sample of 1176 implies about 12% of a population. The upwards adjustment to 16% is plausible given that social desirability bias is a known problem with surveys but a bit more of a judgement call and the numbers adjusting for this should never be presented out of context of some pretty serious hedges.

Re:welcome to the world of statistical inference

[sexconker]

No, it's not scientific.

You are sampling a population about something because the causes, effects, or numbers of that something are unknown - you can't calculate your desired data, so you resort to measuring and extrapolation.

Until you can account for all relevant variables, your sample cannot be said to be sound.

If you could account for all (or most) relevant variables, you would be able to calculate, hypothesize, and then measure to confirm or disprove (and then improve and retest). THAT is science.

It's probably still accurate though.

[Hitman_Frost]

It wouldn't surprise me at all if there were seven million illegal file sharers in the UK. It's hardly an underground thing any more is it?

Re:It's probably still accurate though.

[gbjbaanb]

When you know the total population of the UK is roughly 30 million households, that's a fair chunk of the population. (total population is roughly 60 million people)

Out of the total population, only 18.7 million have broadband. Guess roughly 40% of the population is a pirate then. We should make it legal, government being there for the populace and all that.

This calls for a quote

[SilverHatHacker]

"If they facts don't fit the theory, change the facts."
~Albert Einstein

Idiot (in-Numerate) Government Workers

[omb]

In the 70's Shirly Williams destroyed the British educational system based on political bias, almost everything she did was wrong and did not work. 30+ years later, most of the Civil Service does not understand statistics or any numerate discipline, these idiots now advise the pols!

This stuff does not have ANY substance and certainly dosn't deserve the description Research. It is ignorant crap.

Why are we surprised that nothing that New, or Old Labor does works?

Re:Idiot (in-Numerate) Government Workers

[FourthAge]

To modern Government, statistics are a tool that is used to define reality, rather than understand it.

You're more likely to agree with something if you think 90% of the population are backing it. This is the purpose of opinion polls and focus groups; not to find out what people think, but to find out how they can be told what to think.

The decisions come first. Then, the statistics follow to show that the decisions were right. It is cunningly manipulative. Even those familiar with the old "damn lies and statistics" quote still tend to believe that "scientific" "statistical" data cannot be twisted to serve a political agenda.

If the Government wants you to believe something, that in itself is a good reason to disbelieve it.

mathematics

[martas]

maybe the authors of the study were taught math skills through unschooling?

file sharing software=pirate???

[Anonymous Coward]

using file sharing software does not mean you pirate software or media.....

Re:file sharing software=pirate???

[sexconker]

No, but I'd be willing to bet my genitals that it's true 99.9% of the time.

How's that for a confidence interval?

Re:file sharing software=pirate???

[toriver]

Also: What counts as file sharing software? Do they have a list that includes Opera 9 and 10 (which includes Bittorrent support) or Blizzard Downloader (which downloads patches to World of Warcraft using peer-to-peer tech? Heck, if you right-click a folder in Windows and elect to share it on your network, is that "file sharing"?

Re:file sharing software=pirate???

[4D6963]

Yeah, thanks for the pedantic comment. That's like saying, it's not because you have a TV that you ever watch it, when in practice it's completely irrelevant because it's so rare it's insignificant.

Actually I wouldn't be surprised if there were more people with TVs who never watched than file sharers who only ever download copyright-free material.

Re:file sharing software=pirate???

[JAlexoi]

That is true. I use P2P to download World of Warcraft patches, the Blizzard patch downloader uses it actualy. So I use P2P, but absolutely legally.

So, optimistically, 2.12 million, then?

[Animaether]

136 out of 1176 people in households with internet connections admitted to having used file-sharing software (source: the summary)
18.3 million households in the UK had internet access at time of polling in 2009 (source: http://www.statistics.gov.uk/CCI/nugget.asp?ID=8 )

136/1176 * 18.3M ~= 2.12M

Not sure if "having used file-sharing software" means that they downloaded / distributed at least 1 item - say, a song - via said software and that they had no actual rights to do so (you know, as most people use file-sharing software to distribute Linux distros, or have simply 'used it' but didn't actually download or upload anything... *cough*)...

