It’s rarely black and white

Taking statistics out of context to push a particular agenda is nothing new. But it’s nice to see a pushback.

Fixing the ‘impeach this’ map with a transition to a cartogram
As discussed previously, the “impeach this” map has some issues. Mainly, it equates land area to votes, which makes for a lot of visual attention to counties that are big even though not many people live in them. So, Karim Douïeb used a clever transition to change the bivariate map to a cartogram. Now you can have a dual view.

its-rarely-black-and-white-1

We just need more of this kind of thing over here. For instance:

Show this chart to anyone who says Brexit is the ‘will of the British people’
This chart is not an entirely convincing argument against Leave or Remain, but it does illustrate that ‘the 52 per cent’ and ‘the 48 per cent’ actually constitute much smaller proportions of the UK population than the figure might imply.

its-rarely-black-and-white-2

Poor performance

For such a small number, a school’s Progress 8 score can be quite a big deal. So the last thing we need is an exam board messing up the performance tables process by not sending complete data to the DfE.

Progress 8 error in performance table checking after BTEC gaffe
Peter Atherton, data manager at a school in Wakefield, told Schools Week some schools had received a “nasty surprise” when they went to check the website.

“It could be the case that, if all of these qualifications were missing for your school, that could affect your progress 8 score by quite a lot. Some schools are saying they’re -0.20 below what they were expecting.”

Gaffe is such an odd word, if you think about it. French, I guess. Would that make Pearson a gaffeur?

It is the second gaffe relating to BTECs to hit the exam board this year.

In August, Pearson was forced to apologise after it hiked grade boundaries for its BTEC Tech Awards just days before pupils were due to collect their results, meaning youngsters were handed lower grades than they were expecting.

Requires improvement, I’d say.

Do we know what’s really going on?

It seems there are three kinds of people in the world: fools that believe in ludicrous conspiracy theories; bullies that want to persuade us that established facts are conspiracy theories when they’re plainly not; and us, stuck in this post-truth world, trying to get to the bottom of it all.

What you think you know about the climate is probably wrong – new UK poll
But our lack of understanding of the scale of the issues doesn’t mean we’re not worried. In fact, recent polling of Britons by Ipsos MORI measured record-breaking levels of concern. Our new polling also shows that two-thirds of Britons reject Donald Trump’s assertion that global warming is an “expensive hoax” – and instead two-thirds agree with the recent UK Parliament declaration that we are facing a “climate change emergency, with the threat of irreversible destruction of our environment in our lifetime”.

Things are confusing enough without all these concerted efforts to massage the truth.

Five climate change science misconceptions – debunked
This organised and orchestrated climate change science denial has contributed to the lack of progress in reducing global green house gas (GHG) emissions – to the point that we are facing a global climate emergency. And when climate change deniers use certain myths – at best fake news and at worse straight lies – to undermine the science of climate change, ordinary people can find it hard to see through the fog.

It’s not just the climate crisis, of course. Remember that damned bus?

Citizens need to know numbers
The message on the bus had a strong emotional resonance with millions of people, even though it was essentially misinformation. The episode demonstrates both the power and weakness of statistics: they can be used to amplify an entire worldview, and yet they often do not stand up to scrutiny. This is why statistical literacy is so important – in an age in which data plays an ever-more prominent role in society, the ability to spot ways in which numbers can be misused, and to be able to deconstruct claims based on statistics, should be a standard civic skill.

No idea

I’m glad I’m not the only one who feels a little clueless sometimes. But, as Amit Katwala from Wired UK concludes, “you have to question the ‘wisdom of crowds’ and the ‘will of the people’ when five per cent of Brits don’t even know if they’ve planned their own funeral.”

