Join 3,425 readers in helping fund MetaFilter (Hide)


Data Journalism Handbook
April 30, 2012 10:15 AM   Subscribe

The Data Journalism Handbook is intended to be a useful resource for anyone who thinks that they might be interested in becoming a data journalist, or dabbling in data journalism.

Don't wanna be a data journo? Well, there are lots of pretty pictures like these:

The Handbook At A Glance

Some Favorite Examples

Become Data Literate in 3 Simple Steps

Data visualization DIY: Our Top Tools
posted by Foci for Analysis (8 comments total) 58 users marked this as a favorite

 
Awesome! I'd like to try my hand at it sometime in the near future.
posted by Grimp0teuthis at 11:19 AM on April 30, 2012


Yeah, really awesome info on the data and whatnot. Thanks FfA!
posted by Potomac Avenue at 11:46 AM on April 30, 2012


I find it really disturbing that, as far as I can tell, the only technique for analysis for "data journalism" as given in the handbook is visualization. Visualization is nice, but not a substitute for actual analysis, and, in fact, strikes me as potentially risky trend. Good visualizations should follow good analysis, or there are real dangers involved.

For example, visualization alone makes it harder to separate out correlation and causation, as can be seen in this graph. Statistically, you can use various identification strategies to look at the links between these factors; visually, you can't.

Pure visualizations can also lead to false claims of certainty, as something that can be made to be visually persuasive (for example, by changing the y-axis value) but be statistically insignificant. I understand worries about lying with statistics, but at least there is some underlying numbers that can be checked - visualizations can be even more easily slanted.

Also it is damn hard to do good data work without actually doing some math, though the dictionary talks about cleaning data, it can be really hard to understand issues with a data set without any statistical tools.
posted by blahblahblah at 12:10 PM on April 30, 2012 [5 favorites]


Wait, why was "take a class in basic statistics" not in the handbook? Or even, "read statistics for dummies." This seems like an extreme oversight.

(or anything about the awesome power of R)
posted by redbeard at 1:04 PM on April 30, 2012


Hmm, keep in mind that you're making judgments based on four examples. And a lot of the examples in the "Some Favorite Examples" link contain tons of great data.
posted by itskerem at 1:06 PM on April 30, 2012


Small correction - journalists talk about the tools they use

One sentence reminder that correlation doesn't equal causation

Although "Your search - hypothesis site:http://datajournalismhandbook.org/1.0/en/ - did not match any documents."

Hrm.
posted by redbeard at 4:48 PM on April 30, 2012 [1 favorite]


This is awesome - thanks for the post.
posted by His thoughts were red thoughts at 4:56 PM on April 30, 2012


One of the pages had a link to Poynter free online journalism courses. Thanks for that. I started one and learned a couple things already.
posted by Listener at 5:54 PM on April 30, 2012


« Older Shawn Reeder shoots timelapse video, and his lates...  |  "Sarah Moore researches how wo... Newer »


This thread has been archived and is closed to new comments