Keeping people at the forefront of data stories

Session facilitator(s): William Wolfe-Wylie

Day & Time: Friday, 12-1pm

Room: Boardroom

WILLIAM: Just in case you’re arriving, there’s a link up on the screen, and open to etherpad, and there’s a pile of videos in there. And if you sign up for one, at last part of the interactive part of this. If you’re without a device, they work on mobile devices but if you don’t have any electronic devices, you can pair up, sign up for two profiles and then you can have lots of fun together.

There’s 30 profiles there, so I think everyone here will be able to get their own. How many of you were at the Washington Post guy’s session on the police shootings? Awesome. More than half. This is going to play kind of off of a lot of those themes sort of developing what comes after that.

So first off, introductions. Hi, I’m William, I works at the CBC. I’m a newsroom developer there but we’re a very small team of six people. So we tend to be half editors, half project managers, half designers, half all of those things wrapped into one. So we try to be jack of all trades, so when a project like this comes along, it can be really challenging to, sort of, how to figure out how to tell these stories well and keep people on the forefront of them. So on the etherpad right now, the etherpad is that link, there’s a list of 30 piles. So if you’re joining us, open those profiles on the electronic device of your choice. And for the duration of this hour, you are the advocate of that person in your profile. You are representing their interests in this room. You are their only voice in this world of journalism. So any time that we say something that might run counter to your person’s interests, any time we propose a story idea that would render your person invisible, any time that we present a data viz idea that will not include your person, or that your person is impossible to include in that, throw up your hand, be there advocate, be there voice, be there person because they are not here to represent themselves. So full disclosure: This is—all of these people are missing or murdered women and girls in Canada of aboriginal descent. So as a result of that, there could very well be unfortunate stories shared here today. So if there’s triggering elements of these, if you need to take some time, knock yourself out, if you need to say, off the record, because there’s uncomfortable moments, say, “Off the record.” We can totally handle that. But basically, what we want to do is take all these profiles—these are 30 of more than 250 that we were researching as part of a big project we were working on. And we wanted to talk about ways we could maintain their humanity, their stories, their , while also talking about the systems that failed them. While also talking about why they’re so relevant to the storytelling experience. So what I’m going to do over the course of all this is go over couple of a screen capture of our early design experiments, our early brainstorming, starting with a pitch, a spreadsheet idea, and going through all of our different ideas, the demands from our stakeholders, our investigative journalists across the country. And every time we present a new idea, I want to sort of throw it around the room, to see what people think, where are we failing, what could we be doing better, and what better ideas that we could be sharing with each other. So kick things off, I want to go around the room if you had a chance to read these profiles. These are all real people. None of these are made up. All the photos provided for you are provided by their family and friends. And all the interviews are with interviews of family, and friends, and police. So there’s an element of reality throughout the room today. But I want you to go around the room, and read a detail that you feel particularly struck you, that this person needs this part of their story told. So we’ll start here, and we’ll go clockwise around.

PARTICIPANT: While there’s a sentence that says, she’s been found, her parents are being told that it might have been suicide but they’re thinking that there’s evidence in the case but police are ignoring that. So that’s a fact that needs to have some light shown on it.

PARTICIPANT: That’s really similar to mine where this person, who was found dead, the police also think it’s a suicide. And her sister doesn’t believe that and is having trouble getting information from them. So the police aren’t willing to interactive with the person, it seems.

PARTICIPANT: So I think mine, this person—around her death but not much was known when she was found but they didn’t find any evidence preemptively.

PARTICIPANT: The person that I chose detail that was resonant with me was that she was found in the snow, and killed by blunt-force trauma, just thinking about the last moments of her life, and, you know, being in a really cold environment in Alberta. And killed in a way that wasn’t necessarily humane.

PARTICIPANT: The detail mine that I think would be important to service is that she was a child. She was 17. It actually says that she was 15 or 17. So that, I think, adds another even sadder thing that we don’t even know. But she was a child.

PARTICIPANT: My person was found dead and it just—I think the detail that I would bring up is that it just took a really long time for them—like, for her family to get a report into why she was—why she died and what happened exactly.

PARTICIPANT: My person is named Josephine, and she was also young. She was 17 years old. And the coroner says that she died of exposure, cuts and bruises but the case is being looked into.

PARTICIPANT: My person was also very young, her name was Jamie. And she was celebrating her 20th birthday and her case was being passed between four different police forces and now it sounds like a private investigation that has been taken over. But a very mundane, harmless person that was celebrating her case has been passed along so many times.

PARTICIPANT: My person was found beaten and stabbed to death in an alley.

PARTICIPANT: My person was pregnant with her second child when she died.