But let's presume it does.

Then let's take the low price in iTunes UK of GBP 0.79 per song, then the music industry 'lost' ('cos obviously people had no intention of buying that song that they didn't download / distribute because they were downloading a Linux distro instead *cough*) about GBP 1,671,897.96.

Well, that's peanuts, innit.

Re:So, optimistically, 2.12 million, then?

[DragonTHC]

right, then how did they adjust dishonesty?

They just adjusted from 11.3% to 16.3%

Where did they get this 5% of people lie on surveys figure?

so then, if you take 16.3% and plug it in...

192/1176 * 18.3M ~=3M

that's still far less than 7 million people. The ath don't madd up! ;p

the true "what the fuck"

[timmarhy]

TTWTF here is that someone believes there are only 7 million file sharers in the UK. sure that figure is bullshit, but the number should be a hell of a lot higher not lower.

Re:the true "what the fuck"

[Volante3192]

How much higher?

Keep in mind, before you answer, note the UK only has a population of 60 mil, so try to pick a number below that.

Re:the true "what the fuck"

[sakdoctor]

As a solipsist I'd say everyone does it.

Why the BBC rocks

[MosesJones]

This is yet another example as to why the BBC is the finest broadcasting and journalistic organisation on the planet (I've never worked for them, sold to them or have any other financial connection other than the license fee).

They actually investigated something created by an industry group and found it to be bollocks and then reported it. The BBC are arguably the most "socialist" organisation in the democratic world (funded by a tax on everyone for the benefit of everyone) and yet they still question and challenge everything.

The US seriously needs something that questions vested interests and rubbish statistics as much as the BBC. Jon Stewart and Bill Maher are just comedians and FoxNews is just comedy.

Given a choice between the first amendment and the BBC, I'll take the BBC; its demonstrated more freedom of speech in a week than the US media has in a decade.

Re:Why the BBC rocks

[vivaelamor]

Oh come on, the BBC have reported this number many times since it was first used and you sing their praises because Radio 4 happens to do a show devoted to statistics? I wonder just how much time they will devote to debunking this statistic considering how many times they have quoted it.

Just because the BBC is better than the US networks doesn't mean we should be proud, personally I'm appalled at how low the bar is set.

Re:Why the BBC rocks

[Atario]

funded by a tax on everyone for the benefit of everyone

Is this, strictly speaking, true? I thought you had to pay a yearly television license fee based on the number of TVs you own. No TV, no fee. (Which creates problems of its own, including the necessity of sending around TV detector vans to make sure no one's hiding an unlicensed TV. I'm thinking it would in fact be better if it were a simple universal fee and be done with it...)

Oh, and agreed about the Beeb rockin'.

Re:Why the BBC rocks

[FourthAge]

You can also avoid paying the licence fee if your TV can't receive over-the-air pictures, e.g. if it is disconnected from the aerial.

There was once a "radio licence", you can still see a reference to it in one episode of Monty Python, but this was phased out when almost nobody owned a radio but not a TV.

In the future, I expect the TV licence will be extended to include Internet connections as well, since those can now be used to receive BBC programmes too. At that point, we will see if the BBC can continue to convince people that it is worth the money.

Re:Why the BBC rocks

[_Shad0w_]

You only need one license, you can have as many tellies as you like. Portable tellies used in caravans and the like will be covered by the license for your home as well.

If you have two houses, you will need two licenses though, afaicr - which is why students away at Uni need to buy a license - including if they're in halls - even though their permanent residence might still be their parent's house.

I find the BBC great value and love it dearly. I suspect people will say that's because I'm white, middle class and liberal or something.

Re:Why the BBC rocks

[houghi]

The BBC are arguably the most "socialist" organisation in the democratic world [...] and yet they still question and challenge everything.

