Christine Zhang is a freelance writer and data analyst who loves stats, stories, spreadsheets, and sandwiches. She just finished up a fellowship with Knight-Mozilla OpenNews at the Los Angeles Times Data Desk and has previously worked at the Brookings Institution in Washington, D.C.
Can you start by giving a broad overview of your work and why you do what you do?
Until recently, I was a Knight-Mozilla OpenNews Fellow based at L.A. Times. I worked on the Data Desk — a group of reporters, computer programmers, and statisticians — people who use data to tell stories in many different forms.
As a Fellow, I also traveled to a lot of conferences and spoke about data-related topics. I gave trainings and tutorials on data analysis tools, and taught the basic concepts of data visualization — usually to people who didn’t have a lot of background in these things.
I also got to meet a lot of different and really interesting groups of people from different domains within, or even outside of, journalism.
Now, I’m working on independent data projects, including finishing up a project with the Times and freelancing for places that let me tell stories with data, like the podcast/blog Data Skeptic.
Thinking about this work and what you’re doing, could you hone in on one example or a specific time where you really felt a sense of success?
Success is defined in a lot of different ways, but as a journalist, you always think about what the impact of your story is and how to create the most journalistic impact.
As a data person I’m not sure how we would quantitatively measure impact, but I would say that qualitatively there is the “big picture” impact, where a story can change somebody’s life, or systematically change people’s lives, or effectuate change within the system.
Then there’s a really “small picture” impact in which your story was useful to somebody. One of my successes and happiest memories of a project isn’t something major — but something pretty simple. I made a map of all of the locations in L.A. that were having midnight parties for the new Harry Potter book — a little data project for the L.A. Times that I did over the course of a couple of days. Then I went to one of the Harry Potter midnight parties just for fun, and I was talking to a few people in line — there was this lady with her two girls, and she was saying, “Oh, you work at the L.A. Times. That’s so cool! I just used this map to look up where to find this party.”
The fact that she used something that I had created — even though she didn’t know I had created it, was pretty great. I was able to “change,” or “improve” a person’s life, and then I got to meet that person. Sometimes you forget about the people behind the numbers when you’re so focused on abstract things like computer programming or coding. This was a great reminder of that — my moment of success.
That’s a really cool moment. I’m hearing you say that meeting that person and having that point of contact, even if it might not have been a really serious deep issue, was important — that made you feel the success.
Yeah. I also feel a lot of success when I go to conferences and give talks and tutorials — part of what’s difficult about data is that it can be mystifying to people. A lot of what I have done in this fellowship and what I want to continue to do is to demystify things, and to make sure people know, for instance, “R is actually not that scary — sure, it’s a little bit difficult to learn and there’s a learning curve, but it’s not magic. It’s actually pretty logical, once you pull back the curtain.”
When I gave a talk on ‘Intro to R,’ at the Investigative Reporters and Editors Conference in New Orleans, people came up to me afterward and said, “Wow. I’ve always hated R. Now I’m going to use it more.” This is another way I’ve felt a sense of success.
I’m thinking in terms of what you fantasize about achieving. What you hope teaching R and demystifying data journalism in general will achieve? Once the person gets a little familiar, what kinds of things do you hope will happen? What kinds of things would you love to see?
I would love to see people feel empowered to use data to tell their own stories.
One job of a data-savvy journalist to help people make sense out of data, to create a story with the numbers. Another job is to showcase the data — to make it more accessible to people so they can look at the information and craft their own narratives. I think that’s really important, too.
For example, a map of where people have voted for a certain political candidate — that’s a common data visualization. It’s empowering because you can look up your neighborhood and you can look up other neighborhoods.
You can see trends and form your own ideas about the landscape of America in a sense.
I think doing that with other types of data and other types of visualizations is what I’m hoping to achieve — this is why I think demystifying data is important.
Thinking again about the work that you’ve been doing. What would be an example of a challenge?
There’s so many. A challenge for me is trying to understand how to make the information into a story that has newsworthy impact. There are so many data sets in this world to write about and to showcase. As a journalist there’s always this struggle of, do we want to tell a story with the data we think people should care about — but they’re not really talking about? Or do we want to find a data set on a topic that is already being widely discussed and write about that?