PARTICIPANT: My person left to run errands with her grandmother and I kept thinking about how after she was killed, there might have been no one to keep her grandma company anymore.

PARTICIPANT: My person’s name is Roxanne. And she was the mother of two kids, and she was trying to get her life on track because she was, like, battling addictions. And she was killed with another man she was seeing.

PARTICIPANT: The person I had was a 20-year-old who was found dead and the cops say it’s not a criminal offense. But there were three 911 calls that went unanswered before her death.

PARTICIPANT: My person is a mother to four and the year before she went missing, her sister was murdered, and her body wasn’t found until ten years later.

PARTICIPANT: The woman from my profile is a 22-year-old. Police says that her death showed no evidence of suspicion but her family and friends say that she was pushed out of the window for failing to pay drug debt.

PARTICIPANT: My woman, her body was found naked in the ditch near the dump.

WILLIAM: So these are really, really brutal stories. And one of the core issues that we are asked to explore with this investigation is the element that racism plays in women whose cases—all of these women are of aboriginal descent in Canada, and Canada has a really strong problem, right now, with our history of residential schools—sorry—how many people are Canadian in this room? Okay.

In the early 1900s, the government of Canada set up a residential schools program to pull aboriginal children out of off the territories, to educate them in HRA*EURPBLG largely church run and government run schools basically anglicise them. Most of these were abusive environments. A lot of them died, health care was rarely provided and the last one was shut down in the ’90s. So this is the history upon which where this is starting to come out. People who finished their residential schools went back to their families often suffered from addictions, mental illness, and a lot of PTSD. And as a result of that, they’re basically abandoned in their communities, and this stuff happens. But because of racism in the police force, as well, it’s a running theme, a lot of their families don’t believe their friends and loved ones received the investigations that their cases merited. Starting next week, actually, there’s a federal inquiry into how police investigate cases of missing aboriginal women. So this is now the big running story that’s going to be the focus of Canadian news for the next new months until the federal inquiry gets underway. So this was the project that came to us about a year ago, actually, almost two years ago, with, as so many of our data projects do, a spreadsheet. So literally, this is what we started with in our pitch meeting, was we have a spreadsheet. We’ve been trying to get data from police forces, from private investigation groups, from the federal police force, from the government, trying to figure out how many cases of missing aboriginal girls there are in the country. Because there’s no central database for it. So the investigation unit that were searching for this, said that we might create a wall of faces.

PARTICIPANT: Wendy has no picture.

WILLIAM: Can’t do a wall of faces. Two people don’t have faces. Three people don’t have faces. Okay. Hole number one. So what if we did silhouettes? Generic, background silhouettes where we don’t have a face, and then we can do a map. Everyone likes maps.

PARTICIPANT: I don’t know where my person is from.

WILLIAM: How many people don’t know where the person’s from? It’s not in there? Couple people? Three people? So you’re looking at a subset. So you’re looking at 255 profiles, and you’re looking at 10%, there’s no information for location. A lot of them were taken at a young age, and a lot of them were dropped off in the city where they lived transient lifestyles. We don’t really know. Sometimes they weren’t even reported missing until months later. Someone would say, oh, yeah, have you seen Colleen? Actually now that you mention it, I haven’t seen her for a really long time. I wonder if she left town. Few years later, they find her body. So we don’t have location information for a lot of these people. Before as you’re in data land and in development land, your stakeholders always want maps. Everyone appears wants a—could we put this on a Google map? What if we just put all the data on the map and let people sort through it themselves. So this is what a sample profile page would look like if we could interview their family and friends, if we could interview the people that knew them, the police, we could make a profile for each of them. Profiles might look like this, create a database, show people the impact of it.

Create longer profiles. Start to tell their stories. Have a little data section over here, age 15, home status, the year they were found, or the year they were killed.

PARTICIPANT: Are we going to have specific information about where the cases ended, where they were found. So my person, it’s an ongoing investigation and there’s a code case against the police department filed by the family. And that’s important detail.

WILLIAM: Absolutely an important detail because the family says one thing and the police says another thing.

PARTICIPANT: And it’s an ongoing investigation.

WILLIAM: Exactly.

PARTICIPANT: So that needs to be a major part of it.

WILLIAM: So now we have a subset of people who were murdered before their cases aren’t resolved yet.

PARTICIPANT: In the previous screen where you have the grid of images, how so that being sorted initially?

WILLIAM: This one is being sorted by the order in which they were entered into the database. So essentially not.

PARTICIPANT: I worry that my person is going to be so far downstream that nobody’s gonna click on her image.

WILLIAM: And why do you worry about that?

