Our last Makeover Monday challenge conveniently falls on Boxing Day, so the topic being Christmas dinner works just fine for me. The data is from the BBC and highlights the rising costs of a Christmas dinner in the UK.
The accompanying article discusses the change in costs over time and provides the following visualisations:
What I like about the original visualisations:
- it’s an interesting topic that many people can relate to, so it’s a great opportunity to get people interested in data and data visualisation
- a line chart is a good choice for showing data points over time
- the horizontal bar chart focuses on this year versus the prior year which is probably most relevant for most people so it’s a good way to simplify the data to tell the story
- I like the cost per person break down of the bar chart, because again it simplifies things plus it gives people an easy way to estimate their own dinner, no matter how many people they’ve invited
- the subheadings are clear for both charts, so anyone reading them should know what they’re looking at
- sources are clearly stated below the charts
What I don’t like:
- the y-axis for the line chart should have a much larger range in my opinion. The dramatic looking rise in dinner costs is due to a distortion with the y-axis starting at 50. It doesn’t necessarily need to start at zero, but a wider range would give a more realistic picture.
- the heading for the line chart isn’t great. ‘costs rising again’ to me suggests they have been rising for at least the last couple of years but the chart suggests they dropped since 2012 and only rose this year
- the smoothed line of the line chart creates data points where there shouldn’t be any. The line should only connect the actual data points…
- the colour choice of the bar chart is unfortunate. I get the whole red and green Christmas theme but those two colours together are really hard to look at. Even if you’re not colour blind, I find it unpleasant to look at, the hues are too intense and bright. A change of opacity could have helped here.
What I did:
Before even starting to visualise data, I had to do a bit of research into these UK Christmas dinners. Growing up in Germany (cabbage country!), our Christmas dinners looked quite different and certainly didn’t feature Brussels sprouts or anything with puff pastry. Living in New Zealand and Australia where Christmas is in summer had me consume either a summery BBQ feast or a slightly identify-crisis suffering heavy European style meal.
Either way, I had to learn about UK Christmas dinner traditions first. This helped me categorise the different foods in the dataset into categories. I’m still trying to figure out how mince pies fit into the menu, but maybe life is more exciting with a few mysteries… For now I’ll treat them as part of the main dish.
Following my initial research, I started doing some iterations with the dinner categories, looking at costs over time, etc.
I usually avoid looking at other Makeovers before publishing mine, but with all the other conversations on Twitter, I couldn’t help but see some of the work by my Tableau friends. This led me to looking more closely at the different items included in the dataset. One thing I noticed was that there was fresh and frozen turkey as well as ham. Surely for a dinner party of 6 you wouldn’t prepare all three? And then there is white wine, red wine, beer and whisky. Again, would you really have all of those??? Maybe I am really out of touch with the foodie scene…
Anyway, after playing around with the data my story started to form and I decided to point out that yes, overall costs have gone up, but the type of dinner (turkey vs ham) and the type of alcohol people serve has the biggest impact on their budget.
So without further ado, here is my little dashboard… For the interactive version, just click on the image
I tried to keep things simple, clean and with a somewhat Christmas inspired colour scheme…
In other news…
Now with the last Makeover Monday challenge of 2016 complete, I am very excited as we head into 2017, because I’m teaming up with Andy Kriebel for a year of makeovers and new data sets…