What does their political thinking have to do with whether they challenge anything or not? I would call many unions more socialist then the BBC. And they do challenge everything all the time as well (Rightfully or not is another discussion).

I think you confused "socialist" with "socially engaged" which is not the same thing.

Re:Why the BBC rocks

[evilviper]

This is yet another example as to why the BBC is the finest broadcasting and journalistic organisation on the planet

"The grass is always greener" as the saying goes, so I'd naturally love to believe that. But sadly, I can't, because I have extensive experience watching, listening, and reading the BBC.

The BBC's news reports are almost always moderately-shallow fluff, VERY light on facts relative to their US counterparts, and rarely researched more than summarily, and constantly providing unconfirmed 3rd party information without so much as a footnote.

The levels of journalistic integrity displayed by the BBC would become a scandal at any (real*) major US news provider, either print or the major 3 TV networks.

The US seriously needs something that questions vested interests and rubbish statistics as much as the BBC. Jon Stewart and Bill Maher are just comedians and FoxNews is just comedy.

None of the above are serious US news sources. Try comparing the BBC to the New York Times, or to the morning/nightly news broadcasts by the major 3 US TV networks (NBC, ABC, CBS). For more in-depth issues, try comparing the BBC to Frontline, 60 Minutes, etc., and then come back and attempt to justify your US-bashing... The BBC does a reasonable job, but they can

No, the "local news" that's on several hours a day isn't up to par with the BBC, but the two have completely different purposes and scope, so they're hardly comparable. Nor the "early shows" which resemble talk shows vastly more than a serious attempt at news reporting.

And no, the existence of all the crappy, non-news sources that you (or I) can (and have) point to don't detract from the fact that there are many extremely GOOD news sources in the US. If you want to go that route, the UK is in an even sorrier state... Even the most dedicated tabloid readers in the US would be aghast at the tabloids on the UK news stands. Never-mind the heavy-handed, unbelievably biased pieces of trash which get passed off as documentaries (even on BBC TV/Radio, though not the worst of it).

Re:Why the BBC rocks

[nosferatu1001]

It already does.

If you receive ANY realtime television transmission, regardless of the method used to receive it, you are required to obtain a license.

This covers internet, satellite tv, etc. Only if you exclusively watch non-live programming are you exempt. Such as the iplayer....

Re:Why the BBC rocks

[nosferatu1001]

You only need a licence for your own television if each bedroom is considered a seperate domicile - in other words it has a lock on a door which makes it "yours", wiht a seperate rental agreement. This is why Halls require individual licences, but shared houses don't (as a general rule)

Re:Why the BBC rocks

[gilgongo]

I take it you have not watched the movie of Manufacturing Consent?. You should - I think you'd like it.

Re:Why the BBC rocks

[_Shad0w_]

That's pretty much what I was saying. Just not very well, it seems.

Re:Why the BBC rocks

[FourthAge]

I think many Brits would say that Chomsky's thesis doesn't apply to their broadcaster, only the "evil" ones that are owned by corporations instead of Governments, and are "for-profit" instead of funded by the taxpayer.

Why, it's almost as if the vast propaganda power of a national broadcaster has been used to guide their thinking. Surely not... the BBC is good and would never do such a thing.

Re:Why the BBC rocks

[raylu]

Though I doubt you did it purposely, it was the section you snipped that makes them "socialist":

funded by a tax on everyone for the benefit of everyone

Re:Why the BBC rocks

[vivaelamor]

I have great respect for most people working at the BBC and many of the things it has done. Having said that, I wish to specially exclude Nicky Campbell from that last statement entirely.

I believe the best way of describing the issue about news reporting is to call it 'grading on a bell curve'.

My only real issue with the BBC is how they are funded. Why? Because that issue makes all the of the others moot.

Winston Churchill

[feufeu]

O "Statistics are like a drunk with a lampost: used more for support than illumination."
O "The only statistics you can trust are those you falsified yourself."
Tick one.

Re:Winston Churchill

[El_Oscuro]

I have read a number of studies which say: 87% of all statistics are completely made up.

i think it's...