PARTICIPANT: Because I think it’s really difficult—this is a really difficult topic matter and I imagine that users would have a certain sense of fatigue from exploring even just a couple of these.

WILLIAM: Ah, okay. So we have five cases right here. And so if I click on one, click on the other, and click on a third, and I’m reading long profiles for each of them, how many of them am I actually likely to get through as a reader? And by the time I hit three of them I’ve read about 1200 words about missing and murdered people. Am I going to go through 250 of them?

PARTICIPANT: Short summaries?

WILLIAM: Short summaries. That could help.

PARTICIPANT: Just like a couple lines.

WILLIAM: So our stakeholders really, really, really wanted the map. And so they said, “What if we had it like lines drawn across the map to show where they’re born, and where they went missing and where their body was found?” What if we could show their progress across the country and have a little animated pin at their last known location? How many people would that break for? At least three or four in this room? We found that we didn’t have location information for 30% of cases.

So this was version one that got published. This was published February 2015. So just about a year and a half ago. And it’s quite literally, there’s some filtering. So you can see filter by missing and murdered. Over 18, under 18, filter by province. Over a decade, under a decade. Search by cases and name. And hovering over the page gives you information whether they’re missing, whether they’re murdered, their name, their age, and their province. Clicking on one of them delivers you to that page in a new tab. What’s the biggest problem with this right now?

PARTICIPANT: Well, Wendy’s still missing because she has no picture—one is still missing because she has no picture.

WILLIAM: So we have people with no photo at all. In this version that was published, they get a silhouette, a generic, .png silhouette, and you don’t get a name until you hover on it. So the silhouettes really are fully anonymous.

PARTICIPANT: What’s the reverse scenario where they maybe found someone who they didn’t have any information for. So, like, they have a person with no identifying information?

WILLIAM: There are some cases that we had to add retroactively like that. One case was, actually, I think she might be in one of your profiles. She was found dead of exposure in a ditch. And the family came forward and said, we think that might be our mother. We’re not sure. And police said, well, your mother was never reported missing. And the family came forward and said, well, you wouldn’t accept the missing person’s case. Well, that’s because someone saw her. Well, the family said, could we do some dental records and DNA tests just to be sure. And then that came back, and yeah, their mother was dead in a ditch and the police refused to accept that report. So in some cases, it’s only family advocacy that leads to that link they made. So then retroactively we add them to this database as that information becomes available. But yeah, that’s absolutely a thing that happens. Traffic-wise, this version failed miserablely. Nobody paying attention to this. `

I think in the first week, it did 15,000 page views. Didn’t generate much discussion. Most of these page views were of the friends and family of the woman viewed in the profile. So this was when way sort of had the option to abandon it, or redesign, figure out what broke, and dive forward with it. And thankfully, we chose option two.

So family members said, this feels like a police website. This feels like we just visited the RCMP website and visited their missing persons tab. It feels corporate. It feels ugly. It doesn’t feel like we’re actually talking about people; it feels like we’re talking about cases. So that’s where we led our design route. And so, they wanted the map again. What if we had a map? So the first thing we did is we brought their names out. So the hover effect. We have people who don’t have photos available. These are really poor communities. Especially going—there are some cases going back to the ’50s and ’60s, cameras not exactly the easiest thing to come by, let alone family photoal alumnus. So a lot of family photos, especially for these just aren’t there. So we bring the names, ages, details out of the hover effect and bring them into the main face wall.

PARTICIPANT: For the map part, is there geographic data that’s really meaningful, or revealing a pattern? Like, do a lot of them happen in the same place, or something like that?

WILLIAM: Yes, and they tend to be focused around the locations of those old residential schools. But because we don’t have data for 30% of the cases, it’s not representative enough. And the data that they were able to provide through their research was only city-level data. So we would just have a hundred pins on City Hall. And, especially with an interactive map, you zoom in, okay, so they disappeared from City Hall, or maybe we have a home address. Oh, are we going to give away people’s home addresses now when they’re vulnerable populations, possible sex workers? Seems like a terrible idea.

PARTICIPANT: Yeah, it has all those problems, and right now, it’s at a very common part of the page, like, right now, that’s the first thing you see.

WILLIAM: Exactly. And how many people in this room know which part of the country we’re looking at right now? Pffft.

PARTICIPANT: Did you ever get at why people were so compelled by maps? Was it just like something they had seen before, or was there something that was, like, a more emotional, like, attachment to that location information?

WILLIAM: It was basically what they had seen before. The editors in charge of the project were not—I am going to use the term web native. They’re old print journalists for the most part. And they had seen interactive maps that would worked very well by other organizations. So they had in their heads that a many people was a good way of showing individuals across the country. And so we ended up doing some mobsing like this of what might include a map, mostly to show them why it doesn’t work. As we can see, we did that a couple of times.

