In this blog post I’ll be giving you a brief walkthrough of my initial application to The Data School, on a macro level. I’ll mainly be going over my thought and design process. This post is mainly aimed at prospective applicants for future cohorts. I hope you learn a few things.
What my application was about
My first application was about stop and search in England and Wales and I found the dataset on the Government website. It took a bit of mulling over before settling on this dataset. My tip to you would be to pick a topic you genuinely care about and have an interest in. It’ll really show in your work. The data could literally be about anything you’d like.
Exploring the data
One of the first things I did was to explore the data. This included making a few simple bar charts e.g. the total number of stop and searches over time. It allowed me to gain familiarity and a feel for the dataset (I feel like this is sometimes overlooked). In addition, it let me see if there were any data points of particular interest from the get-go. This is the beauty of Tableau, it enables you to carry out relatively quick analysis and exploration.
Another advantage to exploring the data is that it can give you inspiration to ask the ‘right’ questions. This leads onto the next section.
Asking the questions
Throughout my dashboard (which was in 3 sections), I had three questions in general:
- Who/which police force is carrying out most the stop and searches?
- How many have been carried out over time?
- Who is being stopped and searched?
It’s subtle but notice how the scope narrows down the further you go in the questions. This was a good aspect as it allowed me to come up with a presentation which had a flow, made sense and was like a ‘story’.
When making a dashboard, one thing you could ask yourself is: “What questions do I want to be answered?”. If it helps, let’s look back to my very first blog post that I published about Andy’s Data Viz 101 session:
... we were taught to ask the data 6 simple questions (in this order specifically): 1. When? 2. Where? 3. What? 4. Who? 5. How? 6. Why?
Answering the questions
As a tip for answering the questions that you ask, you may find information online. In my case, I found that there were quite a few changes in policing, e.g. policies, which caused the drop in numbers of stop and searches. I’d say in general make sure that you do some research around your topic; see if you can find things that can enrich your dashboard (though not necessarily supplementary data).
For one reason or another, there will most likely be a scenario where you’ll have to redo your dashboard(s). Whether coming at them from a different angle in terms of the analysis or a different chart type. You may even make a chart and nothing useful comes out of it. Just know that it’s all perfectly OK and normal. It happens to all of us. The journey of creating your application (or any dashboards in general) might look like the diagram below:
P.S Thanks to Luke for reminding me of a similar diagram during training. To quote his PowerPoint slide:
It’s an iterative process!
For my application, I didn’t originally plan to focus on the Metropolitan Police. It was only after doing some initial research and exploration that I decided on this aspect. Moreover, I found out that stop and searches peaked around 2008 and then dropped off in subsequent years. I thought to myself:
As a result, this changed the type of chart that I used. A third example is that I decided to focus much more comparing the ethnicities of the people being stopped and searched. This was because I noticed that there was a big disparity in the rate that people of Black ethnicity were being stopped and searched.
When writing it down and reflecting it may sound simple but, trust me, the whole process was like the squiggly rollercoaster above.
Getting feedback from others
This is a very crucial aspect. It can be your siblings, people at The Data School, or your friends – whoever. This is a way to get a new perspective. With any work, if you use or look at it long enough you’ll develop blind spots; others may give you feedback on things you would have never considered before. Remember, even if it makes sense to you it might not make sense to others.
Throughout the whole process I made a lot of changes to my application and I owe credit to those who I asked for opinions from. One example is that I originally had a treemap on the first page. This was changed to a pareto chart after some feedback.
Final tips and thoughts:
- Sketch out your designs
- Be flexible and be open to suggestions
- Complex does not always mean good, vice versa
Although this isn’t the full picture, I hope it gave you some good insights. For anybody reading this who will be applying or is in the process of applying, good luck!
Thumbnail image by Ian Schneider