[WalesAlex]

time to get 1984 as obligatory reading before anyone gets a say in the subject

Bad Demographics

[ZuchinniOne]

I have a friend in England who was telling me that they actually focused on trying to make younger people the demographic for this study ... and obviously younger people are more likely to fileshare.

I wish this was more than an unsubstantiated rumor, but I couldn't find anything online about it.

Re:If it's bogus, it's probably too low.

[Miamicanes]

> Work backwards from the undisputed declining sales figures of the recording industry.

The main reason for declining sales is the fact that CD sales during the 90s were artificially boosted by people replacing records and tapes with CDs... then replacing them again when remastered CDs were released a few years later. It was a once-in-a-lifetime event for the recording industry that won't be repeated during our lifetimes.

People re-bought CDs they already owned in analog (or optimized-for-analog CDs) because they represented an epic improvement in quality by just about any meaningful standard over the analog media they replaced. Everything that's come out since CDs has only been cheaper, shittier-sounding, or intolerably-crippled by DRM.

Here's an idea for the music industry: ditch the DRM'ed formats, and roll out a music format on DVD media with 96KHz 32-bit stereo PCM. Make the discs gold-colored, call it something like "X-fi", and sell them for $24.95. You'll win on all counts -- genX'ers will go back into highschool mode and buy them to show off how rich they are and/or pretend they sound sufficiently better than 16-bit CDs to justify spending ~twice as much on them, and the fact that every disc will be ~4-8 gigabytes will serve as self-limiting DRM for the next decade or so. Just make sure they still have the MOST compelling consumer benefit intact (and reason why people who buy CDs still DO buy CDs): it's a flawless first-generation master to use for making all your "working" copies for everywhere else.

Re:If it's bogus, it's probably too low.

[mysidia]

It could be that file sharers are slightly more likely to have excluded themselves from or refused to participate in the study.

Re:If it's bogus, it's probably too low.

[aynoknman]

I certainly wouldn't wager any money that it's significantly higher than actual piracy.

Why not?

Given the difficulty determining an reliable answer, you could probably wager any money at all and not have to pay up.

Have fun trying to collect on your bet.

The really big hunch

[CustomDesigned]

is assuming file sharing == illegal file sharing. As others have pointed out, that alone makes the rest of the conclusions meaningless. My stat class said sample size should aim to be around sqrt of population. Of course, smaller samples just lower the confidence intervals. The 1000 people (sqrt(10^6)) is at least in the ball park for a population of 30 million.

Wow...

[Mashiki]

Can I hire these guys to make me money? Seriously, I've never seen number fudging like this.

So, to sum up

[Hawthorne01]

This survey "saved or created" seven million file sharers.

Scoundrel Statistics

[anyaristow]

Even a first year stats student can see it.

This is almost as cliche in arguments of statistics as the car analogy is on slashdot, and it's the sign of a scoundrel. If you actually had a first year stat student's understanding of stats you'd know where the weaknesses actually are, and where all the rest of the smoke blown in this discussion goes laghably wrong.

So let's apply some first year stats to the issue.

First, the sample size. Whether it is numerically large enough to be useful is a matter not only of it's size but also the number of positive results. IOW, a sample size of 1176 is too small if you found 3 of what you're looking for, but if you found 136 (11.6% of 1176), you have plenty of samples. The question is then only whether you had a representative sample.

My next concern would be precision. Using data with three or four significant digits (136, 1176) to make conclusions to seven significant digits (11.56463%) is silly, but that doesn't seem to have happened here. The only number in all of this that is fishy is the 16.3% number. To get three significant digits they'd have to know the number of lying households to that precision. If they had another study that determined this number they might very well have a number to that precision, but I'm assuming they just guessed.

That's still not a problem. If you guess, you run your confidence interval through your formulae (here it's a simple product) to put a range on your results. If it's a from-your-ass guess you might put a 100% failure estimate on your low end (i.e. there might be no lying households at all) to arrive at a conservative range. Here, it looks like they used an estimate of 40%. They should have (and might have; I didn't RTFA) run the un-adjusted 11.6% through the formulae to get a conservative low-end range.

