December 26, 2011
Stat of the Summer Competition: December 24 2011 – March 2 2012
This summer, we would like to invite readers of Stats Chat to submit nominations for our Stat of the Summer competition and be in with the chance to win a copy of “Beautiful Evidence” by Edward Tufte:
Here’s how it works:
- Anyone may add a comment on this post to nominate their Stat of the Summer candidate before midday Friday March 2 2012.
- Statistics can be bad, exemplary or fascinating.
- The statistic must be in the NZ media during the period of December 24 2011 – March 2 2012 inclusive.
- Quote the statistic, when and where it was published and tell us why it should be our Stat of the Summer.
On Monday 5 March 2012 at midday we’ll announce the winner of the Stat of the Summer competition, and restart the weekly competition.
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.
- 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 copy of “Beautiful Evidence” by Edward Tufte.
- The blog moderator will contact the winner via their notified email address and request their postal address for the book to be sent to.
- The competition will commence Saturday 24 December 2011 and continue until midday Friday 2 March 2012.
Statistic: Smokers each cost the economy $139,000 per year.
Source: The Quit Group, quoted on Radio NZ
Date: 1 January 2011
The number is insane, as detailed here. They’re claiming a smoker costs the economy three times per capita GDP.
I’ve not seen the background workings on how they came up with this one; I don’t know whether RNZ or The Quit Group botched it. But the number has surprising implications. The 2010 Social Report says 22% of folks aged 15-64 smoked in 2009. Suppose that it’s only 10% for the country as a whole. Can smoking really cost the country 30% of aggregate GDP? Could we plausibly increase GDP by a third if smoking disappeared?
13 years ago
Statistic: Smoker costs
Source: as above
Date: as above
Update to prior one: The Quit Group reports by email that their figure was a present discounted value of a stream of costs, not an annual cost. So either QG’s spokesperson misspoke or was misquoted. Either way, confusing an annual cost with a PDV of a stream of costs inflates things by a bit over an order of magnitude at an 8% discount rate.
Noted here.
13 years ago
Statistic: Caffeine poisoning on the rise, ”warning call” for health authorities
Source: Stuff.co.nz, The Medical Journal of Australia
Date: 16/01/2012
While the original study (not linked in the news) does show an increase on the number of calls, it also mentions that the NSW poisons centre takes an average of 110,000 calls per year. Thus, the peak of caffeine related calls (year 2010) represents only 0.06% of all calls received by the centre. (Reporting without mention to the reference population).
In addition, the original paper’s claim that “labelling and any marketing of these products should include appropriate health warnings and the national poisons hotline number” is strange, given that energy drinks do not have more caffeine that a standard cup of coffee.
13 years ago
Statistic: “Private economists’ company BERL has put the social cost of alcohol harm at $4.8 billion but it has been estimated as high as $16.1bn.”
Source: TV 3
Date: 20 January 2012
My paper thoroughly debunking the BERL figure, as well as a related Australian figure.
– $700 million of the figure is drinkers’ spending on their own alcohol
– hopeless double counting of productivity costs of premature drinker mortality with the intangible costs of premature drinker mortality (contrary to how Ministry of Transport uses their Value of a Statistical Life figure)
– Vast majority of costs are borne by the drinker. Social costs – costs borne by others – are closer to the $967 million mark.
13 years ago
Statistic: Infidelity: it’s a right-wing, meat-eaters’ thing
Source: NZ Herald
Date: Thursday Jan 26, 2012
It is an example of the media grossly misinterpretting statistics and making them sound bad to the general population when in reality the figures probably largely resemble something close to national averages
13 years ago
Statistic: “Auckland has become a city of extremes, containing more than half of the country’s total students in both the richest (decile 10) and the poorest (decile 1) schools, with relatively few in between.”
Source: NZ Herald
Date: February 10
Really…with with more than half in decile one and and decile 10 that means nearly half are in between not “relatively few”.
