“I was thinking we could do something around prevention,” said Erin Rhoda, the editor of a journalism projects team called Maine Focus at the Bangor Daily News.

The other four of us at the meeting nodded in agreement. Little did we know the thought would consume us for the next two months.

The opiate crisis, which Erin has helped humanize through her reporting, appears in the news every day. Yet it’s usually in the context of another death, depressing statistic or debate around legislation. What can be done to prevent more Maine people from developing an addiction in the first place?

We decided to find out.

Our initial look took us to the schools, the institutions that shape nearly every child. Most addiction starts before the age of 20, and most kids spend more waking time in school than at home with their families, so the schools seemed a logical place to start digging into prevention efforts.

We quickly learned no one in Maine had a comprehensive understanding of what schools are doing to prevent substance use.

So we decided to survey the state’s principals ourselves. It’s somewhat rare for newspapers to conduct surveys like this, but we hoped the results would be useful to many people. We wanted to hear not only what schools are doing prevention-wise but also what they need for support.

We drafted questions, then shared them with nearly 20 prevention and school experts around the state for feedback.

Liz Blackwell-Moore, a member of the Maine Opiate Collaborative’s prevention committee, recommended we ask if the schools are involved with their local Healthy Maine Partnerships, for example. The partnerships are regional public health coalitions; there are 27 across the state.

“This would let you assess whether they are connecting to their local resource or not,” she wrote. A handful of other people chimed in with other helpful ideas we would not have considered on our own.

By then we had a big draft survey with many good, expert-guided questions. But was it scientific?

We asked Lindsey Smith, a research associate with expertise in physical and behavioral health, and a focus on substance use disorders, at the University of Southern Maine’s Muskie School of Public Service.

She and her colleague Brenda Jolly helped us determine if it was too long (it was), what kind of response rate we’d need for our results to mean something (it’s complicated), and other such things that social scientists know about.

Next we approached the Maine Principals’ Association, which most Maine principals belong to, to distribute it. Dick Durost, the executive director, was on board.

On Aug. 22, Dick sent the survey to the 466 principals who are members of the MPA. That morning, we watched the results tick in. By Aug. 24, we had 45 responses.

We wanted more. So we scrounged up $500 for school supplies to encourage schools to participate. Dick announced the gift card in an email and our numbers kept rising. Caribou Middle School ultimately won the $500.

Days later, we were still missing a few big schools, so we called and emailed them individually to encourage their participation. Bugging people for information is what journalists are good at, we laughed to ourselves.

We closed the survey on Sept. 6, with 261 responses. About 70 percent of high schools and a third of elementary schools had answered. But how would we interpret the data?

That weekend, my boyfriend, a math nerd who has worked on surveys before, explained the statistical concept of “P-Values” to me. My colleague Adanya Lustig’s grandmother, an octogenarian epidemiologist, shared her own thoughts.

The informal family advice was great, but clearly we needed more. That’s when Jake Emerson, the BDN’s data scientist, stepped in. He’s a University of Maine Ph.D. student in computing and information science, and he could help us understand the data.

Over the next two weeks, Jake guided us. He explained Cramer’s V, contingency tables, and other statistics fundamentals with patience and humor. (You can read all about what he did here, and you can see the raw data here.)

We shared our results with Lindsey of the Muskie school, too. She advised us on how to interpret our qualitative data (“come up with themes”), suggested various correlations between the survey answers to examine, and gave us plenty of other invaluable advice that only someone with more than 16 years of experience conducting social science research can give.

I wish I could thank in person all the researchers, prevention experts and principals who made this possible. In lieu of that, the written word will have to do. Thank you.