Managing A Quality Service Flashcards
Situation:
Working as a data scientist…
…at the Office for National Statistics.
Produce an urgent, comprehensive analysis on the effect of the lifting of the…
…work from home order on mobility flows across the UK, for the Cabinet Office, which was required to be completed in three days.
Working with Cabinet Office colleagues to understand the true nature of the task, I established that there were a diverse range of stakeholders, both technical and non-technical, and I understood that the complex data had..
…to be simplified in order to communicate the insights to these stakeholders and if necessary to a wider political sphere.
The data being analyzed was proprietary mobile phone data that had been supplied to our government directorate and this analysis was one of the first to leverage data of this kind to answer a policy question as unique as this.
There were four main stages to service delivery: processing that raw data, satisfying the data privacy requirements, conducting the analysis, creating the output and presenting the data product. I was responsible for conducting the analysis…
….creating the output and satisfying the data privacy requirements.
I coordinated with the team members who were conducting the processing of that raw data and the team who were responsible for maintaining data security and we agreed to complete tasks as much as we could in parallel to each other to ensure there were as few bottlenecks as possible.
I arranged daily meetings with members from each team to “peer review” our work to ensure that outputs were validated and robust, and also to provide feedback on the quality of the work.
I reviewed the BigQuery views that were preprocessing the data and my dashboard visualizations were reviewed and feedback was given and acted on.
Rather than a static PowerPoint, I decided to produce an interactive dashboard on Google Data Studio, which at the time was not commonly used in government,.
I anticipated that this analysis would be monitored going forward when new data became available, so rather than create a PowerPoint of the insights that would have to be recreated every time new data was available I chose to create an automated dashboard.
Using a dashboard also meant that the stakeholder would not rely on contacting us and re-running the analysis every time they wanted insights from the data…
…which would have led to a delay in critical decision making.
I started work on the dashboard itself and I carefully selected the best graphics for visualizing that data so that the Cabinet Office could easily identify and analyze trends in the data.
I produced a clear visualization of how mobility trends had been impacted by the change of policy, using labels on the graphs to highlight areas of interest and describe the trends that I was visualizing.
I produced an executive summary to highlight the main insights of the dashboard to make it accessible for non-technical stakeholders and wrote clear descriptions of the methodology being used.
I chose Google Data Studio as it was native to our team’s current cloud provider, GCP, which meant that it easily integrated…
…with Google Big Query, where the data and views processing the data were written.
Moreover, Google Data Studio was easily accessible to the Cabinet Office team, by simply sharing a link.
The data that was being used in the analysis had a strict data sharing agreement in place with the original provider, which limited the sharing of row level data.
In order to meet the stipulations of this agreement, but still share the insights of the analysis with the Cabinet Office,…
…I held meetings with the cloud engineers in the Cloud Architecture team to fully understand the operation of the security measures available in Google Cloud Platform.
I created bespoke security groups for Cabinet Office officials, within our cloud environment, which restricted their level of access at a data level but allowed them to view the output of the dashboard.
The Cabinet Office was able to fully understand the effect of the work from home order.
The information displayed on the dashboard is regularly updated through a data pipeline that gives the Cabinet Office…
…an up to date picture of mobility trends in the UK which is used to contribute to the evidence base for decision making within the Cabinet Office.
