Join the minority, and vote for a commissioner

Written by Web News Editor on . Posted in Features

Turn out predictions for Thursday police commissioner vote are dim. The press reports that chief constables are anguished,  fearing that a low poll will rebound on public confidence in the police.  Lack of media coverage (which in part reflects the fact there aren’t elections in London) may have compounded the problems facing any election taking place in the dark days of November.
 
But some commentators are hopeful. Campaigning has been lively in some areas, and the public have been given a welcome push into asking what they want from the police. The numbers underpinning police work have been worked over, strengthening those demanding what officers do is more securely based on evidence of effectiveness. The parliamentary session organised recently by RSS getstats showed a keen appetite for a statistics-led approach.
 
Whatever the strength of their mandate, Police and Crime Commissioners will have to deploy data and may, if they challenge chief constables, stimulate a more informed local debate. Our paper What-PCCs-ought-to-know-October argues that we need to move beyond measures of crime – not all of which are reliable at local level – to look more widely at what stimulates public anxiety (such as anti social behaviour and social disorder) as well as at the other things that the police spend their time doing, including regulating traffic and acting as a first responder in a great variety of social circumstances.
 

New database makes history more human

Written by Web News Editor on . Posted in Features

An article on the BBC News magazine site entitled ‘In Moscow, history is everywhere’ spells out how a newly published database has enhanced our sense of the human aspects of a tragic period in Russian history.
 
The new Memorial Society database marks the addresses of people killed in the purges of Moscow residents in the 1930s-50s including those killed in the purges of 1937 and 1938 when tens of thousands of people were murdered by Stalin’s regime.
 
It is also an account of the role played by one of the leading statisticians  at the time.  Olimpiy Kvitkin was in charge of what is now understood to be the supremely important 1937 census, a census which sought to count everyone in the Soviet Union. This was not a good job to be in when the census showed that rather than the growth which Stalin had announced in 1934,  the population had actually dropped by 6 million in 3 years, because of man-made famines and other aspects of Stalin’s brutal policies.  Kvitkin was accused of being part of “a serpent’s nest of traitors in the apparatus of Soviet statistics” and later shot.
 
Not only does this story demonstrate the power of statistics –  the information in data –  it also lifts history off the page and makes it more meaningful to Muscovites, literally bringing history to their own front doorsteps.
 

More research, more numbers - more evidence-informed policy

Written by Web News Editor on . Posted in Features

We are all guilty of relying too heavily on personal experience – and not evidence – when it comes to views on how society should work.  We tend to assume our own experience is the measure of how people should behave.  My treatment in hospital may have been excellent…but this is not necessarily how things really are.  We have only to read the most recent Care Quality Commission report to know differently.
 
It’s sometimes the same with policy-makers.  Like the rest of us, they can be persuaded by personal experience and what they see and they may over-rely on anecdotal evidence. How often, to prove the success of their policies, have we heard reference  to ‘witnesses’ – a family I’ve just spent the day with, a factory I toured – rather than research evidence and the testimony of numbers as to what does and what does not work.
 
Certainly, evidence can be counter-intuitive and evidence-informed policy making does not always come naturally to those in positions of political leadership. Research needs to be commissioned, you need to wait for its findings. Outcomes may not support a policy a minister is very keen to set in motion  (no minister wants to be remembered for doing nothing!).  Politicians need to think about where to get their evidence from. From special advisers and experts, from lobbyists or think tanks, the media, members of their constituency or academic research?  What about the influence of  ‘gurus’ or studies undertaken by sole individuals which are very popular and taken up in a big way by the media?
 
Then there’s work to do to get policy makers and researchers to understand each others’ perspectives on research evidence. Policy makers see research as based on common sense, contextual, policy relevant, timely, clearly messaged, jargon free,  short, concise and accessible. Researchers see it as scientific (context free), empirically proven, theoretically driven, needing as long as it takes (i.e. not just timely), with caveats and qualifications where necessary, using language specific to the discipline and detailed, comprehensive and methodological.
 
We need a culture that values knowledge and information, that recognises these differences and builds the capacity of policy makers and researchers. That takes into account the many aspects of policy making and understands its pressures –  political, economic and timescale driven pressures.  Also that takes into account and addresses the influence of the media and its responsibility for accuracy in communicating research to the wider public.
 
