R is an open-source software for data and statistical modelling software. StatsNZ uses R to process and analyse data for their statistical and business outputs. Since the introduction of R into Stats NZ, the number of has grown from a small set of users to around 200 frequent users and many more that have access to the system.
The Statistical Infrastructure team at Stats NZ wanted to better understand these users, how they used the tools and for what purposes, what issues they faced, and their ongoing resource needs – this detail would then fed into the next stage of planning around infrastructure and identifying key areas of focus to support these users.
The project: Identify R user personnas to enable effective generation of statistical outputs and products.
To get a summary view that would be helpful for the next steps, Standard of Proof worked together with the infrastructure team to put in place the following plan:
- Focus group workshops: Based on qualitative sampling framework 8 workshops were run with participants across the organisation. This framework aimed to achieve maximum variation in the perspectives gathered while also approaching saturation—the point at which little new information is being put forward.
- A survey to all R users in the organisation: Using the knowledge gathered through the workshops, a survey was created to query users on their specific use cases and resource needs for R software and related tools.
- A report was prepared summarising the different user personas, their specific use cases along with quantitative feedback from the survey.
To work effectively together with Stats NZ we caught up on a weekly basis to share ideas, set timelines and update on progress. The direct end users of the report were also engaged in the process as we got closer to the finished product.
The final report identified a range of user ‘personas’, their resource needs and perspectives and we were able to describe typical user personas use cases and themes and issues across the user base. The report married together the insights from the workshops with quantitative results from the survey, both sets of data being used to inform the design of a technical infrastructure platform (and environments/policies) that supports and enables effective generation of statistical outputs and products.
We are proud to have supported StatsNZ on such an interesting and challenging project, combining qualitative and quantitative methods within a deeply technical domain across multiple business contexts.
