analysis and use of evidence

TLDR
A full-day workshop for teams who need to turn evidence into sound policy advice. Your group learns to find and choose the right sources, run the checks that make evidence reliable, weigh what it says and how confidently, and shape it into a recommendation a senior decision maker can act on. Practical throughout, with AI tools built into every stage.

tell me more - email
tell me more - form

Good policy advice stands or falls on the evidence behind it. This programme gives your group a shared foundation in analysing and using evidence: finding the right sources for the question, testing them before they go into a case, weighing what they say and turning them into a recommendation that a senior decision maker or an external body can rely on.

AI tools now sit inside every stage of the work, from searching to summarising. Your group learns where they add speed, how to check what they produce and what can safely go into them, so AI-assisted evidence meets the same standard as every other source.

Compliance and rigour are part of the method. Participants work through data protection, privacy, data sharing, confidentiality and licensing requirements as they gather and handle evidence, and build an audit trail showing where each piece came from and who may see it.

Every participant leaves with a decision-ready evidence summary on a real question for their group. It sets out what is known, what is not and how confident the group is, and it carries a clear recommendation.

The programme is adapted to the type of work most relevant to your group now and delivered in line with your team’s and organisation’s values, code of ethics and tone guide.

    • Identify and gather evidence from the right range of sources for the question at hand, including compliant AI tools

    • Test evidence for reliability, bias and gaps before it goes into a case, including where it is AI-generated

    • Weigh conflicting or incomplete evidence and state clearly where confidence is low

    • Handle evidence and data in line with data protection, privacy and other current requirements, and synthesise it into a decision-ready evidence base

    • Turn that evidence base into a clear policy recommendation a senior decision maker can act

    • Evidence bases that senior decision makers and external bodies can rely on

    • A shared standard across your group for what counts as good evidence

    • Clear statements of confidence that make advice easier to act on and to defend

    • Confident use of AI tools in research, with outputs your group has checked

    • Recommendations that trace back to their sources, with an audit trail to match

  • Discovery: A conversation with you and a short pre-workshop questionnaire. It covers the questions your group works on, the sources it uses today, how it uses AI tools and the data and compliance requirements that apply. The results shape the case studies and checklists used on the day.

    Full-day core workshop: In person or virtual. Every session is built around practice, role play and habit building.

  • 09:30 – 09:45 | Welcome and expectations
    Introductions and aims for the day. Participants add to an expectations wall what they want to be able to do with evidence by the end of the session, and each group names the policy question it will carry through the day.

    09:45 – 10:15 | Evidence myths and mindsets
    Groups sort common beliefs about evidence into myth and truth, for example that published means reliable, that more sources mean stronger evidence or that AI summaries are neutral. They pick the two most active in their work and agree the habit to practise through the day.

    10:15 – 11:15 | From question to sources
    Groups turn their question into the decision it informs and the evidence that would answer it. They map the range of sources on post-its, from official data and published research to practitioner and stakeholder evidence and AI-assisted search, cluster them and choose by fit to the question. Participants practise prompting AI tools to find sources and log what each returns.

    11:15 – 11:30 | Break

    11:30 – 12:45 | Testing evidence
    Groups work through a pack of sources, including an AI-generated summary, and run a reliability checklist covering provenance, method, sample, currency, independence and consistency. They cluster the bias and gaps they find, check each AI claim against its cited source and agree which evidence earns a place in the case.

    12:45 – 13:30 | Lunch

    13:30 – 14:15 | Weighing the evidence
    Groups place their evidence on a grid of reliability against relevance and work through the conflicting and incomplete findings. Participants practise writing confidence statements of high, medium or low, and state plainly where confidence is low and what would change it.

    14:15 – 14:55 | Handling evidence and data
    Participants work through a compliance checklist covering data protection and privacy, data sharing, licensing, confidentiality and what can go into AI tools, and note who to involve at each point. In a short role play they respond to a colleague who wants to paste a dataset into an AI tool. Each group starts an audit trail recording where each piece of evidence came from and who may see it.

    14:55 – 15:10 | Break

    15:10 – 16:00 | Evidence into recommendation
    Each group synthesises its evidence into a one-page decision-ready summary: what we know, what we don’t and how confident we are. They turn it into options and a recommendation, take peer feedback in “I like / I wish / What if” rounds, refine it and defend it to a challenge panel.

    16:00 – 16:30 | Commitments and close
    Each participant writes a pledge card with one change in how they find, test or present evidence. The group agrees the shared evidence standard and checklist it will use in its work, revisits the expectations wall and closes with questions.

Communicate - a Teamshaper training module