Why I’m obsessed with people who respond ‘don’t know’ to really obvious YouGov questions
Scrolling through the results of similar polls over subsequent days, weeks and months, I found a country that is deeply confused on a lot of seemingly straightforward issues. Two per cent of Brits don’t know whether they’ve lived in London before. Five per cent don’t know whether they’ve been attacked by a seagull or not. A staggering one in 20 residents of this fine isle don’t know whether or not they pick their nose …

Digging into the demographics is fascinating. The further north you go in the country, the less likely people are to know whether or not they can swim. Older people are less likely to have been surfing, but more likely to know whether or not they have.

Seventeen per cent of 18- to 24-year-olds don’t know if they’ve ever attended, watched or listened to Glastonbury, compared to zero per cent of the over-65s. Generally (based on my admittedly unscientific whizz through the funniest-looking poll questions), the 18- to 24-year-old age bracket seems the most likely to respond ‘don’t know’ – which could be down to their relative lack of life experience, or perhaps general indifference to surveys as a whole. By the same token, maybe it’s unsurprising that the over-50s seem pretty sure of themselves.

League table fatigue

Another weekend saw another trip to a university’s Open Day, and another PowerPoint presentation full of league table statistics…

Unknown pleasures: exciting new uni rankings
As this commentary from my colleague Professor Mike Merrifield observes using university league tables is not a great way to choose where to apply for a university place both because of the way they are compiled and their inherent flaws.

But there are plenty of them out of there and they never stop coming. And we are now in peak league table season. So, put your sceptical face on and have a look at a couple of the most recent major offerings.

Meanwhile:

Vast differences exposed in graduate outcomes
New data published today shows the wide variation in graduate outcomes depending course and institution.

Government reveals student loan contribution
Data published today shows forecasts for student numbers, the cost of student loans and loan repayments in England.

Self-improvement

The Economist’s charts are usually very clear and helpful, but that’s not to say they can’t be improved – as they themselves show.

Mistakes, we’ve drawn a few
At The Economist, we take data visualisation seriously. Every week we publish around 40 charts across print, the website and our apps. With every single one, we try our best to visualise the numbers accurately and in a way that best supports the story. But sometimes we get it wrong. We can do better in future if we learn from our mistakes — and other people may be able to learn from them, too. …

Misleading charts
Let’s start with the worst of crimes in data visualisation: presenting data in a misleading way. We never do this on purpose! But it does happen every now and then. Let’s look at the three examples from our archive.

Mistake: Truncating the scale

self-improvement-1

Statistically insignificant?

One of the dangers at just looking at the numbers.

Progress 8 scores for most schools aren’t that different
There were over 300 schools with P8 scores between -0.05 and +0.05 – a difference of over 300 rank places (10% of schools) between the highest and lowest scoring of them. But what do these numbers mean?

Let’s say the score for School A was +0.05 and School B was -0.05. Taking the numbers at face value, one interpretation is that if you picked two pupils with the same KS2 attainment, the two pupils would have the same grades in seven of the subjects included in Attainment 8 but the pupil from School A would have one grade higher in one and only one subject than the pupil in School B.

Is this an educationally important difference?

It depends?

And talking of Progress 8 confidence intervals…

statistically-insignificant

xkcd: Error bars

In HE, everyone’s at it *

* Being deceitful, that is. Or maybe just willfully ambiguous?

Let’s start with Alex Hayman from Which? University

Students need clarity when choosing a university
It has been almost a year since the Advertising Standards Authority upheld a number of complaints about misleading information in HE. Despite the clear warnings, we’ve investigated and found at least six universities included examples of unsubstantiated or unverifiable claims about their standing on their websites, in likely breach of advertising standards. This just isn’t good enough.

Various examples are listed. This is an interesting line from the Guardian.

UK universities ‘still inflate their statuses despite crackdown’
The continued use of assertions about high international status is evidence of the strain universities are under to increase their domestic and international student recruitment, as well as the effects of global rankings.

Hmm. But, of course, it’s not just the universities that are happy to stretch the truth.