I think that line is a bit difficult — I personally think everything is interesting, but will someone else think it’s interesting? If they don’t think it’s interesting, how do I make it interesting? I’m a pretty quantitatively-minded person, but I’m always curious about information and presenting it in the best way. I try to find an angle that people can identify with.
Related to that challenge is when the analysis is complicated. How do you explain things in terms that people will understand — in terms that resonate with people? My background is in academia, and I worked at the Brookings Institution, a think-tank, before I did this fellowship in journalism. The tendency in academia is to first describe your methods, your analysis, and your data set. Then, you go into your conclusion from the data and present what you found. Basically, you have a whole lot of introduction and background and research and then you get to the point.
In journalism it’s usually reversed. You have to get to the point pretty quickly because nobody is going to read your thing if they can’t figure out what the point is after the first paragraph. You might do all the research and background, but you don’t necessarily write about it in your piece. Thinking about that inversion is sometimes a challenge with data journalism — especially if you want to tell stories in a responsible way.
Yeah. I’m sure you see a lot of things written in ways that drive you crazy now.
I think a lot of it comes from the need or desire to be attention-grabbing. The challenge of how to make information newsworthy — how to tell things in a way that makes people care — that’s really hard.
Sometimes it’s tempting to write a story that’s really compelling but doesn’t use the most responsible techniques of analysis or doesn’t use the right data.
Can you tell me more about how you approach addressing the first challenge you mentioned? Especially between stuff people “should” care about and newsworthy stuff. How do you decide?
That’s a really interesting question because, in my opinion, what qualifies as “journalism” is a huge philosophical question. On the one hand, you don’t want to be patronizing toward people and tell them, “this is what you should care about.”
At the same time, the role of the free media in a democratic society is to create a well-informed public. I read a blog post from Brookings that tried to define the role of a data journalist — I believe the quote was something along the lines of, “We try to tell stories in a thoughtful and impactful manner.” The wording was a bit high-level, but it really gets to the crux of what I try to do.
It’s important to have a newsroom full of people who come from very diverse backgrounds — who have different ideas about the world — so that the stories told also reflect that diversity.
I think a lot of the time people refer to things like “breaking the story.” There is a writer, Rebecca Solnit, who gave a commencement speech at a Journalism School that defined “breaking the story” in two senses. Breaking the story could mean that you find a really interesting scoop and you write it up — that’s breaking the story because that’s something people didn’t think about.
Then there’s breaking the story as the physical act of breaking something. The “story” in this case would be the narrative, or the dominant philosophy. Breaking the story could mean completely changing the way that we think about something. Thinking about that in these two ways makes it a little bit easier to determine what is newsworthy.
I think the challenge I had with the Brookings quote is the word “impactful,” because it’s so coded. It’s like, “Affect some type of change.”
I remember it more precisely now. It was, “communicate ideas with great consequence in a thoughtful and impactful manner” — but what is impact, right? One way of thinking about impact is “breaking the story,” challenging or reversing the dominant idea, or breaking the story by finding a different angle.
Yes — reframing. There’s all that great Robert Edmond work about framing. Asking, “what even gets talked about?” That’s a great example. Turning now to the broadest issue in Mozilla Universe. How would you describe the open web or a healthy web? In your words?
I think that a healthy web is a place where people feel safe and empowered. It is a place where different people from different groups can connect with each other in a way that they wouldn’t have been able to otherwise — they can easily find each other.
In the very beginning of the web there was this high-level language of how it could be this great equalizer — how it would be amazing. The web of today, though, is not that — we need to bring it back to what it was before or what it was conceived to be.
An open web allows people to openly access information and openly access each other.
Can you give me an example of how some of these aspects of being able to access others, being able to access information have been important for you, personally?
Yeah, for sure. One of the reasons I was able to get the OpenNews fellowship or found out about the fellowship — although I attribute it to a lot of things — but mainly there are two reasons. One is the Hacks/Hackers Meetup group, which is an organization that brings together coders and journalists.
I was in D.C. working at Brookings, and I was trying to understand my next steps, as a person who likes data storytelling. I thought journalism would be a really great move.
I didn’t know where to start — how to get into journalism — especially as somebody who didn’t do it at the time and had no background in it whatsoever. That was pretty daunting. For me, trying to find and meet those people — there was not a great way of doing that in person and without the web.
