SRCCON:WORK Talks: Sisi Wei
Day & Time: Friday, Dec. 8, 2017, at 3pm
RYAN: All right. I think we are ready. So our final speaker for SRCCON:WORK is — maybe we can get the projector up actually. Our final speaker for SRCCON:WORK is Sisi Wei. Sisi is a data journalist and editor-in-deputy for news apps at ProPublica. How many of you all are familiar with ProPublica’s Data Institute? Awesome! The Data Institute is an intensive training workshop and Sisi helped create it. In that process, she learned a ton about applications and about hiring. So in an alternate timeline, or perhaps after this talk, in the hallway, Sisi will be here to present some very different research — her deep, deep dive into dot-grid notebooks, how to evaluate paperweight and spacing, and you should ask her about pen bleed. There’s a whole world, but to take us into the final breakout sessions at SRCCON:WORK, she’s here to challenge us into thinking about networking and making recommendations to each other. So we’re excited for her to share with us some goals that can truly help create more diverse teams in newsrooms. So Sisi, thank you, this is a challenge that I think we need to hear.
SISI: So requisite mic check. Can you hear me? Fine. So when someone you know is hiring for a position, and someone you think is talented is applying for that position, something that we do these days is shoot off a recommendation. That might be calling a colleague, or emailing someone, I met someone at SRCCON today. Hey, I know you’re hiring, I have this fantastic applicant who applied, let me tell you about them. Now for people who are hiring, there are very good tangible benefits to this because when you receive recommendations, most of the time what happens is organizations get a pretty talented, excellent candidate when you hire that person and it saves everybody time and the people who are doing the hiring really feel a little safer because someone they respect has vouched for this person.
But even though I have made and received many recommendations like this, every time something has always felt a little bit off. And, in fact, sometimes receiving recommendations like this have made me feel very uncomfortable.
It was kind of like something unfair was happening in the ethos but I didn’t know what it was, and I couldn’t do anything about it. And it wasn’t until the past year or so that I have been able to figure out why. So here’s what I’d like everyone here to consider: everyone thinks of recommending people as a way to promote one deserving person and that person is probably very deserving, very talented. But when we move to make that recommendation, what we’re by default doing is literally pushing down everybody else who applied. Now every person who gets one of these recommendations gets them because they know someone who knows someone who knows the hiring manager but 99% of your applicants are not gonna know somebody. They can’t get the hiring manager’s network to recommend them. And yet, because of that, are they somewhat inherently less deserving of getting your job? Should their application start off as disadvantaged in yet another way compared to someone who just happens to know somebody?
I know the answer is no. So who are the people most likely to not have those connections? Those are the people who never worked, didn’t go to the same schools as us, didn’t get the chance to go to the same conferences that we do, people who don’t look like us, who aren’t our colleagues, who aren’t in our socioeconomic class, there’s a million things that I can say here. But when we talk about newsrooms, and how they’re too monolithic, when we talk about, aren’t they the people that we’re trying to hire into our newsrooms and onto our teams? But I bet you’re thinking right now, Sisi, what if I only recommend women or people of color, or people that are different than me; is that okay? And here’s the difference between hacking the system and trying to fix it, right?
So yesterday I thought it was very heartbreaking to hear Disha’s talk about her not being taken seriously until someone was willing to vouch for her as basically the status quo. It shouldn’t be that way, but it is. And when we send recommendations for journalists of color, we’re hacking the system, which is a good thing but there are also cons.
So the first con is we’re still promoting one person, pushing down everyone else, including other journalism colleagues. But second, we’re actually propagating the system that in order to get hired, skills and expertise, that’s not enough. You have to know somebody to get that job. And that is just not good enough for me and I don’t think it should be good enough for any of us. What I want is for us to leave that aside and come fix this system, right? So how do we fight back against this idea that the only way an application can float to the top sometimes is by knowing someone?
Thankfully, there are ways. In fact, I have a couple suggestions for what we can do, both as individuals, and for those of you out there, who have the power to influence how your organizations actually hire, how we can build a better hiring process to counteract these things.
