Teaching journalists to do data
I struggle with how to teach data reporting. I want to give the students an introduction to how to do it, and how to make it work online (conceptually, for the audience, I mean), but I only have four weeks. I want to make sure what I give them has a practical application, but I can’t load them down with a gigantic project.
Ideally, I would like them to analyze some meaningful data and then explain it to an audience. I would also like to have them present the data visually online, using simple tools.
All of our journalism students get some Excel training in a course called Applied Fact Finding (it’s our public records how-to class). This would be a second step, aimed at getting them to do both some analysis and some formatting for presentation.
One aspect is making sure they understand charts such as these, from The New York Times in July 2008, explaining the financial crisis. Yeah, the one that is spiraling out of control right now. The Times actually published some very scary visuals three months ago.
Another aspect is maps, like the always-updated electoral votes map at The New York Times.
In the comments on my post yesterday, Jude Mathurine wrote:
… computer assisted research, reporting and analysis for contextual online journalism as a foundation prior to writing and editing. Also, how about: Manage metadata for search and social promotion. Storytelling using UGC and mashups like polls, geomaps, charts, galleries and timelines …
Angela Grant wrote:
… understand when stories can be enhanced with online database applications and interactive graphics like maps and timelines …
And Brendan Watson wrote:
I would add basic research methods. In the Internet age, journalists have access to reams of data, both for reporting and analyzing the impact of their reporting. They should be able to make sense of this data, or at least understand how others make sense of this data.
All of them are right, but I’m still struggling with how to turn this into an assignment that students, working in groups of three, can effectively learn from.
Categories: data, reporting, teaching, training
If a journalist lacks basic data finding and analysis skills, then all they can produce is a piece with factual claims by the people involved, and balance will demand that competing factual claims also be included, leaving us with the next-to-useless he said/she said type of stories we see so often.
Here’s an assignment: pick a political candidate’s factual assertion (about, say, the percentage of small businesses that would be affected by a tax hike) and have them (a) run down the underlying data that on which it is based; (b) derive at least two independent factual statements (not necessarily contradicting the original claim); and (c) evaluate the opposing candidate’s counterclaim in the same fashion.
Now, if you could only find some politicians that are making various claims that you could check …
If you want someone to be a guinea pig and be the first to try the assignment, I’d be down for it. I need to get better at CAR.
I wonder how ahrd it would take the real-life example and make it work in a class?
With reckless ignorance to your students level of expertise, I find I’m routinely troubled by reporters missing the boat when it comes to backing up claims with some actual numbers.
Quick example would be the story about voter registration being up. It can be done with anecdotes and some overall numbers (typically from a mouthpiece) but would be better done by actually getting the data and see how much it’s up, where it’s up, who’s making it go up, etc.
What about giving them several story assignments and ask them to identify which ones would be capable of being CARed and how they would do it? Data skills are great, but they can be taught. Knowing when they should be used is a harder thing to learn, IMO.
Gathering and comparing numbers can be as problematic as balancing he said / she said. I think anyone using numbers as facts must have a basic understanding of the assumptions for those numbers.
Some things are so simple, like margin of error for polls. Check out this LA Times fishing expedition in an effort (it seems to me) to prove with numbers which candidate “won” a debate. Spoiler alert: Obama’s “lead” increased from 48% to 49%.
Hold the presses. No mention of the MOE in the story. I had to click on an image to see full poll results, then scan through the tiny type below to find a 3% MOE. Statistically, McCain is just as likely to have “won” as Obama.
This is a non-story that should have been axed at the copy desk.
One of the best sites for serious digging AND story ideas is the Census Bureau. In the Washington DC area we were always lucky because we could get someone from the bureau to come in and hold the students’ hands as we mined the depths of the databases.
But there is still a lot people can do on their own using the American Community Survey portion. Students can break down local communities by ethnic groups, income, age, education, etc.
One example I showed my students was the high college educated level in Fairfax County (I think it was about 90 percent) AND the high income level. Then I pulled up another county with a much lower college education level and compared that to the income in that county.
We then discussed if this is a one to one relationship. I made the students figure out other ways they can start with those figures and then find additional data to look at the assumptions one might make and how those assumptions are right and wrong.
No sooner did I hit the submit button than I received Al Tompkins’ daily article from Poynter Online.
Topic: How many have health care in your county? Source for research: A new Census Bureau report.
Here is the link for Al’s article: http://www.poynter.org/column.asp?id=2&aid=152003
A timely issue that requires research.
[...] Hm. Keine deutsche Seite dabei. Seltsam. (die meisten Links via Teaching Online Journalism) [...]