I’ve come to the realization in the past couple of years that teaching computer-assisted reporting requires first teaching how to approach stories with a "data state of mind" — a term that I’ve co-opted from the famous Barlett and Steele phrase, "document state of mind."

There are certainly many reporters who already do this — even if they don’t do their own data analysis — and these reporters are most likely the ones to eventually gravitate to CAR. Typically it starts when they come to the realization that they want to do their own analysis to measure something or find a pattern or trend, but they don’t know exactly how to make the software work.

Reporters who don’t already have this data state of mind tend to think in terms of human sources or reports or summary statistics. Their aim on stories tends to be to find somebody or something to tell them what to write. They don’t have the inclination to go measure it themselves — or possibly they haven’t thought of it simply because they don’t know how.

This past month, one of my colleagues at the Pioneer Press published a story that serves as a perfect example of how you can get a better story by approaching an idea from a data state of mind.

Jeremy Olson, our health care beat reporter, is one of the most CAR-savvy reporters I’ve ever worked with, and he routinely publishes stories that benefit from his skills.

The most recent one, "Accidental deaths in Minnesota soar as more people abuse or aren’t aware of the risks of prescriptions," started from a run-of-the-mill press release e-mail. It talked about a federal study showing that overdoses of prescription and over-the-county drugs was the leading cause of accidental poisonings among American children.

Olson wanted to find out if this was also true in Minnesota.

At that point, he had two options. The first was to call state health officials and ask the question outright. The second was to see if he could measure the trend himself.

A reporter who doesn’t have a data state of mind would have chosen the first option.
Olson chose the second option. He decided to turn to our state death database, which we obtain annually and keep in-house, and look for deaths from overdoses. It wouldn’t quite match the federal study since that looked at hospitalizations overall, but it would give him a place to start.

The data showed him that there had not been an increase in deaths from medicine overdoses among young children. However, it pointed him to another story: a huge increase in the number of young adults who died from medication overdoses. It also directed him to the lead of his story: Cody Nelson, a young man who had died from taking too much Tylenol.

Olson called state officials and confirmed his findings and took the reporting further from there, resulting in a compelling 1A story.

I asked him what he thinks he would have gotten if he had chosen the first option and not looked at the data first. He thought about it for a minute and said he is fairly certain he would have walked away without any story since the fact that there have been more deaths among young adults wasn’t fully on the radar screen of state health officials.

It’s the perfect example of coming up with better — and exclusive — stories by having a data state of mind.

I’ve uploaded a PDF of Olson’s story.