Conducting a survey on your beat
This week, Kristine spoke with reporter Nicole Hayden, who covers homelessness at the Oregonian. When local government didn’t have sufficient data on outreach efforts to those in need of housing, Nicole got the idea to conduct the survey herself and write about the results.
Kristine Gill: Can you tell me a little about your beat and how you got the idea for this story?
Nicole Hayden: So, I just focus on homelessness. And that means, you know, keeping up with city council decisions, but more so focusing on projects, doing big picture policy things that we think can have an impact. I guess my specialty is figuring out how to do social science research in newsrooms, and so I've done surveys like this before at prior newsrooms.
And with this one, we decided to do it because the city was getting kind of their side of the story about a certain situation, and then folks living unsheltered that I talked to said something that contradicted the city. And we couldn't get data that kind of proved either. And that's when we decided to do surveys – when there isn't enough public data to answer an important question. Then we decide to kind of create our own data.
So what was the specific issue in this case?
Essentially, what the city was saying is that when they do sweeps of encampments—so when they go and kick people out from where they're living, essentially, trash all their items—that they offer services beforehand, that there are these outreach workers that say, ‘I can bring him to a shelter, I can help start the housing process.’ And they're required to do that under the Martin v. Boise decision, which essentially said that Western states can't criminalize homelessness.
So you're not allowed to forcibly remove someone, but if you do, if a city does want to force someone to move from where they're sleeping, then the only way they legally can do that is if they offer an equal or better alternative, and typically that’s offering a shelter bed. So anyways, the folks on the ground that I had talk to on a weekly basis continuously told me over and over that they were not seeing any outreach workers and that no one would offer them help before a sweep. So we wanted to show that this wasn't just an anecdote by one or two people, because the city can easily dismiss that. So we wanted to get really quantitative value for this qualitative experience. And that's how we decided to do the survey.
How long had you been hearing these anecdotes and looking for data before you decided to conduct a survey?
It was almost a year, maybe nine months, that I kept hearing the same thing over and over before we decided to look into it further. I feel like less than a year in (at a paper) it's hard to really understand which topics that you should devote time to.
What’s a regular day like for you covering a population like this? Who are you talking to?
Definitely, you know, city council, nonprofit folks, like you really would any other beat. But I tend to spend as much time as possible and try to root my stories in real people's experiences. So this whole past week, I've been going out to meet with one particular individual who's homeless to kind of track her story. And I also stop at other encampments and kind of walk through, so it's a lot of in-the-field reporting. Not necessarily because I need to do a story but to check in and see what's going on with folks.
I imagine you’re running into a lot of the same problems these social workers are, trying to find people who don’t have a phone or address.
Sometimes I feel like I'm running around the city trying to chase people down who don't want to be found, but usually folks stay in the same area and even if they don't have a phone, you can track them down.
And how long have you been on the beat?
I've been at the Oregonian on this beat for a year. I've been a journalist for about nine years and have written about the intersection of public health and poverty the whole time.
In the past when you’ve done similar surveys, how did you make those decisions?
Historically, local county, and state governments are really bad at keeping track of data about homeless issues, so that’s also why we decided to survey at my prior paper.
And then, unrelated to homelessness, we also did a survey at the Coachella Music Festival about sexual harassment and assault experiences of women. So that's another thing where we found the local police department really wasn't keeping tabs on harassment or assault cases related to the biggest music festival. So we decided to step in and do something.
And I believe you worked with a company to conduct this survey at The Oregonian.
Yeah, so all the prior times, it was just me doing the surveys and that can feel insanely impossible and difficult. And one perk of doing it by yourself is that you have complete control and you can make sure you're asking the questions the same way every single time so it's consistent, but it's definitely less stressful if you can work with folks.
Other people in the newsroom who volunteer to help you can be helpful, but in this particular case, we contracted with Street Roots, which is a homeless nonprofit. And they essentially hire folks to sell their newspaper and do outreach. And also, they hire people to do surveys so they were already kind of experienced and immersed in what we were doing so we paid six of their workers to go out and walk around with us and ask folks questions. And that was also helpful because they were all either currently or previously homeless. So they already had relationships and trust built. So already having trust in the communities and you're doing something like is helpful.