PARTICIPANT: Regarding details, are you showing just the official report, or what the families are saying?

WILLIAM: Both. Where the police would speak to us, where including their details in official reports. And there’s sometimes links to PDFs of official reports. But families would speak to us, and especially where that information would contradict the police information, both are included.

PARTICIPANT: What type of detail is included in this page?

WILLIAM: This page is literally, first name, last name, age, province, and whether they’re missing or murdered. Then it got into their head we could show a graphic with lines showing the moving across the country from place-to-place. Okay. We can’t show city-level information because that was just be silly on a map. We can’t show home addresses because that would be silly. What if we showed them moving between provinces and general regions on a map. But it’s not a map; it’s a graphic showing the arcs of their travel.

PARTICIPANT: That sounds backwards. There’s no faces. You’ve turned them into metadata, basically.

WILLIAM: Exactly. So we —

PARTICIPANT: Also, you’ve lost your narrative about what’s happened, what was moving around. There’s no—it’s not clear that there’s been an unjustice that has been done.

WILLIAM: How many people are represented by that line? Everyone who moved between Winnipeg and Smiths Falls. Just this one person? What about everybody else? `

So we thought county level. People who are in Winnipeg, people who are in Sioux Falls. It goes back to your earlier point, as well. And looking at what it might look like as a smartphone app. What if we took those filter ideas from the first, show them as murders unsolved, filter by year, age, and you could actually bring out details of individuals as a smartphone app, browser browsing through it as a web app? What do you think about that idea?

PARTICIPANT: Too many images for a phone.

WILLIAM: Way too many. At this point in the game, we’re dealing about 240 cases—240 people. So here we have full name, age, their status, their home reserve, their communities. So it’s the Sekani Creek First Nation. So the time they were last seen, their home province. Any of your profiles invisible to this idea?

PARTICIPANT: That has to be a lot of text that you would have to be reading, you know, the full story, and what the police say, and just to read that on your phone and want to keep going, it’s hard enough on a desktop to do that.

WILLIAM: Yeah, was there another contradiction back there? Yeah, so we come back to the same thing of, okay, if we want to focus on people, how do we make the leap from providing access to all the reporting around all these people to individual stories and humanity and emotional hits, to then extrapolate back to the data. And of course, be able to facilitate back and forth between all of this. And so we thought about a rotator. What if we could rotate through as a slide deck, as an embed inside of the big investigative stories that we’re telling, we could rotate through pieces of data. One in 7692 aboriginal Canadian women are missing. That’s four times the average for non-aboriginal women in Canada. Source: Statistics Canada 2006.

PARTICIPANT: It’s a very weird number. Like…

PARTICIPANT: But the perspective is also, sort of, minimized in part, I mean, the important part is, this is so much more prevalent among aboriginal women than not.

WILLIAM: Yeah, so for this you want to take that four and a half and make that the big part. Swap those numbers around. And if we did that, and then put it in a story that’s your typical investigative feature: Lead, paragraph, paragraph, paragraph, slide deck of interesting stats, paragraph, paragraph, paragraph… face wall of people? How’s that work for presentation layer, to allow people, sort of, grasp the scale of what we’re talking about? How many of you believe that your profiles in front of you would be adequately featured? That the families of the people in front of you would feel like their story had been told.

PARTICIPANT: I mean, reducing it to statistics, again, eliminates the humanity from it, especially if you’re putting the humanity down—so far down the page.

WILLIAM: Yeah. So we figured, what if we combined those, and we created an interactive that allowed you to filter the face wall while seeing the data summaries at the top of it. So you could select missing or murdered, age range, province, decade or year, search by new location by name, and then see missing on the top, murdered on the bottom, year by year on the columns, see, where individual cases placed in time and case type. So you could see prevalence, you could see a little bit more about perspective for this type of case. You could see—and where an individual placed on this page.

PARTICIPANT: Tons of blocks and photos.

WILLIAM: So too difficult to, sort of, take in all at once to, sort of, understanding what you’re looking at?

PARTICIPANT: The bottom line also, sort of, looks like a police website. Like, that’s the same way a mugshot lineup is displayed. It doesn’t look like a memorial because you’re going to have to scroll to the right endlessly.

PARTICIPANT: When you have a problem each of those individually, you just combine those into a bigger piece where you have both of those realizations together. So one looks like a police site, and the other looks like a graph that’s just like dehumanizing.

WILLIAM: So you’re sort of like the worst of both worlds, right?

PARTICIPANT: I mean, this is a problem, whenever you’re using filters, you’re mals values for each person. Almost every one of these here.

