How The UK Gov't Extrapolated 136 Self-Reported File Sharers Into 7 Million
from the nice-work dept
A year ago, Julian Sanchez took the time to dig into the numbers that the US gov’t was using to explain the “cost” of “piracy” on the economy and found that they were completely bogus, based on an offhand mention decades ago, based on no research, and then twisted, pumped up and given a government “seal of approval.” It looks like something similar has happened in the UK. Joseph Young points out to an investigation into the UK’s oft-repeated number of 7 million illegal file sharers as being the reason for kicking people off the internet. But a dive into the details finds a massive extrapolation. That 7 million number is based on 136 people responding to a survey paid for by music industry lobbying group BPI, and conducted by Jupiter Research (now owned by Forrester). Not surprisingly, the research director who ran that study was also a guy who has claimed in the past that music can’t be free and that without copyright there’s no way for musicians to get paid.
Think those survey questions were unbiased?
Either way, the survey reached 1,176 net-connected households, of which 136 (11.6%) said they did file sharing. The researchers then just decided that 11.6% was too low, and bumped it up to 16.3%. Why? “To reflect the assumption that fewer people admit to file sharing than actually do it.” Fine. But how was the number picked? They won’t say. So… you’ve basically added 40% to the number there for totally unclear reasons. Then… the extrapolation gets more ridiculous. Jupiter Research said that there were 40 million people online in the UK, and they applied their made up 16.3% number to that number. Only problem? The 40 million number is made up to. The real number was 33.9 million. Thus, their “estimate” was first boosted by about 40% and then another 20%. And all that on top of likely leading questions to appease the music industry lobbyists paying for the study. And this is what the UK gov’t is basing its decisions on?
Filed Under: extrapolation, file sharing, uk
Comments on “How The UK Gov't Extrapolated 136 Self-Reported File Sharers Into 7 Million”
Shocked
I’m shocked! SHOCKED!!!
… well, not that shocked.
Re: Shocked
Survey Taker: “Has your government ever deceived you?”
Surveyee: “No, I knew they were lying to whole time.”
Re: Re: Shocked
Survey Taker: “The box only has a spot for yes or no what do your want me to check?”
Re: Shocked
here is some more shocking news:
1) water is wet.
2) the sky is blue.
3) size matters.
Re: Re: Shocked
Your right size does matter. And the industry has a huge incentive… er pocketbook. That they put to good use. Why get real results when you can buy the ones that suit you?
Own Goal
Actually inflated figures are a bit of an own goal – don’t they indicate that filesharing is actually unstoppable without prosecuting more people than voted for the government at the last election….
You would have thought that a lower figure would have indicated that the proposed measures had some chance of success….
Rounding
So basically they (based on the survey) tripled the actual amount in their estimate. (Yeah I know I did a little rounding myself 11.6% of 33.9M = 3.9M
What definition of file sharing are they using?
Did they ask file sharing or music sharing or being a music file leech?
I would be most worried about the number of people sharing. IE: giving away the the product you are trying to sell.
And this is what the UK gov’t is basing its decisions on?
No, don’t be silly. They’re basing it on who takes them out to dinner at the nicest restraunts.
Statistics...
Statistics has rarely used to tell the truth. Its used to make a point and truth rarely enters into it.
If it was not for those damn ‘file sharing’ kids th entire world could have avoided the recent recession. Just think of the BILLIONS of lost sales in Music, Movies and Software. The sky is falling and my retirement too.
does the research talk about file sharers or illigal file sharers if I send my vacation pictures to a friend I’m sharing files but not harming the music industry.
Re: Re:
I really wanted to explain to you how sending vacation pictures is destroying the music industry. Unfortunately I’m not delusional enough to come up with a reason. Just wait for Dark Helmet, I’m sure he’ll have a reason.
Everyone uses statistics to prove their point.
“Either way, the survey reached 1,176 net-connected households, of which 136 (11.6%) said they did file sharing. The researchers then just decided that 11.6% was too low, and bumped it up to 16.3%. Why? “To reflect the assumption that fewer people admit to file sharing than actually do it.” Fine. But how was the number picked? They won’t say. So… you’ve basically added 40% to the number there for totally unclear reasons.”
I love statistics. I can make it sound really impacting “added 40%”, or minimal (4.7 percentage points). Both are 100% accurate. It’s always funny how people usually choose the route that makes their case sound stronger.
I do agree that the use of any arbitrary number (and this one does sound arbitrary) in a survey/poll pretty much negates the validity of said poll, and I would throw out any data/results derived from it.
Re: Everyone uses statistics to prove their point.
I love statistics. I can make it sound really impacting “added 40%”, or minimal (4.7 percentage points). Both are 100% accurate.
Yet the correct way is only one: stating that they added 40%, since the topic is number of “file sharers” and the number being inflated.
Re: Everyone uses statistics to prove their point.
If there were 40 million people online, then 16.3% is 6.5 million people. If there were 33.9 million people online, and the surveyed number of 11.6% is used, then that is 3.9 million people.
So we show that out of the extrapolated number of 7 million people who share files, 3.1 million were invented by the surveyors without any reason.
So the argument is not about how statistics are used or misused, the argument is that these surveyors are liars.
–GJ–
Re: Everyone uses statistics to prove their point.
Math fail!
Adding “4.7 percentage points” may be correct, but it is completely meaningless. Adding 4.7% to .00000001% vs. adding it to 95.3% completely changes how much it effects the results. Mike correctly stated how the final results was inflated by showing how it was changed, you failed.
Re: Re: Everyone uses statistics to prove their point.
I never said Mike was incorrect, just pointing out how stats are used. The 4.7% increase would be based on total internet users (not just file-sharers) so it’s an accurate statement.
If you had 1 File sharer out of 100, and that number increased to 2, the number of file sharers doubled (a whopping 100% increase). Can you see the difference in connotation? We saw an increase of 100% in file-sharers, or we saw an additional 1% of all internet users share files.
yup
This is what happens when politics and research meet.
My stats better
100% of the record lables would rape you in your sleep of they could.
100% of record lables would kill you to earn $5 if they could.
Job Done
As a comment in the original article notes that means that just by unfudging the figures file sharing has dropped from 5.6m to 3.9m a 30% drop.
That’s an impressive reduction and all that is required to achieve it fully is a practically zero spend, (OK Mandy might have to fork a few quid to get to a press briefing or two) and for our glorious, unelected, twice expelled current First Secretary of State, Secretary of State for Business, Innovation and Skills to tell the truth.
A 30% drop in a few minutes for free, requiring no legislation and no pissing off voters – thats got to be tempting?…
Unfortunatly it would also mean Mandy telling the truth – see the rub?
Alternatively he could just resign in disgrace, he’s good at that.
Error
OK Stat lovers.. Here’s how it works.. There was a reason for their so-called perceived madness..
Accuracy is the difference between the measurement and the true value. It can be broken into two parts:
Systematic error (or bias) is a fixed error that is common to all measurements that are made, and doesn’t vary from one measurement to the next.
Random error is error that varies unpredictably from one measurement to the next.
The researchers to get an accurate result incuded the systematic error of people being honest..Whatever that may be.
A measurement is a sum of the true value and a number of irrelevant factors (errors).
We can write this as
M = T+F1+F2+F3+F4+…
where M is the measurement, T is the true value, and F1, F2, … are the effects of the various other factors.
Hope this helps..