Skip to main content

Insight, quickly: analytical tool for plant health

Trees support biodiversity and contribute to wellbeing – a new analytical tool helps us monitor and safeguard them against spread of diseases.

As a statistician working in Plant Health for Defra, my role is to provide high quality statistics, analysis and advice to help us make better decisions. In my team, we have been working with our stakeholders to bring analytical insight to critical decision making that helps us manage or eradicate pests and diseases. These decisions take into account information such as:

  • potential spread of the pest or disease;
  • impact of different actions on spread; and
  • costs associated with taking action.

We want to improve early identification of tree diseases, so we can plan our response and assess interventions against policy objectives.

To help with this we developed a new analytical tool.

Designing the tool

We had two key criteria to assess our needs for the tool whether:

  • it would help us work differently
  • the benefits would outweigh the cost and effort

It quickly became clear that a new tool would help us work differently, by being able to respond much more quickly. At the moment, our analytics comes from bespoke models and reports which take time to produce. Besides producing the current reports, the tool will also allow us to quickly model spread, impact, cost and resource over time.

The second is a little trickier to answer. Once we are engaged in the development process, there is no guarantee that things will go to plan. What we can do is factor in some resource, estimate some costs, and allow for some contingency – and monitor progress during development

The devil is in the detail

We knew the tool could be transformative, so in spite of our concerns about the challenge of delivery, we pressed ahead. As statisticians, we came into our own, using our skills to identify the scope of the tool, expected deliverables, limitations and strengths; we anticipated the data that might feed into analysis, and what we might expect to emerge.

There are three key things that I have learned from working on this project:

  1. It’s important not to jump too far ahead – but to keep considering who will use the tool, and for what purpose.
  2. We need to engage with technology choices. There are rarely any right answers but, we can’t, and shouldn’t, ignore the vast amount of new and shiny tools we can use to realise our vision.
  3. We need to think of the short-term as well as long-term, recognising that some tools and technologies become obsolete very quickly and others have security implications; some are not easy to transfer across systems and agencies or just require too much cost and resource to get up and running.

We need to retain the core statistical principles: being clear about what the data produced means; and what level of confidence users can/should have in using it.

Statisticians in government

Reflecting on what it means to be a statistician in government, the data and analysis landscape has changed immeasurably over the past few years, but our overarching professional principles still stand. We can take our ‘old’ skills of manipulating data in spreadsheets and generating statistical findings, and combine them with ‘new skills’ such as coding and visualisation to present data and analysis in more exciting ways. Some of our ‘old skills’ are those that are most important in our role – ensuring statistics are based on sound principles, quality assured and robust – and these won’t change.

For me, applying our statistical skills in this new context comes down to embracing and not feeling overwhelmed by the new directions in which data and data analysis in government are heading – meeting the civil service competencies of seeing the bigger picture, changing and improving, and providing value for money.

Those considerations are very much underpinned by high quality statistics, analysis and providing the best advice to help us make better decisions.


Sharing and comments

Share this page