13 years ago
Statistic: “Unfit to teach – the roll of shame”
“Hundreds of teachers have criminal convictions and many are not fit to teach” – opening para in DomPost story, reprinted on Fairfax websites
Source: DomPost
Date: 13 February 2012
The DomPost story says a total of 293 teachers were referred for criminal convictions, 103 for competency and 277 as a result of other complaints. However, according to the story’s fine print, it appears only 50 of these teachers were de-registered after investigation over a two-year period.
Given the Teachers Council says there are 100,000 teachers, and (say) 25 teachers were de-registered in each year, this can hardly lead to the conclusion that “many” are unfit to teach.
http://www.stuff.co.nz/dominion-post/news/6406110/Hundreds-of-unfit-teachers-in-class
13 years ago
Statistic: Departing students owe more than those who stay
Students who head overseas after completing their tertiary education tend to have bigger loans than those who stay and work in New Zealand.
Source: NZ herald
Date: 13 Feb, 2012
I am confused by the title and the content here.
The article showed some comparsons but i dont understand the author’s purpose.
Having talked to a colleague and i know what heppens here. Graduates with more loan tend to go overseas to earn more to repay or just avoid to repay.
The title makes me think as long as graduates leave NZ they will owe more.
13 years ago
Statistic: Child-abuse on rise in Oamaru
Reported child abuse in Oamaru almost doubled last year, latest Ministry of Social Development statistics show.
Oamaru recorded 57 abuse cases for 2011, compared with 29 the previous year…Oamaru’s child-abuse figures peaked in 2009 with 73 cases, after two years of lesser offending in 2006 (32) and 2007 (36).
Source: Oamaru Mail
Date: February 3, 2012
Given 2011 falls about halfway between the 2010 and 2009 figures I would not presume some definite upward trend.
Also one of several crimes where increased reporting is presumed to be mean increased offending when it can indicate increased awareness and intolerance of this sort of behaviour.
13 years ago
Statistic: ‘Mondayising’ holidays would cost $400m
The department had told Mr Key the cost to the economy for an individual day if Mondayised was $200 million or $400 million in a year when both were Mondayised.
Source: New Zealand Herald
Date: Feb 13, 2012
Using an absolute total to make it sound like a huge financial burden. I am not sure how often Easter and ANZAC coincide but 2011 was the only time in the last thirty years so generally six out of ten years employers have accept the cost of both holidays when neither has fallen on a weekend.
13 years ago
Statistic: Old people have displaced more than 40,000 teenagers from jobs in the past five years as more choose to stay on in the workforce and employers shun youth for experience.
Source: New Zealand Herald website
Date: 17th February 2012
I think this is a bad statistic for the following reasons:
1) The statistic above uses the word ‘displaced’ to suggest that the decrease in the number of 15-19 year olds employed in the last five years is a direct result of over-65s choosing to stay in the workforce rather than retiring. However, the chief piece of evidence, cited from the Salvation Army report, is indicative of a correlation between the increase in the number of over-65s in employment and a decrease in the number of youths employed in the last five years, not causation. There could be factors resulting in the decrease in the number of youths employed other than the increasing number of over-65s in the workforce and employers’ attitudes towards 15-19 year olds and over-65 year olds.
2) Even if the rise in the number of over-65s in the workforce did account for the decrease in the number of 15-19 year olds in the workforce during the last five years, there is no evidence to suggest that this rise accounted for the displacement of all the 40,000 (and more) teenagers from jobs in the past five years.
3) While it may seem intuitive, there is little evidence provided in support of the claim that “employers shun youth for experience” and that this ‘shunning’ is the reason for the decrease in the number of 15-19 year olds in the workforce. It is merely a hypothesis.
13 years ago
Statistic: the $1.35 billion Ultrafast Broadband (UFB) and $300 million Rural Broadband Initiative (RBI) will add $5 billion to GDP over their 20-year life time
Source: Alcatel Lucent
Date: 21/2/12
(1) the “raw data” linked to by NBR shows that the calculation is based entirely on project COSTS.