Anyway, the number they finally used was 7%. One significant digit. That doesn't imply the same precision as, say, 6.7% would. In fact, if their figure for the number of lying households really was accurate to one digit (i.e. 35-45%) then rounding their final result to one digit was the correct procedure. If it was just a guess they should have run the absolute low estimate (probably, zero lying households) through to get a range.

So, with actual first year stat knowledge it's possible to actually state what might be wrong with the study, and not resort to "any first year stat student" hand-waving. It's clear that the most-cited criticism (the sample size) is the result of ignorance and group think, not actual knowledge of statistics.

Re:Scoundrel Statistics

[HiThere]

Your criticisms are largely valid, but I still think the sample size was too small. After all, they couldn't know before they did the study what percentage would answer what way ... not unless the study was rigged.

Of course, it also depends on what the purpose is. If it were for marketing, then this might be a quite acceptable procedure. In that case a large amount of error wouldn't cause significant problems to anyone. But if it's being used to lobby for laws, then it's just that it won't cause any problems for *them*. That the results have been adjusted to be something that can be released to massage public opinion. Etc. In such a case I have a much higher bar for study requirements, and it requires that either the population tested be standardized to eliminate bias (which is impossible if you don't know where bias is coming from already) or it needs to be a MUCH large random sample.

A good study in this area would first investigate the characteristics of a large population WRT standardizing their likelihood of file-sharing. This step in itself would involve many thousands of people in many different social, economic, and geographic strata. (You might want to steer clear of race or national origin. It's likely significant, but too touchy.) After you've done that, then you can standardize a random sample for study WRT characteristics associated with file-sharing. And at THAT point you might be able to establish reasonable guesses at error bars.

Re:Scoundrel Statistics

[Bigjeff5]

Survey sizes of around 1000 are pretty standard. If you run the survey and get 3 positives out of 1000, you say "Oh shit, sample size is too small", then run the same survey with 5,000 or 10,000 people to catch a larger number people you are targeting - i.e. we're looking to see what percentage of people practice illegal file sharing, we need to find at least a decent number of illegal file sharers so we know our survey is accurate.

It's not a matter of knowing what you'll get before hand or rigging the study, you have to have a jumping off point somewhere, else you'll never do the study. If you get untrustworthy results, you simply adjust your sample size and conduct the survey again.

A good study in this area would first investigate the characteristics of a large population WRT standardizing their likelihood of file-sharing.

That is completely unnecessary if all you want to know is what percentage of people practice illegal file sharing. And I'm not sure what you mean by "standardizing" their likelihood of file sharing. Huh?

This step in itself would involve many thousands of people in many different social, economic, and geographic strata.

The government does this thing called a Census every few years, that collects just such information from close to 100% of the population, making it extremely reliable.

(You might want to steer clear of race or national origin. It's likely significant, but too touchy.)

Why? If it can affect your outcome, it should be in your demographics, otherwise your study is unreliable. Why the hell is that information "touchy"? Does a white guy not relize he's white? Or did someone forget to tell the Irishwoman she's from Ireland? What the hell?

Re:Scoundrel Statistics

[dman123]

My next concern would be precision. Using data with three or four significant digits (136, 1176) to make conclusions to seven significant digits (11.56463%) is silly, but that doesn't seem to have happened here. The only number in all of this that is fishy is the 16.3% number. To get three significant digits they'd have to know the number of lying households to that precision. If they had another study that determined this number they might very well have a number to that precision, but I'm assuming they just guessed.

It wasn't that precise. The original number was 17.0% and the article poster just converted it from metric percentages so Americans wouldn't get confused.

Re:Scoundrel Statistics

[MeNeXT]

Nice argument on sample size but was it representative? Ignorance? Was 1,176 a sample large enough to represent the the 40,000,000? I would assume not. You could assume so. The fact would still be that we would both be assuming.