I went on Meetup.com and found the Hacks/Hackers group in D.C. Going to Hacks/Hackers and other events for aspiring journalists was very helpful for me. It was one of my friends who I had met from the Meetup who told me about the fellowship.
He said to me, “You seem to be talking about a lot of themes that are important to the fellowship — maybe think about applying to it.” Without that extra push, I wouldn’t have applied to it or even thought about it. That really helped me. That’s a way friendship has helped me.
The second reason is a blog post from OpenNews itself. My friend had told me about the fellowship on the last day before applications were due! I was pretty hesitant about applying at the last minute, but I read on that blog post that many last-minute applicants actually do make it to the final rounds of the process. So there’s an example of a data story having a very real impact on my life.
A very common thing I’m hearing from people is that it was a friend or someone they met who was like, “Hey. Did you consider Mozilla? They’re more than Firefox. This is what they do.” Then it goes from there. Did you have the perception that Mozilla was sort of more? When did you realize what the broader mission was?
A little bit. I started to realize it right after I had applied for the fellowship. I was going through one of the rounds of interviews for the fellowship. I told somebody I worked with about the fellowship, and he had heard of the Mozilla Foundation because he knew some of its grantees.
I realized, “Oh wow. The Mozilla Foundation funds a lot of things in addition to scholarships, including other fellowships and a lot of things related to openness and the web.” Then, I noticed the extent of their reach when I went to MozFest for the first time in London. That was a really great and overwhelming and amazing experience all at the same time. That’s when I realized the scope of what Mozilla strives to do.
What were your big takeaways from your first MozFest?
Oh my gosh! Community for sure. It was so great to see people supporting each other and encouraging each other — even if they worked in competing organizations. Journalism is such a competitive and difficult industry sometimes.
Even if they were work competitors, people were working together to create something that would be useful to everybody. I really appreciated that aspect of it. That would be my main takeaway — a genuine sense of being a part of something people really cared about, expanding across different fields.
The next question is getting more specific about Mozilla. How you got involved with them. What that’s been like. I realize you’ve answered some of this already. Feel free to repeat or just elaborate.
I first got involved because of this fellowship. I would say that my involvement culminated in the 2016 Mozilla Festival in London.
Before going to the festival, I had wanted to give a “lightning talk.” I asked the OpenNews people, who were running the Journalism track, whether or not I could give a lightning talk. They said, “Actually, we haven’t done lightning talks at MozFest before. You might have gotten that confused with other conferences.”
Then we all thought about it a little more and wondered, “Well, why doesn’t MozFest have lightning talks?” Lightning talks are great. Short five-minute talks on any given topic. You get to hear from a wide range of people.
The Mozilla Foundation funds so many different types of fellows, so OpenNews and I thought that we could have the inaugural lightning talks this year, and span all the fellowships — not just the OpenNews fellowship.
My first year at MozFest was this sense that there was a whole other group of people on the Open Science floor. There were also other types of tracks. That was cool, but I didn’t necessarily know how to get involved with those groups or how to access them.
In the process of facilitating the lightning talks this year, I was talking to fellows from the other groups that Mozilla funds and felt more involved with all the different aspects of Mozilla.
I’m really curious as to what you observed cutting across these different tracks and groups. Were there any commonalities that came to mind? Any big differences?
Yeah. There were definitely commonalities. As a data person, I automatically think about open data. The understanding that certain types of information should be free — and freely accessible.
Going back to how this can empower people to tell their own stories. People can’t tell their own stories if they don’t have the tools or the information to do so. I think what I noticed among all the different groups is that they are helping to provide people with either the tools to analyze information or easier ways to access it.
One good example of that from this year was when I went to the Open Science floor and sat in on a project called “Open Knowledge Maps.” This project showcases research studies that have been published in Open Access journals. Open Access journals exist and you can look up different research studies, but you’d have to go to the website of each Open Access journal in order to get that information. This project tries to consolidate that information and create a map of open knowledge.
I really enjoyed that presentation. I thought it could be a great tool to be used within journalism and science. As journalists, part of our job is to take information and make it useful to people by telling a story with it.
The information contained in research studies is a really good example. If there was a map of all the research studies on a particular topic, that would be a great tool to help us tell those stories in a more meaningful way. This is just one example of crossover potential and achieving the shared goal of openness.
