April 2, 2012
Stat of the Week Competition: March 31-April 6 2012
Each week, we would like to invite readers of Stats Chat to submit nominations for our Stat of the Week competition and be in with the chance to win an iTunes voucher.
Here’s how it works:
- Anyone may add a comment on this post to nominate their Stat of the Week candidate before midday Friday April 6 2012.
- Statistics can be bad, exemplary or fascinating.
- The statistic must be in the NZ media during the period of March 31-April 6 2012 inclusive.
- Quote the statistic, when and where it was published and tell us why it should be our Stat of the Week.
Next Monday at midday we’ll announce the winner of this week’s Stat of the Week competition, and start a new one.
The fine print:
- Judging will be conducted by the blog moderator in liaison with staff at the Department of Statistics, The University of Auckland.
- The judges’ decision will be final.
- The judges can decide not to award a prize if they do not believe a suitable statistic has been posted in the preceeding week.
- Only the first nomination of any individual example of a statistic used in the NZ media will qualify for the competition.
- Employees (other than student employees) of the Statistics department at the University of Auckland are not eligible to win.
- The person posting the winning entry will receive a $20 iTunes voucher.
- The blog moderator will contact the winner via their notified email address and advise the details of the $20 iTunes voucher to that same email address.
- The competition will commence Monday 8 August 2011 and continue until cancellation is notified on the blog.
Statistic: Smoking costs the NZ health system $7 billion
Source: The Press
Date: Saturday, 31 March, 2012
There are all kinds of reasons the number doesn’t pass any smell test. First off, it’s about $1600 per capita. Second, it’s half of total health system expenditures. Third, it’s twenty-times bigger than the estimate produced by Action on Smoking & Health and the SmokeFree Coalition; it’s 28 times bigger than the estimate produced by the Cancer Society.
Discussed here.
The Fairfax papers are horrible for this particular kind of innumeracy. Any big scary number is plausible if it’s about the costs of something the journalist doesn’t like. And there seems to be zero fact checking on the stories either. It takes less than a freaking minute on Google to find NZ-based estimates on this; instead, the author extrapolated from foreign stats. The top hit I get on a Google search here has the Cancer Society saying it costs $250 million per year. Maybe that’s wrong, but it sure as heck ought to have given the Fairfax journalist some pause. There’s just a bit of a gap between $250m and $7000m.
13 years ago
Statistic: There are between 246 and 1,476 people in Palmerston North not addressing their gambling problem.
Source: Manawatu Standard
Date: 3 April 2012
The Manawatu Standard informs us that [i]n Palmerston North, there were 66 people in counselling, but between 246 and 1476 others were likely to have problems that they were not addressing.
Sixty six seems like a ‘solid’ number coming from recorded counselling patients; however, the estimated interval looks curiously accurate, not between, say, 200 and 1500 but 246 and 1476. 246:1476 is exactly 1:6, which is the same ratio presented by the The Problem Gambling Foundation (0.3% to 1.8% of adults have gambling problems in New Zealand). The current population of Palmerston North is estimated at 82,000 people, which multiplied by 0.003 and 0.018 gives 246 and 1476 respectively. However, this does not account for the roughly 20% of the population that are under 18 yo.
The publication not only provides a misleading sense of accuracy, but it also applies the estimated percentage of problem gamblers to the wrong population.
13 years ago
Statistic: Smoking costs
Source: as above
Date: as above
It should have been obvious to me before, but here’s exactly how the journalist screwed this one up.
First, he cites a WHO figure on tobacco’s burden on the health system (6-15% – higher than either existing NZ figure, but let’s leave that aside).
Second, he cites an Australian figure on tobacco costs in the aggregate relative to GDP. He doesn’t notice that this is a measure inclusive of premature mortality, presumed productivity losses, smokers’ expenditures on cigarettes and the like. So he takes this as a measure of the burden imposed by smokers through the tax system.
Finally, he extrapolates from a misinterpretation of what’s likely the Collins & Lapsley measure of the aggregate social costs of smoking in Oz to say that smoking costs the NZ health system $7 billion. That’s more than 50% of the health budget, and consequently many times outside of the range he cited only a couple lines earlier on the WHO estimates. But a lack of basic numeracy meant no flags went up.
This is one reason I absolutely despise “Social Costs of X” studies. It is so easy to read them as providing a measure of the costs that X imposes on others when they’re really mostly measures of how much smokers spend on tobacco and the monetized value of their reduced life expectancy. I’m pretty sure that’s what’s here happened.
The same thing happened in Oz with Collins & Lapsley’s measure of the Social Costs of Alcohol which has alternatively been portrayed by legislators and the press as the crime costs of alcohol, the productivity costs, the mortality costs, or the health costs falling on the taxpayer; the measure is inclusive of all of that and more….
13 years ago
Statistic: Annual crime statistics
Source: Dominion Post
Date: 3 April 2012
Cheers and jeers for the Dominion Post reporting:
Cheers:
The Dom Post reported a Lower Hutt Detective Inspector as noting that ‘public perception was that crime was increasing’, even though the number of crimes has fallen in Lower Hutt and nationally. This is a great reminder of the importance of statistics, especially in public policy.
Jeers:
The article (and many others) reported that ‘recorded crime rate is at its lowest in 15 years’, and then reported the raw number of offences (406,056). The changes to crimes also appeared to be calculated from the raw numbers. It would be much more helpful to see real crime rates — numbers of crimes per 100,000 people or suchlike. For example, the increased crime in the Waikato (up 1.7 per cent) could be a reduction in the crime rate, depending on population growth.
13 years ago
Statistic: Crime doubles close to liquor outlets
Source: Sunday Star-Times
Date: 8 April 2012
If I were going to set up a business in a socially deprived area and had a choice between a liquor store and a Louis Vuitton store then I would naturally go for the former. So there is no surprise that certain areas have certain store densities.
If they did a study where they used distance to the nearest Louis Vuitton store then I’m sure that they would find that areas with greater distance have more crime. Would they then argue that Louis Vuitton stores should be put into poor areas to reduce crime there?
Even if you do a really careful analysis and adjust for all the factors involved and still find a relationship between store distance and crime it doesn’t mean that stores cause crime. It could just be that people of a criminal disposition like to (or can only afford to) live in areas near liquor stores.
13 years ago