So individuals, what can we do? First, instead of sending out a recommendation, dedicate an hour or two to the person you’re trying to recommend instead. Help them with their application. Coach them. Edit them. Explain to them niche jargon. Especially if these candidates are unlike the ones that we already have in the industry they probably aren’t familiar or aware of the unwritten rules of journalism that we just take for granted. One of my summer fellows was a computer science student who loved ProPublica but she didn’t know anybody in the organization much less anybody in journalism when she first came to us. And when I helped her apply to jobs, what she needed most was actually the understanding of what a developer-journalist résumé and portfolio was supposed to look like. What do we value and how should she present it in her work? So I helped her edit and rearrange her résumé. I showed her examples of portfolios from people probably in this room, reviewed what she came up with, and now those are tangible skills that she can take to any job she chooses to apply for. And second, as individuals, one of the most worthwhile and difficult things that I think we should be doing is to identify people who are not planning on applying for something but they should and to give them the support to help push them past that last step.
So one of my colleagues has a mentee who was a college student at the time and he had never applied for a single summer internship and it turns out the reason behind it is because he always worked the same summer job in order to pay for college and he was afraid that if he didn’t go back for essentially one quarter, right, that his hourly wage won’t drop back to minimum wage instead of what he approved.
Now what my colleague basically encouraged him to do was just apply. We’ll take it one step at a time, we’ll see what happens. The long story short is that a large, national magazine offers him an internship and coaches him into his summer job to see what he can do. Turns out they never had a problem with it. And the student just needed that push from his mentor to even ask in the first place and that was even more valuable to him than any recommendation would have possibly been.
So those are some ideas of what we can do as individuals. But what about those of us who can affect the hiring process, right? Now before we answer that question, I want to make sure that we’re on the same page for all the sort of onslaught of things that we’re fighting against. We talked about recommendations but there are millions more, right? If you want to read the studies, for those of you who have not, please ask me afterwards. But it’s pretty much a given that human beings have biases based on sex, race, and that bias is worse the further you get in an application process but the worst part is — the research I’ve seen so far, people are willing to change what’s good or bad based on who they want to hire based on their discrimination. Some of you might know this story. One applicant named Michael was street wise, he was tough and he worked in rough neighborhoods. The other applicant, Michelle was educated in how to administer, and she was a political candidate. Who was a better fit for this job? The evaluators said they picked Michael because they said it’s more important to be street wise than have better connections. And that’s fair. That’s a fairly reasonable explanation to come up with. But in part two of this study, what they did was showed the exact same people’s descriptions but swapped the names. The evaluators picked Michelle because she was educated and Michael because he was street wise. So people will actually just redefine what they think makes a good candidate based on what they want and who they want.
But it’s not over. So on top of that, there’s one more layer to this research that basically showed that when making these decisions, evaluators perceived themselves as actually being more objective than normal because they were comparing people based on merit, right? So let’s just pause on that for one second. The conclusion I took away is that not only are human beings more likely to discriminate, we then justify that discrimination in order to work on our behalf of our biases and then we actually feel like we’re more objective while we’re doing it. And to bring the psychology sort of a little full circle, there’s something out there called the Mirror Exposure Effect. So as humans if we’ve seen, or heard something at all, we’re more likely to prefer it just because it’s more familiar to us. That’s why advertising exists. They want exposure. So it works the same thing with names. All else equal, if someone sends me a name, and I read it, if I don’t know anything else, right, I’m already more likely to prefer this person and I might not even remember why.
Okay. So how do we really fight this? What do we do to try to eliminate, as much as possible, our biases on sex and race, as to something like being exposed to names. A proposal is to use blind applications. What this basically means is when you’re reviewing applications, turn off people’s names and turn off people’s email addresses. We tried it for two years at the ProPublica Data Institute in which we got hundreds of applications where I probably got dozens of names given to me every year. But because the process was blind I didn’t know who was who and I also didn’t know people that applied and it entirely changed my understanding of how we can be fair in the entire process and it’s something that I’m considering doing in a fellowship program next year. The beginning on your websites, make sure that what you’re asking for on the application is basically as clear as possible. There’s nothing implied, no explanation needed for the people you prefer because it’s all out there. And then, stick to this criteria when you’re actually reviewing applications later.