Statistics – through quantitative research –  has an important role to play in challenging existing assumptions that x or y policy will have or has made a difference.  Researchers often find it easiest to engage ministers with research as policy is being developed and less keen to continue to engage with it, whereas the real effects of policy usually only emerge years later  - in education, as long as 6+ years later – by which time the support for any real statistical analysis may have waned (especially if the research finds that the policy has  not had the intended effect.)
 
And then, many problems are complex and resistant and we need statistical and qualitative data to address them…policy makers, researchers and research mediators working together through the entire policy development, monitoring and evaluation process: designing research, generating research questions, verifying findings etc. There are some great outfits already doing this, but we still need more…if you are interested in reading more about education policy making, turn to our report on Past Marks, a recent seminar in parliament.
 

Health reporting may be on the mend

Written by Web News Editor on . Posted in Features

NHS Choices reckons health reporting has been improving with ‘wonder cures’ hitting the headlines less often and peer-reviewed medial reports covered more responsibly. But the paper illustrated, the Daily Express, is a ‘dishonourable exception’ to the trend, according to the Department of Health-supported information site, which monitors how reports of new therapies and treatments are handled by the media.
 
However ‘headlines can often give a different impression’. Rely on them and you may get a sensationalist and mistaken sense of the news.
 
It’s not that news media have stopped churning out stories about miracle cures, for example (from last year) that curry might help stave off dementia and exercise could change your DNA. In both there’s a kernel of truth: the former reported animal studies and the latter was based on informed speculation about the course of genetic expression. But the facts did not lend themselves to either the headlines or the ‘story’ as presented.
 
Journalists still cannot be trusted with the evidence, especially when it is a matter of probability and subtle distinction between absolute and relative risk. Last June media ran with a story linking men’s tea consumption and prostate cancer. There is a link but so tenuous as to make the story a piece of scaremongering.
 
During the same month another statistical story ran, along the lines that we are 14% more likely to die on our birthday than other days of the year. Unpack the figures and it turns out to be a lot less mysterious.
 
Unfortunately for bloggers, Tweeters and journalists, the message in 2013 remains the same as during last year: being accurate may mean junking the story, unless you are prepared to educate your readers in the subtleties of risk and probability.
 
 

Wise old trees offer data insight into rise of Genghis Khan

Written by Web News Editor on . Posted in Features

Historians have always known it. The weather can play a role in the rise and fall of civilisations and empires. Until now, their focus has mainly been on the role of weather in their decline (for example, the effects of drought on Mayan city states) but new data suggests that the weather has a role to play in the rise of political structures too.
 
Dendochronologist (tree-ring researcher), Dr Amy Hessl of West Virgina University and  Professor Neil Pederson of Columbia University’s session at this month’s American Geophysical Union conference, looked at tree ring data which suggests that a very long  period of wet and warm weather from 1208-1231 helped Genghis Khan and his horde to conquer half of Eurasia and build the world’s biggest contiguous land empire.
 
Until recently, it was thought that periods of drought and survival instincts drove Mongol hordes to invade neighbouring territories. However, it seems more likely that good weather,  good grazing and a well fed herd provided the optimum environmental conditions for Genghis to build an empire. This week’s  Economist article ‘A horde of data‘ is clear, “no one thinks that the Great Khan himself had nothing to do with it’. But his strategic genius might have been naught if the climate had provided him only with broken down nags.”
 
In the second half of the 20th century, Mongolia warmed by 2 degrees centigrade. Knowing more about their past climate may help Mongolians understand how to handle climate change today and so the multidisciplinary project on the energetics and ecology of the Mongol Empire continues, with collaborators from history, archaeology, paleoecology and now ecosystem modeling on board. Data analysis is key to their work and the intention is to  analyse earlier data relating to the first millennium AD  to find out more about the impact climate had on earlier Mongolian tribes too.
 

Resolve to make the world more statistically savvy

Written by Web News Editor on . Posted in Features

Happy 2013 ! An opportunity for a fresh start, a new era. A time when most of us (let’s admit it) have given some thought to New Year resolutions. But which resolutions? and how good are we at keeping them?
 
The top 10 most common resolutions* are listed at Statistic Brain and we can reveal that…at no 1 we have… ”Lose weight’ and at no 10 we have ‘Spend more time with family’. (So, not much emphasis on improving society or the environment then?). However, ’Learn something new’ is on the list  too and all the evidence is that we are more likely to succeed with education-focused self-improvement than we are to lose the pounds.
 