Essay mills: ‘One in seven’ paying for university essays
More students than ever are paying someone else to write assignments for them via “essay mills”, a Swansea University study has revealed. The survey of more than 50,000 students, found 15.7% admitted to cheating since 2014 – up from an average of 3.5% over the last 40 years.

[…]

It showed the amount of students admitting to contract cheating, when students pass off a custom-made essay as their own, has increased over time. But Prof Newton, director of learning and teaching at Swansea Medical School, suggested the number could be much higher, as students who have paid for essays to be written are far less likely to volunteer to take part in surveys on cheating.

Indeed. But what if you want to exaggerate how much work you’ve done, but are a little short on funds? Free fonts!

Times Newer Roman
Introducing Times Newer Roman, a font that kinda looks like Times New Roman, except each character is 5-10% wider. Fulfill lengthy page requirements with hacked margins, adjusted punctuation sizing, and now, Times Newer Roman!

in-he-everyones-at-it-2

Times Newer Roman is a sneaky font designed to make your essays look longer
According to Times Newer Roman’s website, a 15-page, single-spaced document in 12 point type only requires 5,833 words, compared to 6,680 for the standard Times New Roman. (That’s 847 words you don’t need to write, which is more than twice the length of this post!)

[…]

Of course, it’s the digital age, so there are some downsides: Times Newer Roman will only work for assignments you have to submit by hand or in a PDF. If you’re sending in a Word document using a custom font that professors almost certainly don’t have installed won’t help. Similarly, Times Newer Roman is only useful for hitting larger page counts; if you have a strict word count limit, you’re out of luck.

Happy statistics day

It’s GCSE results day and, despite the new grading system, the news people are bringing out updated versions of their usual it’s-getting-better-it’s-getting-worse stories.

GCSE results day 2018: New ‘tougher’ exams favour boys as gender gap narrowest in seven years
Girls remain in the lead, with 23.4 per cent achieving one of the highest grades, which is the same as last year, compared to 17.1 per cent of boys, up from 16.2 per cent last year. But the gap in top grades between boy and girls is now at its narrowest since 2010, with boys just 6.3 per cent behind girls, down from 7.2 per cent last year.

GCSE results rise despite tougher exams
A total of 20 of the most popular GCSE subjects in England have been graded for the first time in the numerical format – plus English and maths, which were introduced in the new format last year. These include history, geography, sciences and modern languages, all of which have been designed to be more difficult.

Of those achieving all grade 9s – and taking at least seven of the new GCSEs – almost two-thirds were girls.

GCSEs: boys close gap on girls after exams overhaul
Boys appear to have been the major beneficiary of the overhaul of GCSE examinations taken in England for the first time this summer, as results showed across-the-board improvements in boys gaining top marks while girls saw their share of top grades dip.

Across the UK the proportion of students gaining an A or 7 and above, the new top grade used in England, rose above 20%, with boys in England closing the gap on girls with an almost one percentage point rise to 17.1% with girls unchanged at 23.4%.

In the reformed GCSEs in England, 4.3% of the results were the new highest 9 grade, set at a higher mark than the previous A* grade. The figures on Thursday showed 732 students attained seven or more grade 9s.

Despite the improvements by boys in England they were still outperformed by girls at the highest level: 5% of entries by girls received 9s, compared with just 3.6% of boys.

GCSE pass rate goes UP – but fewer students get new top ‘9’ grade compared to old A* mark
The overall pass rate – the percentage of students getting a 4 or above or a C or above – was 66.9 per cent, compared with 66.4 per cent last year.

But just 4.3 per cent of exams were given the new 9 grade, which was brought in to reward the absolute highest achievers. Just 732 students in England got a clean sweep of seven or more grade 9s.

Previously around seven per cent of exams scored the top A* grade.

The more detail-oriented education sector websites are worth a read, if you really want to dig down into all this.