How big is the homeless population?
In Portland, there are around 4,000 folks who are homeless, we're just focusing on unsheltered homeless, though – so people who are literally outside. And so it doesn't include people like sleeping on couches or in temporary shelter.o that unsheltered population was about 2,000.
That is based on the point in time count, a federally mandated account of homeless people counties have to do every year to understand how many people are in need. And so we chose to survey 300 people, because that would give us a 5 percent error margin, and it mattered that our sample size was representative.
How did you design the survey? I think the survey was 26 questions?
So I start by just writing the questions that I want answered, and then also, you know, basic demographic questions. But then we approach experts to review our questions to make sure we didn't miss something that we should be asking, and to make sure we're not asking questions in leading ways. You don't want people to answer the way they think you want them to answer, you just want a truthful answer. And so you have to make sure you write questions in a fair way.
And so we approached research experts and homeless experts, both local and nationally, to help review our questions and we went through multiple drafts before we did our actual survey.
OK and I know you did something like 50 of these surveys yourself. What did that look like for you?
Oh, yeah, that was just me. So I was out every single day with the teams doing the surveys to make sure that they were doing it correctly and that we're going to a variety of different places. Some people hadn’t done surveys before so I would just stand next to them and help them ask the questions. But aside from that, on my own, I did 50 surveys. I wanted to make sure that I was on the ground and seeing what they were seeing and actively involved, because just looking at the data without context isn't always the best way to go about doing journalism.
And let’s see, it took about a month and maybe 10 or 12 days (to conduct the surveys.)
Some were really quick, like five minutes, then at times maybe it's like half an hour or longer? It depends. You know, some people are open to doing a survey and just want to answer the questions and move on. For other people, it's the first time someone asked, and they want to be heard so they tell you everything. And definitely we don’t want to rush people when they're telling us a dramatic story.
And then other times people are experiencing mental health issues but if you're patient and listen you can get answers. This one guy was trying to explain to us that an octopus was coming to take your stuff or something. So at the forefront it kind of sounds like he isn't able to answer questions, but he really is, you just have to wait there and be patient. So we want to make sure the most vulnerable individuals get to answer, so some took longer, but mostly it can go like five or 10 minutes, really fast.
What has the reaction to the story been?
It was overwhelmingly positive by the like general readers, really great, supportive emails.
However, the county and the city are less positive. They essentially think our research method is flawed. And they essentially tried to imply that people who are homeless are liars. So how can my data be accurate if everyone lied to me? Which is disheartening for them to say. And then one policy director in the county tried to tell me I should have gotten a HIPAA release signed by every individual so I could go and check to see how many people were lying to me about the services they received. So essentially they're trying to discredit people's lived experiences.This has just been to me in emails and phone calls.
I like that this story was concise but also included an anecdote. Did you find Jennifer in the course of conducting the survey?
No, so we made the survey anonymous so that people could feel safe talking to us. So we kind of tried to keep it separate. When we wanted someone on record, we went back out and looked for folks. But we wanted everyone to feel safe, so we did not try to in the same moment say, this survey is anonymous and then also try to get their picture taken.
So I went through a few different sources before I found (Jennifer) mostly because in other places I’ve reported, it's been really easy to keep track of folks and I've had relationships that have lasted one or two years. But in Portland, it's been pretty interesting reporting because people get kicked out of where they're living so often and in the process, lose their phone. So it's super hard to keep track of people here because they're just moving around losing things. And so trying to find someone who I could keep following up with and get photos with was hard. So we found her afterwards and she was willing to let follow her around.
What advice would you give to people who are covering homelessness? How do you stay safe? How do you find people?
In terms of being safe, I feel pretty comfortable going out by myself. I don't go out late at night to random places I'm not familiar with. But if you're trying to get integrated, and you're new on the beat, definitely talk to activists who are working on the ground, volunteering. They probably know more than anyone.
