BBC Fusion Data Day 2013 #bbcdataday

We're liveblogging from a day of talks on data journalism at the BBC's Fusion Data Day. You can also follow on Twitter #bbcdataday

17/07/13 Reverse order Timestamp format


  1. Bill Thompson introducing the data day with a little history. The first session on 'exploring data' begins in a few minutes.

    Help Me Investigate
  2. First speaker is @kevglobal talking about a story on shale gas #bbcdataday

    Help Me Investigate
  3. "Fusion Tables is a simple and easy way to use what are usually called pivot tables" - @kevglobal

    Help Me Investigate
  4. [email protected] demonstrating pivot tables in Google spreadsheets - personally I avoid 'column' and use 'filter' instead for clearer overviews #bbcdataday

    undefined Help Me Investigate
  5. [email protected] demonstrating how the =IMPORTHTML function can be used to load tables into a spreadsheet - for more see tutorials like this

    Help Me Investigate
  6. Tool 3 from @kevglobal is Tableau Public for data vis - free download here: http://www.tableausoftware.com/public/download

    Help Me Investigate
  7. Second speaker FOI expert @rosenbaum6 - "If you're a beginner [Excel] is the tool you're going to use"

    Help Me Investigate
  8. Simple calculations, sorting and filtering are the first things to learn. The second is pivot tables. - @rosenbaum6

    Help Me Investigate
  9. If you want examples of FOI-led stories see my bookmarks at http://pinboard.in/u:paulbradshaw/t:foieg

    Help Me Investigate
  10. To read about @rosenbaum6's story on MOT failures see http://datajournalismhandbook.org/1.0/en/case_studies_13.html

    Help Me Investigate
  11. Change over time, league tables, association, big numbers are all possible data stories - @rosenbaum6

    Help Me Investigate
  12. [email protected] has written about surgeons' mortality rates here :http://www.bbc.co.uk/news/uk-politics-22489062

    Help Me Investigate
  13. Mortality rates reported differently: BMJ 'within expected range'; Mail/Telegraph: wide variation of performance; Independent: surgeon fears; Guardian: flawed data; BBC: 'historic move'

    Help Me Investigate
  14. "When you began asking questions, this is what you heard" - http://www.bbc.co.uk/news/uk-politics-23248754 - "... what you needed to do was look at the data adjusted for risk [not raw data]" @rosenbaum6

    Help Me Investigate
  15. Now talking about designing editorial products using data is @robinpembrooke

    Help Me Investigate
  16. Robin's role includes reducing risks of possible duplication of planning and commissioning, wasted time/effort, and lack of depth

    Help Me Investigate
  17. Identifying storylines, events and data associated with them is key - @robinpembrooke

    Help Me Investigate
  18. Circa, ITV News and VOX Media given as examples of new ways of looking at storylines

    Help Me Investigate
  19. BBC have published details on their storyline 'ontology' here: http://www.bbc.co.uk/ontologies/storyline/2013-05-01.html

    Help Me Investigate
  20. BBC News Labs Hackday on Oct 24/25 venue TBD

    Help Me Investigate
  21. RT @kevglobal: The BBC has an internal tool called the Juicer that automatically updates connected stories via metadata.

    Help Me Investigate
  22. Tagging your content so that it fits into that storyline system will become more important, says @billt

    Help Me Investigate
  23. The second session is looking at visual journalism, with @BellaHurrell @walt_jw @Leimdorfer and Helen Sears

    Help Me Investigate
  24. Mapping corruption: http://www.bbc.co.uk/news/business-23231318

    Help Me Investigate
  25. Affordability of housing is one area where visualisation and interactivity can perform an important role http://www.bbc.co.uk/news/business-23234033

    Help Me Investigate
  26. "The budget calculator is the grandaddy" of interactive features at the BBC - @bellahurrell http://www.bbc.co.uk/news/business-17442946

    Help Me Investigate
  27. "Express what you expect to do in one short sentence ... which can then be used in broadcasts" - @bellahurrell

    Help Me Investigate
  28. Find out if the data exists - combining different datasets can take time, and you need to make sure they are compatible - @bellahurrell

    Help Me Investigate
  29. The fuel price calculator combined 10 datasets http://www.bbc.co.uk/news/business-21238363

    Help Me Investigate
  30. On #distribution: "Enable users to share a key fact to boost reach" suggests @bellahurrell

    Help Me Investigate
  31. "What do you want a software to do? Which of these is most important?" - Andy Leimolorfer at #bbcdataday

    Help Me Investigate
  32. 'Keep it simple, keep people in mind and have fun' - Helena Sears talking about data stories on James Bond movies and cats

    Help Me Investigate
  33. "Journalists are probably the most demanding consumers of news" - Miles Bernie at #bbcdataday

    Help Me Investigate
  34. "Tags and using good tools to search archives very important part of research for stories" - @zillahwatson

    Help Me Investigate
  35. "The co-relation between a series of events (not always linear) can be a very important factor to identify or spot a story" - @jeremytarling

    Help Me Investigate
  36. "Don't let lack of data stop you from using data" - Harry Robbins

    Help Me Investigate
  37. "Discovering the data, discovering the story, having conversations with the data and asking lots of questions while looking at the data - key points to keep in mind when attempting to tell a story using large data sets" - Tony Hirst @psychemedia

    Help Me Investigate
  38. "Anticipating problems with FOIs is a good way to ensure you get useful results from it" - @paulbradshaw

    Help Me Investigate
  39. "Take numbers away - journalists have this skill with words, but not always with numbers" - @paulbradshaw

    Help Me Investigate
  40. "This is not data journalism, this is journalism" - @kncukier

    Help Me Investigate
  41. Do journalists need to learn coding? The world is not divided into muggles and coders [email protected]

    Help Me Investigate
  42. There is no such thing as clean data sets, it's just dirty and dirtier - @veltman

    Help Me Investigate