(2) GDP is a measure of value-added, which (summarising aggressively) is defined as Benefits (a.k.a. Value) MINUS Costs.
(3) Those cash costs could have been used for other stuff (schools, hospitals). It was not flown in by some munificent benefactor, but raised from Kiwis via taxation.
(4) So the investment cost must be deducted from any GDP benefit.
That’s enough to make the case for this being outrageous stat of the week/summer. We’re talking about $billions of our taxpayer funds here.
13 years ago
Statistic: “Students flee NZ over job fears” – several statistics given.
Source: NZ Herald
Date: 25 Feb 2012
There is no context in the article – have such surveys been given before in NZ? in other countries? The headline is also misleading.
The survey is an online one (albeit by a well known polling company) – some comments about methodology should surely be given.
Finally, how hard is to link to the actual report, so readers can make up their own mind? Have Julian Assange and data journalism passed the Herald by? For example, I would like to see the exact wording of the survey questions.
13 years ago
Statistic: Divided Auckland
Source: New Zealand Herald
Date: February 2012
February was the Herald’s poverty/inequality month with a series of articles, columns, and an editorial with the theme “Divided Auckland (http://www.nzherald.co.nz/22divided20auckland22/search/results.cfm?kw1=%22divided%20auckland%22&kw2=&st=gsa). But it was flawed because they assumed as truth something that is only a hypothesis and assumed that correlation implies causation (that inequality causes poverty).
The essence of the story is that (Brian Fallow http://www.nzherald.co.nz/wages-and-salaries/news/article.cfm?c_id=277&objectid=10785754) since 1988 the real (after inflation) income of poor people has risen 13%. Yet at the same time the diseases of poverty have increased (http://www.nzherald.co.nz/economy/news/article.cfm?c_id=34&objectid=10786897).
Why?
The Herald doesn’t explicitly ask this question let alone investigate an answer. They explicitly assume that the cause is inequality per se (http://www.nzherald.co.nz/nz/news/article.cfm?c_id=1&objectid=10788241). Tapu Misa says “the more unequal a society, the sicker and more dysfunctional.”
This disfunctionality is only a hypothesis, not a fact that can be used as explanation. And anyway, even if it applies to the world in general it may not apply to New Zealand in particular.
The series hit its prejudiced low with the article “wealth gap linked to rise in infections” (http://www.nzherald.co.nz/nz/news/article.cfm?c_id=1&objectid=10786897) which reports on a health study. The article reports an “increase in infections which were associated with a widening wealth gap …”.
But the study did not actually measure any changes in the wealth gap. It reported that poor people had more infections than rich but it did not actually measure any CHANGES in the gap between poor and rich. The Herald with its blinkers on did not really read the article.
I wonder if increasing poverty in the face of increasing incomes among the poor is due to a disproportionate increase in housing costs. The Herald had an article on housing but the Herald did not investigate this aspect. Instead they claimed, with absolutely no evidence, that “the growing gap between rich and poor has made an extra one sixth of Auckland truly tenants…” And anyway, what’s wrong with being tenants? Fewer Germans own their own homes than New Zealanders.
Many words were devoted to how little New Zealand rich are taxed compared to other countries (“NZ tax on rich among lowest in the world”), the implication being that tax rates should rise. But given that “the top 10 percent of households contribute over 70 percent of income tax” (Bill English) I would think that getting more people into the top 10% (you know what I mean) would be a more effective way of raising taxes than raising tax rates. That would actually increase the wealth gap but provide more spending on the poor. But the Herald, with its blinkers on, doesn’t entertain this possibility.
Misa says “the causes of poverty are multidimensional and complex.” Instead of just blaming the rich the Herald should investigate these causes.