GCSE results 2018: How many grade 9s were awarded in the newly reformed subjects?
There has been a curious amount of interest in how many students might achieve straight 9s in all subjects. It seems to have started with a throwaway remark on twitter by the then-chief scientific adviser at the Department for Education that only two students would do so. Tom Benton from Cambridge Assessment then produced some excellent research showing that it would, in fact, be several hundred.

Today Ofqual has answered the question once and for all. A total of 732 students who took at least seven reformed GCSEs achieved grade 9 in all of them. Given that fewer grade 9s are awarded than grade A*, it should come as no surprise that fewer students will achieve straight grade 9s compared to straight grade A*s.

But it can get a little heavy-going at times.

GCSE results day 2018: The main trends in entries and grades
Across all subjects, 21.5% of entries were awarded a grade 7/A or above, compared to 21.1% last year. At grade 4/C or above, 69.3% of entries achieved the standard this year, compared to 68.9% last year. Both figures have been on something of a downward trend since 2015, so this year’s figures arrest this decline.

GCSE and A-Level results analysis
Explore trends in national entry and attainment data between 2014 and 2018 in:
All subjects
Additional mathematics
Additional science
Art and design subjects
Biology
Business and communication systems
Business studies
Chemistry
Citizenship studies

GCSE 2018 variability charts: Are your results normal?
Each year Ofqual produces boring-sounding variability charts. It sounds dull but they show how many centres, i.e. schools or colleges, dropped or increased their results compared with the previous year. This means that if you dropped, say, 25 per cent in one subject, you can see how many other schools also saw the same dip.

Let’s give the last word to the JCQ and Ofqual, and have done with it: I’m getting a headache.

JCQ Joint Council for Qualifications: Examination results
Each year, JCQCIC collates the collective results for its members from more than 26 million scripts and items of coursework. We only publish collated results from our members though and cannot supply regional, centre or candidate information.

Ofqual Analytics
Ofqual analytics presents a selection of data in an engaging and accessible way by using interactive visualisations. We hope this innovative approach to presenting data will make it easier to understand and explore the data we produce.

Map of GCSE (9 to 1) grade outcomes by county in England
The map shows reformed GCSE full course results (the percentage of students achieving specific grades) in England by subject and county for the summer 2018 examination series as well as the summer 2017 examination series. Data in the map represents the results that were issued on results day for both years (23 August 2018 and 24 August 2017) and do not reflect any changes following post-results services.

happy-statistics-day-2

(All I know is that my boy got his GCSE results today too, and they’re a credit to the amount of time and effort he’s put in over the years.)

100,000 happy moments

Nathan Yau has a fascinating look at what makes us happy.

What makes people the most happy
What made you happy in the past 24 hours? Researchers asked 10,000 people this question. More specifically, the collaboration between the University of Tokyo, MIT, and Recruit Institute of Technology asked participants on Mechanical Turk to list 10 happy moments. This generated a corpus of 100,000 happy moments called HappyDB.

With how things are these days, I was happy to read over and analyze such a happy dataset.

All strapped in?

Nearly 16% of US consumers now own wearables
“Fitness bands continue to outsell more advanced smartwatches,” reported Lauren Guenveur, Consumer Insight Director for Kantar Worldpanel ComTech. “In the fourth quarter of 2016, just 35% of wearables purchased in the US were smartwatches, a decline from 40% in the third quarter of 2016.

Well, I’m very happy with my fitness band, though it feels like we’ve all voluntarily bought electronic tags like a load of criminals.

All the graphs you can eat

What’s better than data and loads of graphs? Data and loads of graphs about food and drink, of course!

Britain’s diet in data
The British diet has undergone a transformation in the last half-century. Traditional staples such as eggs, potatoes and butter have gradually given way to more exotic or convenient foods such as aubergines, olive oil and stir-fry packs. Explore the changes across four decades and hundreds of food and drink categories in this interactive visualisation, featuring data from the Department for Environment, Food and Rural Affairs (Defra).