WILLIAM: Yes, exactly. And that’s one of the biggest problems that we had to overcome with this. `okay, under 18, and over 18. The age here was ambiguous. I mean, the range provided for the case was all under the age of 18. But there are some people that were just poorly documented that we just don’t know sometimes. So how do we categorize them? Take a best shot? It’s ultimately what we had to do, is get friends and family to say, well, how old do you think she was? And they would say, I don’t know, 19, 20, I think, probably?

I mean, she got carded sometimes. Filter by province. Home province? Province where she was found? Which province? Which one matters? What are people expecting when they use that filter? So then we came up with the idea that representative cases who were emblematic of larger systemic issues: Young people, police ignoring the initial reports, ignoring 911 calls, people who were found decades after they had died, that they could be put up in the top rotator, with emotional quotes from friends and family that outlined their frustrations and their issues.

PARTICIPANT: Like, I kind of—I mean, I don’t want to call it a boring case but she was found strangled, and they found right away. She’s not going to be emblematic, she’s not going to be representative at all. And her family is talking to the police. And probably not you, given that.

WILLIAM: Actually that family was really open to us. There was a lot of distrust with the police. One of the women on this project was a woman that I amed Connie Walk and she grew up in Saskatchewan. And chefs say that when police came to the community, they would hide in the she said, they would hide, because they couldn’t trust the police, that’s who brought pain and suffering in the world. And when their families go missing and they say tell us what happened, nope, but journalists who were born and from the communities where they were raised, tell us what happened, we tend to get a little more results that way. So in her case, her family talked a lot to us, actually. This is version two that went live. `

So this is Leah Anderson, she was murdered in 2015. And featured is a long interview with her aunt about what this 13-year-old wanted to do with her life, what she feels the police failed to do in her case. And then below that, you can sort of see that we have a headline. And this is the starting of about four paragraphs outlining the nature of the project, who we spoke to, how we gathered the data, and who’s featured. So there’s about 15 different cases that are featured in this top rotator right now, each of which, clicks through to a lead profile.

PARTICIPANT: Where are the rest of the profiles on that first page? How do you access them?

WILLIAM: Further down. As you scroll down, this is what you get. Sorry, that’s a design mock. So as you—below that, you get the wall of faces. We actually also published that square graphic that allowed you to filter and, sort of, see the graphing. That was part of version two. And then, clicking on each one of these, pushes down these profiles and inserts mini profiles which, I think someone on the side mentioned as an idea of scrolling through, and they can be tapped with the keyboard, or clicked through to see more of them as you scroll through. The big goal with this was exploration, discoverability. The big problem with the first one was that you couldn’t dive into any more details. You couldn’t see anything more about them. And we spoke to a few of the families, we spoke to the reporters and their frustrations with it and the big thing that they wanted us to take away from that was that you could learn something about these women, about these trends just by browsing through it. So browsing capability was the number one thing, so you could move from Charise, To Don, to Delane. So actually what you’re looking in front of you is essentially just a quick rework of these.

PARTICIPANT: Hey, did you how many people were actually sharing these on social because there are pretty prominent buttons.

WILLIAM: We did. It was a pretty small number. Most of the shares came when we were promoting the cases of individuals when they came into the news when we were sharing them a lot when they were featured on a nightly broadcast of this person’s still missing, do you know anything? Or police may lead. But we also coordinated with the social media team so that every two hours, one of these is posted across social, and so that would promote some interest in that particular case. But overall, sharing on this initial Facebook one was relatively low.

PARTICIPANT: So you featured a certain number at the top, right?

WILLIAM: 15, yeah.

PARTICIPANT: So what went into deciding which of those 15 those would be?

WILLIAM: Those were found out by the investigative lead on the project. She decided basically, the cases that had the most information about it, so when you clicked through them, you got least 800 words about their case, about what went wrong, what failed them, so that you can dive in really deep with a lot of information. And also, public interest. If there had been a lot of interest in their case when it first happened. Leah’s case developed as a national story when it happened because she just disappeared. She was young. She was adorable. She was very photogenic, and she was studying to be a pilot. And then she was found dead. So it generates a lot of interest when you take off those public interest boxes. So when you do it in a long form way in the boxes, that was another reason to put them in the top row. Problems with that? I mean, clearly but…

PARTICIPANT: I think it’s still a good solution, though, because people are not going to be absorb 30 faces, or a hundred faces like before. But to see one face, you have a little more time with just one person.

WILLIAM: Yeah.

PARTICIPANT: Were the features being cycled through, or same 15 for a duration?

WILLIAM: It was the same 15 for a six month period but they were randomized on load. So if you were visiting the page, you were unlikely to see the same person at the top of the page twice.

