http://rowsdowerisms.livejournal.com/ ([identity profile] rowsdowerisms.livejournal.com) wrote in [community profile] politicartoons2013-03-24 12:17 am

[identity profile] wight1984.livejournal.com 2013-03-24 09:35 pm (UTC)(link)
Very emotive. Not exactly what I'd call a fair presentation of the statistics though.

It lacks a source.
It lacks information on what kinds of deaths are included
It doesn't place the statistic in context (which would mean comparing it to the size of the overall population).

I'm not saying that gun-regulation isn't justified, but if it is justified then we should be able to use statistics properly to prove the point.

[identity profile] wight1984.livejournal.com 2013-03-24 09:43 pm (UTC)(link)
Thanks, that solves one of the three problems I raised.

[identity profile] wight1984.livejournal.com 2013-03-24 09:55 pm (UTC)(link)
[livejournal.com profile] madscience did acknowledge that 'correlation is not causation' and seems to be aware of the problem with using International comparisons to make judgements about national policy: link (http://politicartoons.livejournal.com/3601073.html?thread=81865905#t81865905)

However, the same problems apply even more so to your comparison. It's no good comparing a nation currently without gun laws to a nation currently with gun laws. You need to see what the general pattern is in regards to homicide rates in nations that introduce a ban (or increased regulation) on firearms.

Also, my understanding of the changes to how crime is recorded in the UK since the handgun ban is that the general effect has been to produce lower amounts of recorded crime for the same level of actual crime. I'd have expected that issue to have skewed the statistics in favour of the pro-gun regulation argument.

[identity profile] wight1984.livejournal.com 2013-03-24 10:00 pm (UTC)(link)
Are the statistics inclusive of suicides?

A link to wikipedia doesn't help with the problem with statistics as used in the meme, no.

The problem with the meme is that it gives us the number of deaths in a population without putting it in proportion against the total population number. The second statistic is essential for justifying the extent of the problem under discussion and should be included within the meme (I shouldn't have to search elsewhere for it)

[identity profile] madscience.livejournal.com 2013-03-24 10:22 pm (UTC)(link)
"Recorded crime has gone up in large measure due to new methods of recording crime"

Got a source for that claim?

[identity profile] american-geist.livejournal.com 2013-03-24 10:38 pm (UTC)(link)
Racist trolling is still racist.

[identity profile] american-geist.livejournal.com 2013-03-24 10:39 pm (UTC)(link)
I don't think anyone here would disagree with that.

[identity profile] madscience.livejournal.com 2013-03-24 11:17 pm (UTC)(link)
And that's relevant to the reported increase in crime in Australia because...?

[identity profile] wight1984.livejournal.com 2013-03-24 11:21 pm (UTC)(link)
"Actually the wikipedia link "

Is the wikipedia link part of the image posted?

If not then the image still made bad use of statistics.

[identity profile] wight1984.livejournal.com 2013-03-24 11:29 pm (UTC)(link)
"If by acknowledge, you mean pay lip service to, then sure."

I'll take an acknowledgement over no acknowledgement at all.

"You've got it backwards. "

The information on that is more mixed than last time I looked into it.

[identity profile] wight1984.livejournal.com 2013-03-24 11:50 pm (UTC)(link)
"I don't know how it could possibly be a bad use if the data is presented accurately."

It is very possible to use statistics in a misleading manner. It happens all the time in politics. You never seen a politician paint a inaccurate picture using accurate statistics?

In this case, a number of deaths within a population is given without giving the total population number. That's bad statistics as it gives no sense of scale, and thus can make a problem look bigger than it really is.

An example is International Aid. Every time I see a report on UK contributions to International Aid, it's always 'look at this really big amount of money we're giving away'. The number given is accurate and will seem very big to individual person, but is actually a tiny amount of the UK budget. This gives people a very false impression about the proportion of money spent on International Aid by the UK.

That's bad statistics, despite the numbers being correct.

[identity profile] wight1984.livejournal.com 2013-03-24 11:55 pm (UTC)(link)
My own position is that I don't think there is strong evidence that gun regulation has much of an effect on homicide rates (in either direction).

I don't think people are stupid for thinking otherwise, but both sides of the argument seem to do a lot of clutching at straws for statistical backing. Using the UK's low homicide rate to argue for gun regulation strikes me as an obvious example of dodgy statistics, but (as made clear in my opening) I'm not convinced by the opposing argument either.

Page 2 of 5