That’s a really great example. Is it also making it easier to search across all the journals?
Yeah. That’s what I really liked about it. In addition to mapping the most commonly talked about topics or the most popular research articles on any given topic, it allows you to search the universe.
You can see what’s trending in the Open Access universe, that can you give you a hint that there might be something newsworthy here we can translate? Cool. That makes sense. Can you give me an example of a time when Mozilla has had some kind of impact on your life, your work or even something as broad as the L.A. Times?
I feel like every time you ask about impact, I talk about people. At my first MozFest, I met somebody there from the Digital Humanities space. His name was Andrew Prescott — he’s a professor. He was speaking about his work on relating to digital humanities — a field that I didn’t know much about or even what it meant.
I’m somebody who really likes the humanities and cares deeply about them — a session I co-led at MozFest this year was called Poetry for Programmers. I also care about data. I think digital humanities is a great coupling of the two and has a lot of potential within journalism.
Andrew Prescott introduced me electronically to somebody at UCLA, Miriam Posner, who is the Program Coordinator of the Digital Humanities program at UCLA. I recently gave a talk there about data journalism and what that means to me. How the digital humanities can be used in a journalistic sense.
Mozilla has really helped me by connecting me with different groups and people and by encouraging me to think about things in a different way — to think about how groups can match up with each other, even if they haven’t in the past or they don’t usually.
Even the term “data journalism” is about different things coming together.
There has been a trend toward what has historically been known as “precision journalism” or “computer-assisted reporting.” As journalism becomes more and more digital, so also does the data analysis element of journalism as social science.
My interest in data journalism began when I saw that the social sciences and journalism had this amazing crossover that I’d be interested in getting into. Ever since then I’ve realized it’s not just social sciences and journalism that have crossover — it’s other things as well. Open science, and open journalism, and the open web, and security. Digital humanities, coding, and poetry — all types of stuff — it all relates.
Mozilla is a great organization because it not only funds all of these different spaces and makes them possible, but it also encourages people to think beyond their own space — to think about how different spaces can collide.
In your time that you’ve interacted with Mozilla, was there a point where it didn’t meet your expectations or would you have any very pointed pieces of feedback of gaps or things that could be better?
That’s a difficult question. Some things could be better. At MozFest, the lightning talks were great because they allowed me to meet people who were Mozilla fellows in other areas.
But that was the first time I had interacted with the other Mozilla fellows — which was strange to me. It was also almost kind of a shame because my fellowship culminated in MozFest this year.
Oh, that’s the end of it.
That’s the last big conference that all the fellows went to in 2016.
It’s the last time you’ll all be together.
Yeah. It’s the last time we were all together [at the time of the interview, it hadn’t been announced that the fellows would all go to NICAR in March 2017]. It’s a shame to have met the other Mozilla fellows at the last gathering of all the fellows — you know what I mean? I think it would have been better to have a beginning meeting of all the fellows or an introduction email or an online community where the fellows could all talk to each other.
That would be my one piece of feedback — I think it would really help in encouraging people to think about the collision of spaces.
Has this collision of spaces been a theme that has come up within the journalism fellows? Or is it something that you’re particularly interested in?
I think it’s a theme that’s come up among the fellows in general. I don’t know about the other fellowships, but the OpenNews fellowship really strives to take people from all different backgrounds — not just journalism.
In the past five years there have been fellows like myself who have never worked in journalism before. The goal is to bring different perspectives to the industry. The collisions are necessary for the fellowship and for success.
There needs to be stories about advocacy — and data journalism behind advocacy — literally, these areas all fit together.
Exactly. And security. One of the fellows — Martin Shelton — he studied security and privacy issues as part of his Ph.D. Bringing that to the world of journalism is huge. This is a topic that really needs to be discussed and will continue to grow in importance.
These are the questions we’re asking across the network. Do you think that any of these types of stories might be useful to you if at all?
Yeah. I think so because it would be really great to hear other people’s perspectives on the open web, and it would be really useful to hear about the different backgrounds that others come from. I think about the collision of spaces all the time, and in order for spaces to collide, I need to hear other perspectives — that’s something that’s really important to me.