So if education looks to be the way to go and you are thinking of strengthening your stats know-how and skills, why not start by taking our stats quiz and see how you do?
 
Be inspired by Statistics 2013, the International Year of Statistics. This statistics awareness and engagement initiative is supported by over a thousand organisations worldwide, united in their ‘resolve‘ to strengthen statistical understanding and good use of stats.
 
Looking for more ideas?  Reading  ’Some Statistical Habits to add or subtract in 2013‘ and Tips for a Statistically Savvy 2013 by Carl Bialik for the Wall Street Journal, on developing statistical resolutions, should help.
 
And to help us further fine-tune our resolutions, please let getstats know about any statistical bugbears – in the media, in the workplace, in school, in parliament or in any other aspect of life – you would like to see eradicated in 2013.
 
___________________________
 
*Although based on (not so recent!) US research, the list’s ‘weight loss, get fit, spend less/save more’ refrain is timeless, universal and uncomfortably familiar and so ‘just for fun’, we took a closer look. The Statistics Brain listings were published in December 2012, citing University of Scranton, Journal of Clinical Psychology as the source. The underpinning research ‘Auld Lang Syne: Success Predictors, Change Processes,and Self-Reported Outcomes of New Year’s Resolvers and Nonresolvers‘ by Dr John Norcross and colleagues, was originally  published in 2002. The research is based on a study of 282 ‘resolvers’ and ‘non-resolvers’ in 1995 into 1996.
 

When no change is big news: RPI announcement

Written by Web News Editor on . Posted in Features

After a short review, the Office for National Statistics has decided to leave the Retail Price Index (RPI) unchanged. Instead, a new additional index of inflation – RPIJ (the ‘J’ stands for Jevons, a new geometric formula) – will be brought in by March 2013.
 
It’s a very important issue. The RSS statement on the announcement spells it out: “how inflation measures are calculated is not just a technical issue for professional users but one of widespread public importance affecting millions of UK citizens“. RPI and Consumer Prices Index (CPI) figures impact on government out-goings such as pensions and benefits. Also on taxation levels.
 
The National Statistician’s decision to leave RPI broadly as it is (with some tweaking to the measurement of private housing rents) means that index-linked final salary schemes are safe for now – these would have been affected if RPI had been reduced. Train fares, student loans, some utility bills etc might have been affected, but stay the same for now…the list goes on.
 
What are the RPI and CPI? The RPI has been around since the 1950s/40s and is a measurement of the change in the cost of a basket of retail goods and services. The CPI was taken up in the 1990s and measures the inflation of the price of consumer goods and services purchased by UK households. They sound the same but they are different. The RPI doesn’t include households in the top 4% income bracket and pensioners who rely on state benefits for at least 75% of their income. The RPI tracks owner-occupier housing costs, including mortgage interest costs and council tax.  The CPI doesn’t.
 
Perhaps most importantly, the RPI and CPI use different types of averages to reduce 180,000 individual price quotations – a monthly ‘basket’ of goods and services – into fewer individual item indices. The RPI uses arithmetic averages whereas the CPI generally uses another kind of average, the geometric mean. The latter is less or equal to the arithmetic mean.  Generally, this makes CPI around 1% below RPI.
 
Statisticians have long been concerned that these statistical treatments should generate such a substantial difference in the two indices – other countries do not seem to have the same issues. There can be no case for a continued ‘pick ‘n’ mix’ approach to use of one or the other index in different situations according to preference. The ONS review was prompted by wanting to address the  ”formula effect” –  the gap between the RPI and CPI –  and to find out more about how these differences arise.
 
Watch this space.  The new RPIJ measurement will be published in March 2013. In the meantime, the ONS is continuing to pursue its research programme in the area of consumer price statistics and to work with users to maintain the quality of its consumer price statistics.
 
 

Join the RSS

Join the RSS

Become part of an organisation which works to advance statistics and support statisticians

Copyright 2019 Royal Statistical Society. All Rights Reserved.
12 Errol Street, London, EC1Y 8LX. UK registered charity in England and Wales. No.306096

Twitter Facebook YouTube RSS feed RSS feed RSS newsletter

We use cookies to understand how you use our site and to improve your experience. By continuing to use our site, you accept our use of cookies and Terms of Use.