Visualizing the analytics of chupacabras story production
We have arrived my friends! It’s spooky season!! For one weekend of the year, the entire country takes their costumes off and we get to see who they really are on the inside…
Or is that the other way around…
In any case, one of the most frightening things in data visualization is a bland chart. You never want something drab or uninteresting telling your story for your client, your boss, your family member… for whomever you’re designing this chart, really, because you want to make sure that 1) the message gets across, and 2) everyone has a little fun on the way. So, why pick a control chart?
First of all, what is a control chart and why is it important? Control charts are just one visualization that comes out of the #projectmanagement field which tell us a number of critical items. Firstly, control charts explain the (sometimes fallacious) idea that there is ‘control’, or, more specifically, a parameter within which a process can be considered ‘in control’. Consider — for example — a factory. We know that production of widgets should be happening at a certain rate per hour. This can also be applicable in a construction environment: workers being able to install a given amount of linear feet of plumbing, or laying an amount of bricks. These time calculations allow us to make appropriate time estimates (and by extension, including wage rates/fringe, materials pricing, and insurance costs -amongst other things,- budget estimates) in order to make sure we complete a project on time, and on budget. You might also include them in a #cybersecurity / #infosec environment: how many #pentest or other attempts were made in a certain time period. How many were successful? How many are ‘within normal range’…?
A control chart has to have the following critical features:
- You need a control band. This you need to hammer out with your project/production team to figure out what’s ‘normal.’ Normal is a funny word in 2020, but, put some thought into it and decide on the necessary band width.
- Determine how many times you’re seeing processes ‘out of control’. If you’ve ever wondered at the provenance of the expression, now you know. It means something that costs, has a time duration, or in some other measurement doesn’t fit inside the normal parameter (consider the entire battle over BMI and those metrics. Excellent reason to re-examine whether your control parameters were a correct assumption,) is out of control. Some variation is expected, and even warranted. But if a process/product is outside the parameters >/= 7x, it’s out of control.
- Explain what happened. This is a great spot to make liberal use of the Tool Tip’s generous functionality (I have call outs in the viz on purpose.) But make sure your project/production team understands the 5W1H of the out of control item. This will let you know if you need to change a process, readjust the control band, or stop what you’re doing altogether.
After that it’s just making the chart look visually appealing and telling the story effectively. Have you used control charts in your daily work? Have any interesting stories about how they changed what you were doing? Use them for your own personal data analytics? I’d love to hear your story!
Link to the live chart is here.
