Where we go wrong

  1. Identify & engage decision makers
  2. Have a clear question
    • Who is going to do the analysis?
    • Ensure there is budget for it
  3. Simulate & analyse some dummy data
    • Test & minimise survey questions (survey fatigue)
  4. Decide on collection & safeguarding approach
  5. Write up & record methods
  6. Train staff (collection & analysis)
    • Keep the raw data (flexibility)

Where we go wrong

  1. Ensure the data is collected in standardised way
    • Record absences (zero is a number)
    • Keep data tidy (machine readable)
  2. Analyse & safeguard the data
    • Guided by your organisational data strategy
  3. Exploit the pattern
    • Improve “best practice”
    • Share your knowledge (audience appropriate)
  4. Influence decision makers
    • Promote the evidence
    • Change the world

Automate data collection

Use a tidy format

  1. Every column is a variable (measurement)
  2. Every row is a single observation
  3. One value per cell

Untidy

Tidy

Where we go wrong

  1. Ensure the data is collected in standardised way
    • Record absences (zero is a number)
    • Keep data tidy (machine readable)
  2. Analyse & safeguard the data
    • Guided by your organisational data strategy
    • Find the pattern!
  3. Exploit the pattern
    • Improve “best practice”
    • Share your knowledge (audience appropriate)
  4. Influence decision makers
    • Promote the evidence
    • Change the world

Data strategy

A set of institutional guidelines, protocols, and systems to ensure that the knowledge gained from data is meaningful, safeguarded, accessible, shared and is used for decision making.

Test: If there is a risk of duplicating effort then you need a data strategy

Critical to have Director buy in

Where we go wrong

  1. Ensure the data is collected in standardised way
    • Record absences (zero is a number)
    • Keep data tidy (machine readable)
  2. Analyse & safeguard the data
    • Guided by your organisational data strategy
    • Find the pattern!
  3. Exploit the pattern
    • Improve “best practice”
    • Share your knowledge (audience appropriate)

Where we go wrong

  1. Ensure the data is collected in standardised way
    • Record absences (zero is a number)
    • Keep data tidy (machine readable)
  2. Analyse & safeguard the data
    • Guided by your organisational data strategy
    • Find the pattern!
  3. Exploit the pattern
    • Improve “best practice”
    • Share your knowledge (audience appropriate)
  4. Influence decision makers
    • Promote the evidence
    • Change the world

Summary

  • Good data integration beings and ends with decision makers
  • Most data problems occur in the Plan phase
  • Automation can solve data collection and safeguarding issues
  • Organisations always have the same types of data issues
  • All medium and large organisations should have a data strategy
  • No organisational data strategy will work without director buy in

Thanks for your attention