It could be hard to convince a nonprofit to let you follow around their caseworkers because of HIPAA, but these more guerrilla activist groups are doing things and they are going to let you follow along and connect with people. And so that's really helpful. They have so many eyes on the ground that if something is happening, they immediately know what's going on and who to talk to.
The ACLU has contractors that work in the community that can be a really good source, especially if they're working on, like, group lawsuits against the city. They know a lot of people with lived experience. Neighborhood associations are good if your city has neighborhood associations, but really just pushing yourself to at least once a week, go and walk around in a new area and meet new people is the best way to really understand what's going on.
What kinds of stories do you think lend themselves to a survey like this?
It only makes sense to do a project like this when you really need to, since it's so much work. You can do it on any beat and it can be as big or as small as you want. It can look like just putting a Google form up online to gauge someone's opinion.
We like to be able to claim that we have a 5 percent error margin but if reaching that many people isn’t within the capacity of the newsroom, you can talk to like 20 people and having 20 people talk about their real life experience is still quite significant. So if there's just a question that you're having a hard time answering, surveys are a really interesting way to do journalism. It's definitely helpful for any population that is vulnerable and maybe doesn't feel safe.
I would also say never design a survey just by yourself because you're never the expert. I will never be a survey expert compared to actual social science research researchers at universities where this is their full-time profession. And as a journalist, we can make our research slightly more squishy since we're not trying to get published in like a scientific journal. And no matter how many people we talk to, it will still be important, but definitely getting your survey reviewed by as many other topic and survey experts as possible is necessary and then testing your survey before you go out into it.
So take a day and survey maybe five or 10 people and you'll see where the glitches are in the survey and ask them for feedback and sometimes, the folks you're interviewing will suggest another question. So definitely spending a lot of time designing and testing the survey.
I think I talked to maybe five people (who reviewed the survey in advance.) Portland State University has their own homeless research team, which does actual professional surveys like this, so talking to them was really helpful. And then we reached out to a national homelessness coalition and then Street Roots, who we contracted with.
Do you plan to use other data from the survey for future reporting?
Yeah, so this story, it just focused generally on how often folks come in contact with outreach workers. And what we haven't published yet, but will come in a couple of weeks, is the specific question about when the city does sweeps if they are offering people help before they kick folks out. So that will be its own story.
Tips:
If you can’t find data, create it with surveys.
When conducting a survey, ask questions for the issue you’re focusing on but include other topics so you can use the data down the line for other angles.
Consider working with a local group, nonprofit or experts to either design or conduct the survey.
Free tools such as Survey Monkey can help tell you things such as the margin of error for your sample size. https://www.surveymonkey.com/mp/sample-size-calculator/
What we’re reading
Fortune: COVID is making it harder for schools to practice restorative justice
The Lily: As disabled women, our lives have always been erased. But in a pandemic, they’re urgently at stake.
AL.com: Police in this tiny Alabama town suck drivers into legal ‘black hole’
Fortune: Omicron is milder than Delta. So why are hospitals full?
Awful Announcing: MNF crew criticized for referring to Ben Roethlisberger’s sexual assault allegations as ‘immature’
Fortune: A new contact lens that slows nearsightedness is a boon for optometrists
Poynter: I’m not mad about Ben Smith; I’m mad at all of this
Teen Vogue: What it's like to be the child of a January 6 insurrectionist
St. Louis Post Dispatch: Missouri’s governor still wants to make reporting on the state’s data screwups a crime
The Atlantic: The Absurdity of Renting a Car Will No Longer Be Tolerated
Miami Herald: How decades of problems converged the night Champlain Towers fell (amazing visual storytelling!)
Job board
Morning Brew is hiring a social media editor
The 19th is hiring a designer
The Atlanta Journal-Constititon is hiring reporters to cover the pandemic’s longterm effects on Georgians
The Virginian-Pilot/Daily Press is hiring a food writer, web producer, health reporter, city reporter and education reporter
The Seattle Times is hiring a digital editor
The Las Vegas Review-Journal is hiring an education and health reporter
The San Diego Union-Tribune is hiring a money reporter
The Marshall Project’s new Cleveland team is hiring an editor in chief
The New York Times is hiring a politics reporter to cover the southeast