I’ve put together a spreadsheet of Google Mobility data aggregated for Australia from 15 Feb 2020 until 3 April 2022. I will update it sporadically, but in the meantime it’s interesting to play with. There are several caveats around this data. One is that the baseline used is January 2020, which makes sense for the US, but not so much for Australia, as January is summer holiday time for many of us. This means that the absolute “difference from baseline” isn’t that useful, but you can cheerfully compare months across the dataset.
Apple’s mobility data is based on a day from midnight to midnight, US Pacific Standard time, which is not a day in Australian time, which is another thing to be wary of. I don’t know how Google calculates their day.
The spreadsheet is here as a google doc. Please note that its use is subject to the Google Terms of Service, as the original data came from Google.
What can you see in the data? I can see, for example, big drops in retail activity, but many of them correspond to things we already know about, such as Christmas Day when most stores are closed.
There are so many potential explorations with a dataset this rich. Compare month to month, lockdown to lockdown, local government area to local government area…
Now including all states and territories!
There is so much to explore in this dataset, it has great potential for an open ended classroom exercise. Download it as a csv and play with it in Python. Make graphs in Excel/Numbers/Google Sheets and consider readability, labelling, axes, etc. Compare different parts of the data. Share back the ways you explore this data – either to email@example.com or in the Teachers Using Data Science Facebook group and I will post them here.
If you’d like to fund ADSEI to build full lesson plans with rubrics, curriculum links, and resources, get in touch!