PARTICIPANT: Was there specific attention paid to people that—where there was some kind of a police misconduct or where the investigation just fell through, and then they found out that the cops didn’t respond to 911 calls in that case?

WILLIAM: That was exactly what the folks on phase three was about. Which, thank you for that lovely segue. So yeah, after we published this, we have a lot of interest. And, actually, three national media outlets in Canada were all focusing on the same project, at the same time, this really, sort of, started a really big national conversation. And about a year after we first published version one, we published version two, labor day of 2015. So almost a year ago.

And by Christmas 2015, the federal government had announced that they were going to launch an inquiry into missing aboriginal women with a focus on police—the quality of police investigations around these cases. And a focus on systemic issues that have failed them that allowed these cases to go unresolved for so long.

So with that, phase three of our investigation was to focus specifically on police, where the police had said, “This case is unresolved.” Sorry, the police said, “There’s no foul play suspected here.” And the family said… yeah, there is. You ignored this, this, this, and this. And the police would say, “Nope. The case is closed? How many does that apply to around this room?” How many of you, based on whether the profile in front of you have serious questions about the police, and how many of you are sitting there going, ehh, maybe this is just like family trauma and you just don’t want to believe it.

PARTICIPANT: Serious question with the police.

WILLIAM: Anyone in camp two?

PARTICIPANT: Maybe I just don’t have all the information but it’s harder to tell that.

WILLIAM: Yeah, and sometimes it is. So does where we had to go, we’re not going to trust every police officer who comes in with a report but, at the same time, we are going to recognize that there are some systemic inequities that are promoting poor work. So this is one of the mobsing we built for the phase three launch, which is really `building on the initial launch that would maintain a consistent brand idea. This is now a project that has been live for a year and a half. But what we wanted to do was bring these leader profiles into a more interactive space, so that you didn’t have to onto another profile page to read them. They’re leaders for a reason. And you can click on this drawer to pull down. You can also click to see how many are up there. And these are all cases that have been labeled unresolved. Police have said they’re resolved. The families don’t believe that. And so we’re investigating further on their behalf.

We refined the stylings and refined the face wall to make it more easily navigable. Work on more or devices. Actually, now, I could just bring it up since that’s the live version. So my question for you right now is: Still today, who is invisible? Whose stories are not being told yet?

PARTICIPANT: I mean, the lack of photos is really insurmountable in a lot of ways. I’m just not going to identify with the ones that I can’t see. And then now I could back to 1950, or whatever.

WILLIAM: Cheryl was 15. She disappeared in 1987. But there aren’t photos, there aren’t photos. Now, I mean, is there a better way to produce these? I mean, our designer made a generic silhouette that looks like it could be anybody. It doesn’t look fake like a bathroom image. But it’s also very generic. Is there a better way to handle that? I mean, we could assign a sketch artist to visit a family.

PARTICIPANT: Or maybe if they have some detail—some small—I guess long hair, or short hair. Yeah…

PARTICIPANT: I also think that, like, indicating somehow showing the age in the silhouette. So this person the age was 15, is that correct? So she is a lot younger, or she’s in the younger half of the group. I think somehow showing that in the silhouette might be—might work better.

WILLIAM: Yup. We could definitely modify that. What other pieces of data are missing from your people that you feel is really a crucial bit of information that could omit them from storytelling?

PARTICIPANT: Do you mean missing from there, or missing from the profiles, the information that we don’t have?

WILLIAM: Missing from the people in front of you.

PARTICIPANT: Well, we don’t know anything about who they really are, or what they were doing.

WILLIAM: They were pursuing their dreams? Could we include that in some way? I mean, you’re right. Dawn here is 17. She was murdered in 1987. Someone was accused, there was a trial, the person was acquitted. That’s literally all we know about Dawn. She’s a 17-year-old, police accused somebody, it didn’t work out, eh, I guess that’s done now. Is there a way that we could approach—and this is—we don’t have any answers here. Is there any way that we could approach the lack of data as a datapoint. We go to work, telestart throwing up data. This is the messiest data that you could imagine. Where we assign 15 people to go out, and start families. And we built a backend CMS just to handle it all. Nothing is consistent about this. No two people are alike. No two cases are alike. Sometimes we have the most basic information like a first name and a last name. In this case here, we don’t have a first name, or a last name. It was just a baby—7 months old. So we can’t search her by name. She’s listed in the databases as, “Baby girl.” So this is the discussion that I want to have. And in the last ten minutes we are here is, when we’re faced with a lack of data with people, how do we represent their interests when we’re faced with other people competing for the same stories that have so much data? Yeah?