On another note, why adjust up? Why not down? How was downloading explained? Did people download from iTunes, Amazon or TPB? Did people already own the download?

Re:Scoundrel Statistics

[anarchyboy]

Your criticisms are largely valid, but I still think the sample size was too small. After all, they couldn't know before they did the study what percentage would answer what way ... not unless the study was rigged.

By that logic it is impossible to conduct a fair study, either you know roughly what percentage you expect to answer that way and your study is rigged or you don't know what percentage is likely to give that answer so you conduct your study but are told your sample size is too small because you didn't know what percentage would answer that way to begin with even though that was the entire point of your study anyway.

Re:Scoundrel Statistics

[jmerlin]

"The question is then only whether you had a representative sample."

Herein lies the problem. If the users are pulled "at random" from a list of the top 20% of bandwidth consumers in the country, you can damn well bet there is a bias. If it's pulled "at random" from the entire population, this sample size is clearly too small. If it's pulled "at random" from subscribers to a local ISP, the numbers will be innately inaccurate as an ISP cannot determine what a user is trafficking if that user isn't completely stupid. Etc etc etc. It would be unbelievably difficult to pick a true random sample for this study and have it mean anything in an unbiased manner. You can easily plop a list of 5000 people down on a desk who consumed tons of bandwidth and be sure they're either uploading, downloading, streaming servers, filesharing, etc. Then just pick 1176 people and be reasonably certain a large percentage of them (11%!) will be visibly using P2P networks, though nothing is said to the content of those transfers (!!!!!!). Your statistics mean nothing if the sample chosen isn't completely fair, and I honestly don't think anyone here on slashdot can agree that such a sample can be taken in this setting. This study is innately biased, as the method of "choosing" samples will clearly be the one that most dramatically emphasizes the point that is to be made. After all, that is the point of using statistical sciences to argue a point.

Running a few stat's

[Bork]

Pulled out my stat's book and did some rough calculations in estimating Proportion.
Used a population size of 58,000,000
Their point estimate of 0.116
Z(0.025) = 1.96 for a 95% confidence level (2 SD)

With a sample size of 1176, I get a error of plus/minus 1,102,000.
There would need to a sample size of 10,000 to get the error of plus/minus 570,000.

Re:If it's bogus, it's probably too low.

[dAzED1]

I haven't bought a single CD since the RIAA began to wage its war on its customers. Lars will forever be invited to rot in hell, so far as I'm concerned.

Nor, however, have I pirated, copied, "borrowed," stolen, or in any other way illegally obtained copies of music since then.

That music has declined in quality for a while, and that people don't need to replace their record and tape collections, may have more to do with those declining numbers. The market saturated, went in to a slump, and during that period the RIAA screwed themselves.

BTW - the fact that I can record music using relatively cheap hardware these days helps a lot too; we don't really need the mega-shops anymore when people can get together and make their own stuff really cheaply. There is no natural monopoly anymore. The RIAA can't claim a stranglehold.

Practically tax, but with an important difference

[Chuck Chunder]

For all intents and purposes it is a tax, however the way it is raised is important.

Here in Australia we have a similar institution. However rather than being funded by a specific fee it's income is provided from general taxation revenue.

This can pose a threat to it's independence by making it more beholden to the government of the day for it's revenue stream.

it's teh maths...

[Sfing_ter]

on teh internests its teh maths what duz it.
How did Todd Snider put it? The point being - lies - damn lies and statistics.

Lies

[Workaphobia]

I'm really not sure if this is a damned lie or a statistic.

People and statistics

[symbolset]

For measuring ball bearings produced by a machine, statistics make some good sense. After all, if the average size of a bearing measures 3.05mm, +/- .1mm, a curve can be made that measures meaningfully if a specific batch of bearings differs from the mean and so detect when a machine is coming out of tolerance, long before the bearings produced become unsatisfactory.

When doing studies such as this one, they become less meaningful. If, for example, the report is to be meaningful, the understanding of the questions by the respondent, and their applicability to the statement of the report must be understood. Since to disclose the text of the report to the respondent prior to their response would spoil the result, their understanding of the context of the question is inevitably suspect.