Whilst I’m not surprised to see the fall of lard and the rise of olive oil over the last 40 years, why is nobody buying marmalade anymore? Goodness me.

Visualising data; the good, the bad and the quirky

The data’s everywhere, but can we make sense of it? Here are some data visualisation approaches and examples – how to do it, and how definitely not to.

This article suggests the young folks are demanding better presentations of statistical data because of Google or Wikipedia or something. Sounds fishy. Surely us old folks appreciate good design too?

Data visualization drives the era of information activism
Having grown up with the web, millennials are used to having access to all the information they want with just a simple finger tap on a screen. As millennials enter the workforce, they are bringing these expectations into the office, behaving less as data consumers and much more as information activists.

But how far should we go in leading the horses to the water?

Narration and exploration in visualization
What should we emphasize when designing a visualization? Should we explain the data, perhaps through a narration, or should we let readers explore the data at will?

Here’s an example of how an interactive presentation not only helps with a story but can spur you into looking for your own.

What happens when a newspaper editor and a data-viz whiz team up to tell stories
With two wins behind them, Rob and Daniel are already discussing future collaborations—perhaps something with sports or crime data. The idea is to tell compelling data stories that have a longer shelf life. (“If you put time into creating something like this, you want people to see it over a certain scale of time,” says Daniel.)

I do like these images, though. Not sure how replicable this approach is, but it works well here.

These playful, funny 3D printed infographics can liven up any data
Instead of flicking hastily to the next page, your gaze can’t help but linger on the adventures of the miniature figures scaling, swinging from or exploring the plastic pieces.

But here are the real gems. Bar pies? Caramel latte football fields?

BestWorstViz contest result
As with last month’s graphiti contest, picking a winner was ever so hard with lots of wonderfully bad work heading my way. The submissions generally fell into two different camps: (1) the most blatant, inelegant in-your-face explosions of design awfulness, and (2) the more subtly deceptive “wolf in sheep’s clothing” designs. My judgments were based on the degree of violation against each of my three key principles: Trustworthy, accessible and elegant.

Still quantifying ourselves?

daytumapp

Didn’t realise I’d been away from Daytum for so long.

For those that don’t know, and why should you, bloody hell, Daytum was part of a trend on the web some time back for personal data tracking, the ‘quantified self’. Here’s a post from Wired about it, from 2009:

Numbers are making their way into the smallest crevices of our lives. We have pedometers in the soles of our shoes and phones that can post our location as we move around town. We can tweet what we eat into a database and subscribe to Web services that track our finances. There are sites and programs for monitoring mood, pain, blood sugar, blood pressure, heart rate, cognitive alacrity, menstruation, and prayers. […] All this might once have seemed like a nightmare […] But two years ago, my fellow Wired writer Kevin Kelly and I noticed that many of our acquaintances were beginning to do this terrible thing to themselves, finding clever ways to extract streams of numbers from ordinary human activities.

– Gary Wolf, Know Thyself: Tracking Every Facet of Life, from Sleep to Mood to Pain, 24/7/365 (wired.com)

Daytum, developed by Nicholas Feltron (of the well known Feltron reports) and Ryan Case, Daytum allows you to easily capture and categorise small bits of information, and then present these in quite interesting, visual ways with the aim of providing a little insight into how we live our lives. For a while, all the way back in 2009, I was tracking how much coffee I got through, what I was reading, which were my favourite ties, that kind of thing.

I don’t know what made me think of Daytum again, but I thought I’d catch up with it again.

Its blog, however, ends with this post from April 11, Moving West:

We’re thrilled to announce today that we just started a new phase of our careers: we’ve moved to California to join the product design team at Facebook […]

and the last tweet from the @daytum Twitter account is:

Doesn’t bode well, but then, just the other week coincidentally, this one from founder Nicholas Feltron:

So perhaps there might be some life left in it yet?

And I’ve forgotten all about your.flowingdata.com too. But that, like Daytum and much of the quantifying self scene, seems to have gone a little quiet.