PARTICIPANT: I mean, I think you could, like, really randomize the presentation of the ordering of somebody’s profile, and so when you have somebody who doesn’t have any data, they’re featured. And there is, in place of a narrative of their life, maybe, sort of, an explanation about all the factors that make it so hard for them for, you know, to know about it. And so, that those gaps are the systemic things that prevent us knowing them are as prominent as somebody who has a very identifiable story.

PARTICIPANT: Yeah, kind of going off of that, like, what we did in here today, even people who got profiles that didn’t have a picture, maybe felt more connected with whatever profile you’re paired with. So something like that is, every time you visit the page, you have one person. And it might be that they don’t have a picture but you still like…

WILLIAM: So present, like, assigning the reader to a profile.

PARTICIPANT: Yeah, and randomly.

WILLIAM: Yeah. Yeah, so finding a way of introducing them to a random person on the page and saying this is a person that you have to care about right now instead of allowing them to browse through, forcing that situation upon the reader.

PARTICIPANT: Maybe.

WILLIAM: That’s an interesting idea. I like it.

PARTICIPANT: I mean, I’m going to remember this person that’s in front of me. She’s been in front of me—I’m going to look at your product, but this is what I’m still going to come away with, that’s just the reality of it.

WILLIAM: Yeah.

PARTICIPANT: I also think that it’s really important when you have an explanation to explain why they don’t have it. We just did that was not exactly like this, but it was, like, 150 mug shots of people. And we had a few narratives on them. And there was a few that we did have it on. And we asked the government for this information. And they refused to give it to us.

WILLIAM: And the Washington Post presentation today yesterday did about how many FOIAs they filed. And how many were denied.

PARTICIPANT: I like the idea of making the unidentified person as a part of 15, so you would go through, and it doesn’t have information, and then you have the opportunity to tell the reader that clearly, there’s a lack of data for this connect, or say, like, you know, how many people out of the group doesn’t have a lot of information to kind of share and tell that story.

WILLIAM: Yeah, that’s a great idea. Too.

PARTICIPANT: I wonder if the answer is not so much about the data, or the reporting or the lack of data is a racial issue, and a discrimination issue.

WILLIAM: That was actually a side story along the way, especially when the residential schools were closed down, a bunch of their records were destroyed, which is one of the subjects of this inquiry is, destroying government records, that’s not a good thing. But yeah, absolutely. Being transparent about tactics, as well as, transparent about what’s available from where is a really good idea. So I would say—did I see a hand on this side? Maybe making that up out of my head. Cool.

One of the things that we’ve also had a lot of luck with is this sidebar that exists on every homepage. There’s now three spots on the homepage is a call to action—if you know anything else. This feature was passed around a lot of First Nations communities, and a lot of families and friends of people. And that email address became a very active source of tips and extra information for these communities. So while page views didn’t necessarily reflect, like, this wasn’t the most successful feature that we had ever run if traffic is our only measure of success. But in terms of providing closure and providing a resource, and providing an outlet to friends and family of the people who are profiled here, it was an enormous success.

And there’s a—we’re working on a spinoff feature for this fall based on an email that came into that email address that—all it just said was, “I know who killed her.” And it was one of the unresolved cases. And then the person sending the email turned out to be a retired police officer who worked on the case. And he just couldn’t find the evidence to put this person away, but he’s 90% sure he knows who did it. So that’s what we’re working on—that story right now, it’s probably going to release this fall. But having those call to actions on the sidebar turned out to be an enormously powerful thing into furthering the reporting. Does anyone here use SecureDrop at all? Is everyone familiar with it? Cool, just for everybody else, it’s also a thing that we just set up in the past year. But, essentially, it’s anonymous-to-anonymous secured document leaks. So if you’re a government employee, or a corporate employee, and you want to leak documents but you don’t want to make yourself known as the source of the leak, it’s an encrypted two-way communication system that allows you to leak a vast number of documents. And even the reporter receiving the documents doesn’t know who you are. It sets up pseudonyms between them, and pseudonym email addresses between the two people communicating with each other. So as a reporter, you would only know, you got you just got a link from platypus 67. And based on the quality of the documents that you received, you have you decide whether to trust them. populate + 67 only knows that they’re dealing with dolphin68. And they have to deal with you, and trust you based on your organizational affiliation. So the facilitates that two-way communication so that you can leak information anonymously, and securely.

PARTICIPANT: Did you get a lot of people that wanted to contact you through that versus the email address?

WILLIAM: It wasn’t expressly requested for this project. But it was `one of the ways that we received some of the Panama Papers. So it’s been used in the past to great success. Sadly, I didn’t get to work on that project.