And then there's the question of simply understanding the question in the general sense. In an environment where three standard deviations of the question "are you on fire?" nets 80% of a sizeable sample, one must presume that analysis of more subtle questions will not yield meaningful results.

Re:If it's bogus, it's probably too low.

[zippthorne]

650 MB is more than Enough space for "an album"'s worth of music. If you need more quality than a CD offers, consider that you could compress down to 650 from some arbitrarily perfect "master" and do significantly better than the uncompressed CD standard.

1GB flash drives are almost trivially cheap now, and I'm sure mask ROMs could be even cheaper, if anyone bothered to make them in that size. So there's really no reason to expect optical media will continue for very long in the music world. Basically: as soon as they realize they can put a crypto chip on the same IC as the memory controller and have "perfect" DRM and a device that people can jog with, they're going to work hard to phase out optical media.

Enough about sample size

[anyaristow]

Was 1,176 a sample large enough to represent the the 40,000,000? I would assume not. You could assume so. The fact would still be that we would both be assuming.

Assume nothing. Google is your friend.

Google: sample size

First result has all you need.

Figures lie...

[hyades1]

And liars figure.

The best way to shut these slime-oids up would be to conduct a forensic audit of their royalty payments to artists. I bet not one of the companies would come out clean.

Here comes the pleudo-science...

[WoollyMittens]

Real statistics come with a margin of error. I bet in this case the error is infinite.

Re:If it's bogus, it's probably too low.

[bipbop]

Hey, don't forget about non-RIAA artists. Refusing to do business with the RIAA doesn't mean giving up music.

Slight of hand..

[dbcad7]

As is usual (been going on forever).. the whole issue of file sharing and what's illegal and what's not, is clear as mud. Here you have a survey that is asking if someone has used file sharing software.. and then claims that by extrapolating numbers that there are a huge numbers if file sharers.. ok so far.. but then it makes the leap and says that these are "illegal" file sharers.. It would have been more informative to ask those that answered the question as users of the software, do you make songs or movies available for download, or do you just download ? .. these have different legal positions (although maybe not on the UK ?)

Re:If it's bogus, it's probably too low.

[gilgongo]

The main reason for declining sales is the fact that CD sales during the 90s were artificially boosted by people replacing records and tapes with CDs... then replacing them again when remastered CDs were released a few years later. It was a once-in-a-lifetime event for the recording industry that won't be repeated during our lifetimes.

I used to think this but lately I'm not so sure. My feeling is that since the 1950's and the creation of the music recording industry, we have seen an immense amount of music being produced for (or in the hope of) profit. Before then, when people played and listened to music for entertainment, the amount of music produced was far lower becuase there was no industrialised famework to channel funds, cultural cachet and other benefits to both musicians, and more importantly, their record labels distributors and publishers. Now, however we are seeing the return to the ancient ways: rip mix burn, people being able to make and listen to music without any involvement from anyone else. But the huge difference between now and the pre-1950's situaiton is that we have been led to believe that ALL music is wonderful, that the act of creating it deserves remuneration by right, and that it is somehow a sacred commodity. That, however, is bullshit. We are now able to see that bullshit and will come to realise that a) there is FAR too much production of music (indeed arts of all kinds) and the b) most should and will not see the light of day if left to the "free market" of listeners.

Lies, Damned Lies, and Statistics.....

[IHC Navistar]

Could someone please tag this: 'liesdamnedliesandstatistics' please?

Much thanks!

You answered your own question...

[Joce640k]

"Why would the solution to something that is not easily enforceable be to make it legal? "

You answered it yourself - it's not easily enforceable.

What the RIAA/MPAA is trying to do is to get the governments and police forces of the world to enforce something which can't be enforced. The amount of money which could be sunk in this black hole if they achieve it is unthinkable.

The real problem is that the RIAA has spent the last ten years with their hands over their ears going "LALALALALALALA, we're not listening". Listening to customers is usually seen as good business practice, but they're not doing it.

