Tip sheet: Writing stories based on data

By Mirkolorenz (Own work) [CC-BY-SA-3.0 (http://creativecommons.org/licenses/by-sa/3.0) or GFDL (http://www.gnu.org/copyleft/fdl.html)], via Wikimedia Commons

By Mirkolorenz (Own work) [CC-BY-SA-3.0 (http://creativecommons.org/licenses/by-sa/3.0) or GFDL (http://www.gnu.org/copyleft/fdl.html)], via Wikimedia Commons # Link in context

In getting ready for a new semester, I came across this tip sheet from a computer-assisted reporting class I taught a decade ago. Hivemind, what would you add or change to make it relevant to today’s media landscape? One thing I would add would be some tips on automating data collection. # Link in context

Writing stories based on data: things to remember

1. It’s about the story, not the numbers. The data are just a means of illustrating a point. Make it clear why your data matters.
2. Corollary #1: find your focus and stick to it. It’s likely that during your reporting, you will come across a lot of interesting information that doesn’t fit your story. Lose it. As Faulkner is reported to have said, in writing, “you must kill all of your darlings.”
3. When comparing statistics, make sure the comparisons are valid. This is a problem that often occurs when looking at data over time. For example, some years, the government changed the way it counted unemployment statistics. A longitudinal comparison of unemployment rates would need to take that into account. Another example that we discussed in class concerned the way various states defined “sex crimes” for reporting purposes under Megan’s Law. SAT scores are another well-known example.
4. Corollary to #3: if you are comparing two sets of similar data, make note of differences in sampling methods, error margins or other differences that might reduce the validity of the comparison.
5. Place examples in context – but make sure it’s the right context. For example, let’s say I report that Osama bin Laden’s family has given millions of dollars to Harvard University. (This is true.) I would convey the wrong impression if I didn’t also point out that Osama bin Laden was estranged from his family, which denounced his terrorist activities. In addition, Osama’s brother is a Harvard graduate, which explains part of the relationship between the family and the university.
6. Make sure your data and analyses come from authoritative sources. If an individual who works for an organization makes an assertion about an organization’s history or policies, get written documentation or verification where possible. The employee might be repeating something he or she has heard, and it may or may not be accurate.
7. Corollary: The same thing is true for people who work in highly-specialized fields such as health care or law. When I worked in oncology, one of my jobs was to edit a publication that would provide allied health professionals with accurate, research-based information about cancer, because we were constantly getting calls from people who called us about information they had been given by a nurse or other medical professional that turned out to be inaccurate. # Link in context

CC BY-ND 4.0 Tip sheet: Writing stories based on data by Kim Pearson is licensed under a Creative Commons Attribution-NoDerivatives 4.0 International License.

Posted in Research, Teaching and tagged , , , .


My professional background is in public information, magazine journalism, blogging and journalism education. My current research is founded on the premise that democracy requires the broad participation of a computationally fluent citizenry. Civic media industries must reflect the communities they serve at the level of ownership, research and development, news gathering, presentation and community engagement. This adds greater urgency to the already critical need to broaden participation in computing. To that end, I have collaborated on curricular models for infusing computing into journalism education at both the scholastic and collegiate levels, and for promoting civic engagement in computer science education. My current interest is in exploring the potential of stochastic networks and as enhancement to social computing tools for broadening civic participation.
While most of this blog is devoted to my research in computational journalism and trends in journalism education, I occasionally do some storytelling of my own. This blog picks up where my other blogs, Professor Kim’s News Notes (http://professorkim.blogspot.com) and The Nancybelle Project (http://kimpearson.net/nancybelle.html) left off.

Leave a Reply