PARTICIPANT: Can I come back to the social button because they kind of like…

WILLIAM: That’s a continuous war, so I’m fully open to all feedback.

PARTICIPANT: They’re kind of just like—they’re very, especially given a story of this magnitude, and kind of this, you know, where the—where, you know, the content is so serious, like, putting Facebook, Twitter buttons kind of, can become make it devalue the, you know, enormity of, like, you know, the seriousness of each of these cases, and I don’t know, they just look very happy, these buttons, right?

And if they’re not that popular, like not a lot of people are, like, using them, they shouldn’t be there. But, of course, I’m sure that you can find this…

WILLIAM: Yeah, this was the war between our team as the developers on this particular project, and the product team, managing the look and feel, and branding of the website as a whole. We are constantly in a push-and-pull war of what branding limits we’re allowed to push against, and what branding limits we’re completely, 100% bound to. The social media buttons were a rather fierce war that we lost wholeheartedly. Right now we’re incorporating VF4 and social into all our long form social pages like that. That’s something we’re bound by and it’s unfortunate. I agree with you. It breaks the tone of the page. And since they’re not being used, we could use that space for better information like a call to action.

PARTICIPANT: Yeah, absolutely.

PARTICIPANT: So were you able to release these different versions because there was new reporting coming out, and so it was an opportunity to do a new version?

WILLIAM: Yes. Essentially, we released version one, and there was recognition within the company that it was not as well done as it could have been. The lead developer who was working on it also quit his job halfway through the development of this, and so that’s how I became the lead developer on it. And so that also, sort of, changed our approach rather significantly as we sort of had to pivot partway through. He took a lot of expertise when he left.

But then yeah, once we released version two, that was mostly our team lead on the news interactives desk that really fought for version two to show what we could be capable of if given free rein over a project like this. And then after version two was released, we started winning a lot of awards for it. And other news organizations were winning a lot of awards for their projects and then the National Inquiry was announced, partially inspired by our coverage and other news organizations’ coverage, Global Mail, and the Star did enormous coverage on this file. And once the federal inquiry was announced, we were encouraged to continue because we were getting results, basically.

So, it’s never been a traffic success with us. But it’s been a results and journalism success for us. And so that was worth pursuing. We have two minutes left if anyone has any closing ideas or questions or queries. Yeah?

PARTICIPANT: Did you ever feel like you had to—I guess I’m curious, I think a lot of times, in journalism there’s, like, a sense of being separate. You know, separate from the interests of your sources, and apart from them. But it sounds like—I mean, for this project to be successful, you really have to be empathic to people who are advocating for these stories. But even advocate whether it’s within our own design process, or whether with a—with editors. How do you, like, balance those concerns?

WILLIAM: Um, exactly as you said. Actually, I was reminded a couple of times during working on this project about the dark Knight where Commissioner Gordon gives up that kid, Joseph Joseph Gordon Levitt, saying you’re a kid now, you’re not allowed to believe in those perspectives. We took the belief that this isn’t about objectivity anymore. This isn’t about trying to find the middle of the road. We’re advocates on their behalf. They were systematically ignored by the systems that were put into place to ignore them and subjugate them. They were placed in these schools to erase their identities. That was the stated purpose. And so we decided that that bears no middle of the roading. We’re just going to be advocates on their behalf. We’re going to tell their stories and that’s the stated goal of the project. We largely gave up on the idea of being objective around it.

PARTICIPANT: So one thing that you said early on is that one of your reporters was a First Nations person who led and was really helpful in doing that. Do you think that the story has changed the way—you guys aren’t—the way that the organization hires?

WILLIAM: There’s been more focus on diverse hiring practices. Her experiences with this project in particular have definitely been a really enormous success story to saying, like, oh, no, when your entire newsroom isn’t filled with white men, you actually get really, really good results for a number of reasons. So anyone who’s looking for a reason to believe that now has a really solid case. But it’s been a stated objective for a number of years. But I think, I think this, sort of, added a little adrenaline to it, for sure.

PARTICIPANT: What’s the next steps for this project now that’s going to be inquiry and just like reports that are going to be released, I think?

WILLIAM: Yeah, the inquiry is launching next week, and they’re stated to deliver their first report in October. So, right now, we’re pursuing more detail on these cases that are—whether there’s a discrepancy in the quality of the police investigation, we’re trying to get more details on that. And we’re pursuing those individual cases right now. It’s a research project right now and we’re keeping the database updated. It’s not clear if we’re going to get the resources to do another design refresh on this. But when they release the report, depending on what that looks like, we may find the resources to do that. Cool. Well, thanks, everybody. If anyone’s having any other ideas, feel free to approach me later on. But thanks for coming.

[ Applause ]