The world has changed, people don't listen to CDs any more, they listen to mp3, and they want the singles, not a CD with one decent track and a load of filler.

Apple listened and their iTunes business is doing very well thank you very much.

The other elephant in the 'enforcement' room is that DVD sales are booming year on year almost in line with the drop in CD sales. Maybe the public is buying DVDs instead of CDs...? Nah, it couldn't possibly be market forces at work. We'd better spend billions of tax $$$ on law enforcement to protect the buggy-whip makers, just in case...

Re:If it's bogus, it's probably too low.

[MROD]

They *DID* try the DVD format and selling it at that price. It was called DVD-Audio. It flopped.

The reason?

CD is "good enough", just like MP3 is often "good enough" to the majority of people.

Long live mediocrity.

File Sharing IS NOT illegal

[Raven737]

by itself. If you share content that is copyrighted in your country, to which you do not have appropriate distribution rights for that country, then you MIGHT be doing something illegal.

When i download my favorite Linux distro via BitTorrent i am NOT doing something illegal. Having used a File Sharing program does not mean i have ever used it for any illegal purpose.
136 people admitted using file sharing software, they did NOT admit to using it for illegal purposes!

Why have we allowed the assumption that File Sharing = Illegal to become so commonplace?
I am outraged this is not the first point of contention for this study.

Obligatory

[webheaded]

http://xkcd.com/605/

Seriously. No one posted this?

Most File-sharing are not done with softwares

[R3ADER]

I believe that I am correct by suggesting that most file-sharing are not usually done with file-sharing software. Files are mostly being shared on warez forums and a few blogs. File sharing software's really don't make up a huge portion of the file-sharing that really goes on.

mhmm

[Dremth]

This just goes to show that 87% of all statistics are completely made up.

bogus CD replacement theory

[Gary W. Longsine]

This theory gets trotted out at Slashdot fairly often in these discussions. It's entirely bogus. Vinyl LPs that people actually listened too wore out or were scratched much more easily than CDs, and had a much higher general replacement rate. Furthermore, cassette tapes and 8-track tapes also wore out much more easily, and were portable. People often bought the same album in two different formats. So, with as much evidence as you have, I'll claim that the advent of the CD actually reduced the rate of re-purchase of the same album.

Re:bogus CD replacement theory

[Gary W. Longsine]

s/listened too/listened to/

True!

[krischik]

If I had mod point I mod you up.

WoW uses BitTorrent

[krischik]

You are aware that World of Warcraft uses BitTorrent to distribute the updates? So I welcome you to your new world a castrate.

A rare case - you can listen to the FA

[Larry Lightbulb]

The More Or Less home page is http://www.bbc.co.uk/programmes/b006qshd, and it links to the current and previous programmes.

Re:If it's bogus, it's probably too low.

[thatjavaguy]

People will not buy them again in an even higher quality. Barely anyone cares about quality anymore - they listen to compressed music on a poor format (mp3) on a portable device using low-fi earbuds.

If quality is such a big argument - why isn't blu-ray doing better than it is?

Re:If it's bogus, it's probably too low.

[dAzED1]

sorry, I wish I could say I was unclear, but no - I mispoke. Or rather, spoke incompletely, and the later "clarification" didn't fix it. "I haven't bought XYZ from the RIAA since blah..." is what it should have said. I hinted at that later in the comment, but yeah - wasn't clear.

I'm not an "indy" guy for the cutesy reasons most people are, and I have a hard time coming across indy stuff I actually like (since for many of them, the reason they're not RIAA is because the RIAA doesn't want them). But there are those who are indy who are good, and some who are even big.

It allows me to still like Prince - who is and was willing to give away music that he chose to give away, but will seek to protect that which he has not chosen to give away - while at the same time still being pissed off at Lars. I'm an Open Society advocate, but I also strongly feel that it should be done out of respect and choice, not out of theft or piracy. I'm not advocating individual instances of Open Society, but instead an entire cultural shift, which has to be a willing choice.