If you’re thinking about where new solutions or innovation comes from — it comes from that cross pollination and cross collaboration. Of course it can come from being really deep into a topic, but I think I’m like you and believe that having a multi-disciplinary approach, and people from different backgrounds, and different walks of life, is really important to getting things right, and to designing solutions that are robust.
Is there anything more you wanted to talk about?
You had asked what one major challenge was for me. I talked about how it’s hard to make things newsworthy, and hard to figure out what people care about, and it’s also hard to combine the two.
A project I did about women leaders who have been elected around the world is an example of this because it had so much data. It was a project that I did with a co-worker, Andy Roberson, at the L.A. Times.
We painted a picture of how many countries have elected female leaders in the modern age. The most difficult thing with that was that there was so much data and so many different ways to showcase that data.
There were other outlets that had done a similar thing by showing maps of countries — yet we didn’t want to focus so much on the number of countries, but on the people behind the numbers. To do that, we wrote up biographies of every woman leader. We put that on an interactive timeline — you might say we cared more about time than space, which is why we didn’t have the map.
The challenge with that was deciding on the angle or dimension that we were going to go for and the story that we were going to tell. This data can spawn so many different types of analysis — there are whole research disciplines dedicated to understanding gender equality in politics and in other domains, so it would have been really easy for me as a data geek kind of person — who has some background in academia — to go down this rabbit hole, to say “Hey, let me calculate this thing. Let me calculate that thing. Let me do some analysis on all these different dimensions.”
Maybe I could still do that for separate stories. Yet, for getting that one story out, it was important to focus on the one angle that we were going to take with the data. Rather than focus on multiple angles and get really overwhelmed.
That makes sense. It’s the quandary of the Ph.D. student. When you’re writing your Thesis. I’m interested in this topic and all these aspects but you’ve got to narrow it, narrow it, narrow it. The other analogy is something my mom always says, “Choisir c’est de perdre.” To choose is to lose.
It’s a great quote.
Yeah. It’s about opportunity cost.
I think that’s why it’s hard to glean journalistic insights when you have so much data. And I constantly wonder, “did this work?” And even if it didn’t work, what is the solution to make it better next time?
It’s perfectly fine to say when something didn’t work for you. Ideas and how to fix it or ideas on how it might be approached are great in terms of the feedback from Mozilla, but really, just having a formal systematic feedback loop of what’s not working in a way that’s safe. Someone’s listening to you, being like, “Yeah. That makes a lot of sense.”
The challenge that you raised is something that smart, interested, curious people go through. It might be helpful to put the data online and suggesting ways the data could be used.
Yeah. I’ve always found really great when I read research papers, in the conclusions they usually mention, “Here are the things that I didn’t explore but maybe somebody else could.” I feel that is really overlooked sometimes in research.
Of course you can’t analyze something every which way because there’s infinite possibilities. You never have a monopoly on the knowledge. But you’re much more likely to do something to a data set if somebody were like, “Here’s some things you can try to explore with the data set.” Rather than if somebody was just like, “Here’s the raw data. Do something with it.”
Exactly. It’s funny, I did see a story that relates to your data set, probably in the New York Times. It was about times when women take over countries, there are certain common factors happening.
Oh, the “glass cliff.” There’s this theory — I think the NYT piece might have been written before the election — it was saying that oftentimes women are appointed or elected or get to the top of a company or a political party in times of turmoil or conflict.
That’s exactly what I was reading.
In addition to a glass ceiling, even once they’ve broken through the glass ceiling, they risk falling off of a glass cliff.
Right. It’s a shit show — now, you take care of it.
Theresa May was the example.
Yes. Theresa May was the example that I read. It makes me think a lot. Your data set could be used to validate that hypothesis.
Yeah. That’s true. I do think that for women, because there are so few data points in the data set, that makes it challenging to do a lot of analysis on it. The only solution would be collect more data. More women should be in leadership positions.
I think that actually is a great example of how data can be biased, that statistics could be biased. Even with the glass cliff theory, if you just do a simple correlation of turmoil or conflict with female leadership, you would see, “Oh wow, things like women leaders are associated with conflict. This is bad.” But it’s not a causal relationship.
More women but also you could think of some indicators and get economic data for those countries at the time. Mush those things together.
That’s a good idea.