Despite the emphasis on data journalism in the last few years, some journalists still shy away from large datasets. And when journalists do turn to data, it’s often to investigate a negative outlier. Data usually informs journalistic inquiries like: What city has the worst crime rate? Where is governance the weakest? Where are racial and economic inequalities the greatest?

We suggest a different way for journalists to use data — exploring positive deviants, or slices of a dataset that get at the best performers. For example: Which hospital in Texas has the lowest infection rate? In which state is recidivism the lowest? What country has the highest participation of women in government?

A positive deviant is a signal that something newsworthy could be happening. It’s the journalist’s job to get the story behind the positive deviant — and in so doing, uncover information that could be valuable for people everywhere. Perhaps the Texas hospital found a way to encourage nurses to speak up when a doctor fails to wash hands. Maybe state prison authorities began providing mental- health services and drug treatment to recently released prisoners. Those are important stories. (That being said, a positive deviant could also signify nothing. It could be a quirk in the data, a function of demographics, or an inaccurate measurement. As we said, it’s the journalist’s job to find out!)

Positive deviant journalism works backwards from data outcomes. Some journalists hesitate to attempt solutions journalism because they fear being labeled advocates or PR representatives. But with positive deviant journalism, the data will guide you to a story — and therefore will refute any confusion with advocacy.

Next time you look at a dataset, here are a few ways to consider slicing it to find a noteworthy positive deviant. This table is not meant to be exhaustive, but rather to spur your imagination a bit.