13 years ago
Statistic: West Coast/Buller among the happiest regions
Source: Scoop
Date: 20 February 2012
The thing that struck me as odd, almost humorous, about this story which reports that West Coast/Buller is the happiest region in NZ is that in 2008 the survey reported the same region as the LEAST happiest. Surely the region hasn’t changed that much in that little time!
My best guess is that this is an example of getting the sampling error wrong.
UMR interviewed thousands of people and because of this large number they put the margin of error at a very low 1.2%. Here’s what they say:
The analysis focuses on one question, “Using a 0 to 10 scale where 0 means very unhappy and dissatisfied and 10 very happy and satisfied please tell me how happy and satisfied you are with the way your life is going”
“Most of the analysis in this report looks at the respondents who gave a 9 or 10 on the 0 to 10 scale. The sample size for this analysis is n=7071. The margin of error for a 50% figure at the 95% confidence level is +/- 1.2%.”
I don’t know what that means. Where does a 50% figure come in on a 9 to 10 scale? Why are they only looking at 9 and 10? Don’t the unhappy people at 0 and 1 matter?
But the main source of error is not sampling error, it’s questionnaire error because the questions are self-answered about psychological state of mind which will vary if you repeat the survey on the same person and even on the same day.
This is evident from their chart on page 7 of this report at http://img.scoop.co.nz/media/pdfs/1202/HappinessOfNZUpdate_Jan12.pdf . There is a lot of noise and it seems to my eye that the difference between 2008 and 2010 for West Coast / Buller is within the noise. Given the amount of noise I would be reluctant to say that any one region stands out from another.
This difference is surely due to chance and not that region becoming genuinely more happy. But we mustn’t forget the Pike River disaster. Often a tragedy like that has a positive effect on the survivors (which may be why the suicide rate has dropped in Christchurch). But it still looks like noise to me.
[The 2008 report is at img.scoop.co.nz/media/pdfs/0803/Happiness_of_NZ__Final.pdf and says “the least region happiest is the West Coast / Buller.”]
13 years ago
Statistic: That you are alive today to read this sentence proves you have some talent for managing risk.
Source: New Zealand Listener March 10-16 2012
Date: 2 March 2012
The Listener starts their feature story with “that you are alive today to read this sentence proves you have some talent for managing risk.”
This is NOT true for me. Most people that were born in the year I was born are still alive so the fact that I am one of them is not an indication of talent.
In fact, were I to die tomorrow, I would have died well before my time and this would be an indication of a lack of talent.
If not true for me then could the statement be true in general for Listener readers? As “proof” of talent I would require someone to be in the top 5% (P < 0.05) of survivors. Since life expectancy is roughly 80 for New Zealanders then I would guess that the top 5% of survivors would be in their 90's. Are most Listener readers in their 90's? Could be. I haven't seen any statistics.
The same Listener story gives the odds of the "average person" living to 100 as 1 in 50. I didn't find official odds for New Zealanders but this page http://www.guardian.co.uk/news/datablog/2011/aug/04/live-to-100-likely has odds for UKers (download the spreadsheet from the Office for National Statistics).
The spreadsheet is VERY interesting. You might like to think about what you would expect it to look like before you continue reading…
The odds are highest for 99 years olds (of course) at 67.6% and drop to a minimum of 7.2% for 84 year olds. Then they start rising again and reach 29.9% for new borns. The reason for the rise is because newborns have the most exposure to today's high life expectancies.
The average New Zealander is in their 30's (I guess) and, if similar to a UKer, has about a 20% chance of reaching 100. A LOT more than what the Listener says.
13 years ago
Statistic: Christmas rise gives PM $3900 backpay, $150 more a week
Prime Minister John Key will get $3895 in backpay and an extra $150 a week as a result of a 1.9 per cent increase for MPs awarded by the Remuneration Authority.
Source: NZ Herald
Date: 21 Dec, 2012
1.9% is not a large increase in NZ’s public sector. But the media just want to emphasise the absolute value rather than the percentage.
12 years ago