Wednesday, November 09, 2011

THIS IS A BAD THING. Do you agree that (a) it is a bad thing, and (b) something should be done about it?

Here is a picture of a peaceful protest
that has nothing to do with the riots.
Today @programmed2run sent me a link to an article reporting that:
70 Percent Of Brits Would Agree With
Shutting Down Twitter During Civil Unrest
It's a shitpot of bad statistics.  I guess, falling short of a majority, the actual figure of 48% wasn't enough to make the story interesting, even though it's an alarming enough figure in itself.

But then All Twitter, the site bearing the article, don't seem to worry too much about accuracy.  They accompanied an article about the August riots with a picture of a pensions strike march.

The 70% figure apparently includes the 22% who gave the rather wishy-washy answer of "somewhat agree", although I can find no mention of this figure anywhere in the report by Unisys, who carried out the research.  Nor is it mentioned in their press release, so God only knows where All Twitter spirited that figure from.

Even the 48% figure seems potentially rather woolly to me.  The poll was conducted with 973 people by a "random digit dialing (RDD) sample of telephone households in the UK".

Does this mean landlines?  I think it means landlines.  I wonder who gets missed out if you only call landlines.  A lot of people only have mobile phones these days.  I'd be very interested to know if there's a relationship between landlines and social network use.  I have no serious misgivings about the telephone thing, but I'm sure there must be some kind of impact.  And maybe they did include mobile phones as well.

Then one of the other figures in the report caught my eye.  I know that news outlets are prone to mangling statistics, but even Unisys can't manage to be consistent with their language.  According to their press release:
"A similar proportion (46%) accepts that the authorities should be able to access data from social networks if it helps improve public safety."
Oh, well then.  If it's to improve public safety, then I'm all for it.  If the authorities only get to access data when they really need to, like when someone's committed some horrific crime, then I'm surprised the 46% figure isn't higher.  Because at other times, there's no way they'd be allowed to.

Here's the actual statement that respondents were asked to agree or disagree with:
"The authorities should have open access to data about social network users in order to prevent coordinated criminal activity."
Er, not quite the same thing really, is it?  In fact, there's no mention of it protecting the public at all, only preventing crime.

There are two main problems here, as far as I can see.

1) Unisys asked many people their opinions on something they know nothing about.

Unisys did not ask users of social networks.  They asked nearly 1000 people, many of whom will know about social networks, and many who will have very little understanding of them.  Many of the respondents would have been people who have no idea of the power of good that Twitter et al can achieve.
"Hello.  I'm going to ask you your opinion about something you don't use.  And then I'm going to give your ignorance equal weighting to someone who knows what they're talking about."
2) And then they were given a question preceded by a MASSIVELY leading statement:
"During recent unrest in major UK cities, social networks such as Facebook and Twitter were used to coordinate criminal activity.  Do you agree with these statements?"
a) The authorities are playing catch up and need more resources to monitor online behaviour (49% agreed)
b) During outbreaks of unrest, providers should temporarily shut down social networks to prevent coordinated criminal activity (48% agreed)
c) The authorities should have open access to data about social network users in order to prevent coordinated criminal activity (46% agreed)
d) Providers of social networks should get more information on the people using their services before they allow use (42% agreed)
No indication given to the respondents of the many benefits of social networking.  No hint or clue that the clean-up operation after the riots was conducted largely by people who coordinated themselves via Facebook and Twitter.

It's like approaching someone who has no concept of all the useful things a knife can do and saying, "Knives have been used to stab people.  Do you agree that all knives should be banned?"

No wonder the percentages were so high.

No comments:

Post a Comment

Hello. Feel free to leave your comments and related links here. (Be nice.)