Data Journalism 2: Interrogating, Visualising and Mashing

Online JournalismCity UniversityPaul Bradshaw

Data 2: Interrogating, visualising, mashing

Monday, 7 March 2011

5 things you need to know about eachData journalism in actionWalkthrough

Themes


.

Interrogating data



1. Data always needs cleaning up2. Treat the ‘source’ like a source3. Use the right ‘average’ and percentage4. Variation over time & space: context5. Spreadsheet tools are your friend - but always backup copies

5 things you need to know about interrogating data



“What the Independent have done is confuse the UK’s deficit with our debt [making] the debt problem look around eight times worse than it is. And it used the whole of its front page to do so.”

- James BallMonday, 7 March 2011


Measurement doesn't answer anything if there's only one variableStatistical significanceSample size and selectionControls and the placebo effectRead up.

What is the data worth?


1. Variance is interesting.2. Variance is different for different variables and in different populations.3. The amount of variance is easily quantified.

- Philip Meyer, Precision Journalism


Data > Text to columnsFind & replaceConditional formulas: =IF(condition, if met, if not)=COUNTIF(range, test)

Getting data in the right form


Edit cells > common transformsEdit cells > split multi-valued cellsFacet > text facetExport...

Walkthrough: cleaning data in Google Refine


.

Visualising data


1. Choose the chart for the purpose2. It can be used to spot a lead3. Good design is when there’s nothing more to take away4. It should be self-contained & have refs5. Be careful with scales and classes

5 things you need to know about visualising data


or http://chartchooser.juiceanalytics.com/Monday, 7 March 2011



What is wrong with this picture?



http://simplecomplexity.net/statistics-without-context/




http://junkcharts.typepad.com/junk_charts/trifecta-checkup/




ManyEyesTableauWordle, TagxedoBatchGeoGephiDelicious.com/paulb/visualisation+tools

Visualisation tools


.

Walkthrough: visualising data with Google Gadgets


.

Walkthrough: visualising data in ManyEyes


.

Mashing data


1. It is what a journalist does best2. Look for a point of connection: place? Person? Company? Date?3. What an API can do4. What APIs there are5. Mashups can be live, updated or static

5 things you need to know about mashing data




Yahoo! PipesOpenHeatMapMapalistxFruitsScraperwikiMaptube

Mashup tools


Inputs - Fetch Feed, CSV, Data, Page, YQL, Flickr, FormOperators - Filter, Sort, Unique, Union, Count, Split, Rename, Regex, Unique, Location extractor, URL BuilderOutputs - Map, Gallery, List, XML, KML

Walkthrough: making mashups with Yahoo! Pipes


Format the spreadsheetPublish it as CSVCopy linkPaste it at OpenHeatMapFix any problems

Walkthrough: making mashups with OpenHeatMap


Edit column > Add column by fetching URLsUse GREL (Google Refine Expression Language)Search web for help & examples

Walkthrough: grabbing geo data with Google Refine


.

Questions?


Links

OnlineJournalismClasses.tumblr.comDelicious.com/paulb/cityoj09Delicious.com/paulb/datajournalismDelicious.com/paulb/visualisationDelicious.com/paulb/statistics Delicious.com/paulb/mashups


Before the lab: play with these techniques yourself, have problems, find solutions, raise questions. Install Google Refine and Tableau on your laptop to use.- Visualise, interrogate or mash data

Lab


Books

Kaiser Fung - Numbers Rule Your WorldBen Goldacre - Bad ScienceDonna Wong - The WSJ Guide to Information GraphicsBrian Suda - A Practical Guide to Designing with Data


Data Journalism 2: Interrogating, Visualising and Mashing

Technology

data journalism

aboutinterrogating data

aboutvisualising data

aboutmashing data

data worth

visualising data inmanyeyes

cleaning data ingoogle

right form data text