Use good application software so that it lets you track applications as if they were data, and it lets you hide information when you choose to do so. Which, of course, leads me to when the application form goes live, turn off all the email addresses and names right away. Ask people to provide demographic information just so you can analyze your applicant pool throughout the entire process. And leave a good window open. For us, it was three months, it wasn’t that long. But it gives people time to actually work on a project that you’re asking for or be able to understand the industry enough to reexplain their projects based on the criteria that we’re looking at.
And then throughout that window, when those applications are open, check the demographics of who’s applying. If you don’t see enough journalists of color, or any other type of diversity that you’re looking for, that is the time to fix the problem, not when we’re in final interviews later and the applications aren’t open anymore. And at every stage, again, look at those diversity breakdowns to see and make sure you know what’s happening at every stage of the process.
If there are multiple people making the decision, as there always is, once you have your candidates finalized, rate them separately so that everybody has to make their own decision then see if there’s any overlap, discuss, pick your final choices and then reveal people’s names.
So this is a small slice of all of the things that we ended up doing in this process for the Data Institute and if you want to talk more in-depth, please find me later. But after two years of doing it this way, every single one of our participants ended up being journalists of color, and not only that, they had geographic diversity, they had age diversity, and they had media diversity in terms of whether they were working in radio, broadcast, print, or online. And after piecing together old emails the best I could, I found that we only received one recommendation that had actually become one of those participants that we had.
And here’s something else that we learned that was totally unexpected: is that blind applications don’t have to be perfectly blind. So as long as you structure your application so that when someone reviews it, and they see everything else first, and then the résumé and the portfolio last, even if you see someone’s name in that process, you can’t remember it. So, in fact, both years, Elena who also co-organizes the Data Institute with me, after seeing the finalists over and over again, we had to force ourselves to stare at pictures and names to remember who was who. But if you told me about them, I could remember everything else about that person. And if you’re worried that blind applications are too much of a change, try it on something smaller.
The first way we did this was at a best practice mentorship conference. See what happens, see how it feels when you start reviewing those applications. So diversity was the number one thing that people listed in the News Nerd Survey for what we needed in this industry. And as people who care about diversity, we care about counteracting how exclusive and discriminatory networks have to be because of their history. I have a request for everyone here: stop sending people’s names so that they can be at the top of the pile. And this is because the work that we should be doing is making sure that the pile has no top. We should be shuffling that pile strategically so that people who apply at the same time don’t even get seen together. We should take care ourselves so that applicants aren’t disadvantaged because we forgot to eat lunch that day. We should only look at so many applications in a given sitting so that the candidate that we see last doesn’t get it first because we were tired. We should be doing everything that we possibly can to make sure that every single one of our applicants, most of whom we don’t know, many of whom are not like us, gets the best shot possible into getting into our programs and getting hired into our jobs. Thank you.
[ Applause ]
RYAN: Thank you so much, Sisi. Are we hacking the system or are we fixing it? What a fantastic question! What a great way to describe what I hope we’re all doing together at this event. David, Nicole, Sisi, thank you so much for taking this into our final afternoon of sessions, can y’all help me thank them one more time?
[ Applause ]
All right. So here we are. It’s our last afternoon. We are eight minutes from the first of our two final stints of breakout sessions where I hope that all of us will start turning our heads forward thinking about what comes next. It’s a chance to think about fixing systems and making plans together. And I want to thank y’all already from the heart, that I know you’re going to take into these last sets of sessions. I also want to tell you at 6:00 p.m., we would really like to welcome you back here again so that we can say goodbye together as a group.
So you have a few minutes to get to where you’re going but I want to thank you for being with us for our talks and I’m really excited for what’s gonna come out of our final group of breakout sessions this afternoon. Thank you.
[ Applause ]