Here are some links to some of my best recent work:
Weekly print pages
The Reach data unit was tasked with providing data-led print pages for 34 of our weekly titles. Every week we provide localised pages about topics such as homelessness, burglaries and mental health.
I used R to automate this process. My R scripts analyse millions of rows of data in minutes to find the data relevant to our titles. Functions I wrote create the text and the relative sizes of the charts. R prints out a CSV and TSV file that our developer feeds into the Adobe suite to create the beautiful pages you see above.
School absence and deprivation
I analysed the Government’s school absence data and combined it with the latest Index of Multiple Deprivation to see whether pupils who go to school in poorer areas were more likely to miss school without their parents’ permission.
I used R to analyse the hundreds of thousands of rows of raw data.
My hunch was correct: poorer pupils were more likely to skip school than their richer peers. This lack of attendance won’t just be truancy – some pupils won’t be at school because of their chaotic family lives, according to two psychologists I spoke to for the piece.
This app allows the user to type in their postcode to see how bad their area is for burglary. It shows you whether burglary is going up or down and identifies the problem locations in your neighbourhood.
To create this project I worked with the designer and developer on my team to analyse more than one million rows of crime data.
Hundreds of thousands of people have used this widget.
Cyclists hitting car doors
I analysed millions of rows of the Stats19 data from the Department for Transport using R. These are taken from police records of road accidents that involve at least one person getting injured or killed up and down Britain.
I narrowed it down to cyclists getting injured or killed by colliding with car doors. Drivers can get very careless when opening car doors, not paying attention to oncoming cyclists.
The charity Cycling UK advocates using the so-called ‘Dutch reach’ – using your opposite hand to open a car door, which forces your body to turn and increases the likelihood of spotting a cyclist either directly or in your wing mirror.
Tumble dryer fires
Working with Martin Bagot and Nicola Bartlett at the Mirror, I asked fire departments for the number of fires caused by tumble dryers. Within just a few days, we had collated data from most of the fire services in Britain to give a national picture of this dangerous trend – as well as numerous stories for our regional titles.
Where Should I Move To?
This widget caters to those considering leaving the UK in the aftermath of the Leave vote in the EU referendum. Select which criteria for living in a foreign country are important to you and the widget shows you which other EU country best matches your choices. I sourced the data that underpins the gadget and worked closely with our developer, Carlos Novoa, and our designer, Dmitri Thompson on the team to get it up quickly on the Monday following the result.
Property ultimately owned by offshore companies is huge business in Britain. I analysed the database collated by Private Eye to reveal the multi-billion scale of the market – and also to pick out some prominent examples in Liverpool.
I used Tableau to go into the Land Registry’s house price data, looking for flats sold in the Beetham Tower, Manchester – by far the tallest building in the city. I was able to show that it costs about £280,000 to buy a flat there. This story was viewed more than 100,000 times.
Further analysis of the Land Registry data showed that there was a puzzling lack of homes sold at the number 13. Was it superstition on behalf of the buyers – or the developers in the first place?
Smoking in cars
I resolved to investigate crimes resulting from the law change banning smoking in cars with children present. After months of digging using Freedom of Information requests, I was able to expose a major loophole in the law: the ban on smoking in cars was effectively going unenforced. Police and local authorities simply didn’t have the systems in place adequately to enforce the new law.
Robots taking your job
I used the results of an American study into job losses to automation and cross-referenced them with the ONS estimates of the numbers of people in different types of jobs. The findings were stark – in Merseyside, about 23 per cent of people work in the jobs most at risk. How humanity handles automation will be one of the biggest challenges of the 21st century.