FE and HE bureaucracy and data

Red tape costs FE colleges £180 million a year
Further education colleges could make “substantial” savings by cutting the costs of bureaucracy, according to the National Audit Office. The body said today that colleges spend around £180 million a year on administrating funding, qualification and assurance systems. This amounts to £150 per student.

Adventures in Student Development
A while back I explained how I set up the system, and after nearly 2 years of use I can say that it’s still working.

HESA Press release 168 – What is a Course?
HESA, working with representatives from the sector, is pleased to publish a brief report that considers the variety of interpretations and models that exist across the sector. The report analyses how the concept of a course is represented in various sector-level information systems and specifications and how the definition of course changes across the Higher Education lifecycle.

GTD, stats, fees

New GTD Setup Guide for Outlook 2010
For those of you on Outlook 2010, we just released a new Setup Guide to assist you in creating a rock-solid GTD system in Outlook. Since the 2010 version changed some ways things are done in Outlook, we created a new Guide specific to this version.

Correlation or Causation?
Need to prove something you already believe? Statistics are easy: All you need are two graphs and a leading question. Correlation may not imply causation, but it sure can help us insinuate it.

Tuition fees could bring bonanza for humanities
If the same fees apply for all subjects, humanities departments may benefit, says Jonathan Wolff – but creative arts and some social sciences may suffer.

HE fall, environmental data

Figures suggest that the fall in the number applying to university is mostly owing to a glut of applications in 2010
Wendy Piatt, director general of the Russell Group, which represents the UK’s 20 leading universities including Oxford and Cambridge, said it was too early to predict how many students would end up at university next autumn. She said it was unfair to compare the number of applicants for next autumn’s courses with those for this year’s because the previous figures may have been artificially inflated by students applying before the near-trebling of fees came into effect. “Current 2012 figures are actually very similar to figures at the same point in 2010,” Piatt said.

Information is Beautiful on the Thailand floods
Floods. Amazon deforestation. Earthquake destruction. Satellite maps somehow don’t always help us to fully imagine the size of these disasters. Is there a better way to visualize the scale of destruction? Here I’ve been playing with the ranges of various natural and unnatural disasters, pulling data from various media reports and the US Geological Survey.

Worrying statistics?

Worrying statistics

Documenting the curious increase in claimed family deaths — especially of grandmothers — during tests season at college
This gem is from “The Dead Grandmother/Exam Syndrome and the Potential Downfall Of American Society” by Mike Adams (The Connecticut Review, 1990). Adams’ hilarious explanation for this phenomenon:

“Only one conclusion can be drawn from these data. Family members literally worry themselves to death over the outcome of their relatives’ performance on each exam. Naturally, the worse the student’s record is, and the more important the exam, the more the family worries; and it is the ensuing tension that presumably causes premature death.”
www.easternct.edu/~adams/Resources/Grannies.pdf

Choosing a university

bestCourse4me.com – Choose the right course at the right university
Totally independent and free to use, bestCourse4me shows you the link between what people study and their employment record afterwards. … Spearheaded by the Shadow higher education spokesman David Willetts and bankrolled by Microsoft and the philanthropist and software engineer Steve Edwards.

New website gives wannabe students vital information about drop-out rates and earnings
At the flick of a switch you can find out what you earn, for example, if you study physics as against English (about £4,000 more a year, in case you wondered) and the employment rates of graduates of those subjects. You can discover how much more you earn if you study journalism at Westminster University compared with the University of Central Lancashire or philosophy at Durham compared with Bristol, and you can find out the drop-out rate from all institutions.

Feedback

Rosie Waterhouse: Save us from the tyranny of student surveys
Over the next three months, final-year undergraduates at British universities will be encouraged to take part in the National Student Survey commissioned by the Higher Education Funding Council. The NSS is just one of an ever-increasing number of surveys and forums by which students are invited to give “feedback” on their university experience.