Centring your highlighted jitter point
Week 19 of Makeover Monday in 2018 challenged people to visualise the toughest sports, according to survey data conducted by ESPN. With the data being sports related, I followed the finished makeovers with interest and found many people turning to jitter plots to visualise the data.
Whilst jitter plots are a great way to visualise survey data, hey, if they are good enough for Chapter 3 of 'The Big Book of Dashboards' they are good enough for me, there is a little technique I like to apply to jitter plots to make them really pop. When wishing to highlight a chosen data point I feel it is best to centre your highlighted jitter point. This technique has the following benefits:
1. When highlighting a particular point, that point should be the focus of your visualisation. The middle of your jitter plot is naturally where most people will initially look when interacting with the viz.
2. In the example of the Makeover Monday Week 19 data set each sport was rated across a number of different attributes. If you create multiple jitter plots, one for each attribute, and use the standard Random() jitter method, each highlighted jitter point is likely to be positioned differently across each plot. The Random() function will calculate independently for each worksheet, for example jitter 1 may place the highlighted point at 0.1 whilst jitter 2 may place it anywhere between 0 and 1. By using the centring method you are ensuring your highlighted point will be consistently placed in the middle of the axis across all of your jitter worksheets.
How to centre your highlighted jitter point
Centring your chosen jitter point is simple. For example if, using the Makeover Monday Week 19 data set, you wished to highlight and centre data relating to boxing, the toughest sport, you would create the following formula:
[Jitter] : If [Sport] = 'Boxing' then 0.5 else Random() end
This differs from a normal jitter plot calculation which uses:
[Jitter] : Random()
Of course you can replace the hard-coded If statement with one which uses a parameter to enable your viewers to select the value they wish to highlight:
[Jitter] : If [Sport] = [Sport Parameter] then 0.5 else Random() end
Don't forget to fix your axis range to be from 0 to 1!
My favourite example of this technique in action
When it comes to the jitter plots created during this week's Makeover Monday task, there was one visualisation that stood out for me; 'Float Like a Butterfly, Sting Like a Bee' by András Szesztai. He used many visualisation techniques that, when added together, created an outstanding viz:
- His viz allowed people to select a sport to highlight; not only was the highlighted sport presented using a different colour but it was also sized sufficiently large enough to include the value as a label within the highlighted circle.
- There were no individual axis on the 10 jitter plots, instead András cleverly created 2 master axis to the right of the viz, such a clean look, in my opinion.
- The viz was 'framed' using a background image; with boxing being the toughest sport, András included a boxing background image that blended seamlessly into the viz and was the foundation of a very sophisticated and consistent colour theme.
My Makeover Monday visualisation this week
For my visualisation using the Makeover Monday Week 19 data set I went with a play on words based on boxing being the toughest sport, 'In the red corner, in the blue corner' and using a red-blue diverging colour palette to create a visual heat map of sports and individual sporting attributes.