I want you to imagine that you are a Child Protective Services worker. And you have to respond to a report of child abuse. You walk into a home, unannounced, unexpected, certainly uninvited. The first thing you see is a mattress in the middle of the room, on the floor. Three kids lying on it, asleep. There's a small table nearby with a couple of ashtrays, empty beer cans. Large rat traps are set in the corner, not too far from where the kids lie asleep. So you make a note. A part of your job is walking through the entire home. So you start with the kitchen, where there's very little food. You notice another mattress in the bedroom, on the floor, that the mother shares with her infant child.
﻿我想请各位把自己想象成 儿童保护协会的工作人员，  你现在要调查一起虐童案。 你走进一个家，事先未通知其家人， 让人毫无防备。 首先映入眼帘的是，房间正中央 的地板上，放着一张床褥， 上面躺着三个小孩，熟睡着。 附近是一张小桌子， 放着几个烟灰缸， 空啤酒罐， 角落还安置着几个巨大的捕鼠夹， 就在离孩子们睡觉不远处。 所以你记下了这一切。 你工作的一部分，是走遍整个房子。 你从厨房开始， 发现食物了了无几， 你还注意到卧室地板上的 另一张床褥，  是母亲和她的婴孩睡觉的地方。
Now, generally, at this point, two things may happen. The children are deemed unsafe and removed from the home, and placed in state custody for a specified period of time. Or the children remain with their family and the child welfare system provides help and support.
这时，一般会有两种情况发生。 孩子们的生存条件被视为受到威胁， 所以不得不送去寄养所， 在特定的一段时间内， 由州政府代为监护。 或孩子继续与家人待在一起， 但会收到儿童福利处的帮助和支持。
When I was a Child Protective Services worker, I saw things like this all the time. Some far better, some far worse. I asked you to imagine yourself in that home, because I wonder what crossed your mind. What guides your decisions? What's going to impact your opinion of that family? What race, ethnicity, did you think the family was?
我还在儿童保护服务处工作的时候， 这类事件层出不穷。 情况有时候可能好得多， 有时候差得多。 我让各位想象自己身临其境， 是因为我想知道， 大家当时会想什么。 是什么引导大家做出决定？ 是什么影响大家对那个家庭的看法？ 各位觉得这家人属于什么种族， 民族?
I want you to realize that if those children were white, it is more likely that their family stays together after that visit. Research done at the University of Pennsylvania found that white families, on average, have access to more help and more support from the child welfare system. And their cases are less likely to go through a full investigation. But on the other hand, if those kids are black, they are four times more likely to be removed, they spend longer periods of time in foster care, and it's harder to find them a stable foster placement.
我想让大家知道， 如果那些孩子是白人的话， 在那次家访后， 他们与家人待在一起的几率会更高。 根据宾夕法尼亚大学的 一项研究调查， 白人家庭总体而言，会从儿童福利体系中 得到帮助和支持更多。 而且他们的案件也不太需要 经过全方位调查。 但另一方面，如果那些孩子是黑人， 他们送去寄养所的几率是 白人孩子的四倍之高， 而这些孩子待在寄养所的时间 也更长， 也更难找到一个稳定的寄养的地方。
Foster care is meant to be an immediate shelter of protection for kids who are at high risk. But it's also a confusing and traumatic exit from the family. Research done at the University of Minnesota found that kids who went through foster care had more behavioral problems and internalized issues than kids who remain with their families while receiving help and support.
寄养服务本来是为严重受到威胁的  孩子们提供紧急住所。 但就这样将孩子从父母身边带走， 不仅令人困惑，还会造成巨大痛苦。 根据明尼苏达大学的一项调查显示， 曾有过寄养经历的孩子， 比那些继续和父母亲待在一起， 接受后续帮助和支持的孩子们， 更容易有行为问题和性格问题。
The scenario I mentioned earlier is not uncommon. A single mother, living in low-income housing with her four children. And the rats make it almost impossible to keep food, let alone fresh food in the home. Does that mother deserve to have her children taken from her?
我先前提到的情境屡见不鲜。 一个单身母亲，和她的四个孩子， 生活在一个低收入人群的房子里， 鼠患让食物尚且难以保存， 更不用说要在家里找到新鲜的食物了。 但难道仅凭这一点， 母亲就该跟她的孩子骨肉分离吗？
Emma Ketteringham, a family court attorney, says that if you live in a poor neighborhood, then you better be a perfect parent. She says that we place unfair, often unreachable standards on parents who are raising their kids with very little money. And their neighborhood and ethnicity impact whether or not their kids are removed. In the two years I spent on the front lines of child welfare, I made high-stakes decisions. And I saw firsthand how my personal values impacted my work.
Emma Ketteringham 是一名家庭事务律师， 她说如果你住在一个贫穷的街区， 那么你最好成为一个完美的父母亲， 她说，我们的社会对于那些 低收入的父母而言， 有着一系列不公平且 难以达到的抚养孩子的标准。 而他们的所住的街区和所属的民族， 直接影响到他们的孩子是否 会被送去寄养所。 我在儿童福利的前线上工作了两年， 我做的决定都是高风险的。 我切身体会过我的个人价值观 是如何影响我工作的。
Now, as social work faculty at Florida State University, I lead an institute that curates the most innovative and effective child welfare research. And research tells us that there are twice as many black kids in foster care, twenty-eight percent, than there are in the general population, 14 percent. And although there are several reasons why, I want to discuss one reason today: implicit bias.
现在，作为弗洛里达州大学里 社会工作系的一名教师， 我领导的学院， 进行的是最创新最有成效的 儿童福利研究。 研究结果显示，寄养所里的黑人儿童 是其它种族儿童的两倍，  有28%这么多。 而其它种族的儿童也就才14%。 虽然其中有着许多原因， 但我今天只想谈论一个理由： 隐性偏见。
Let's start with "implicit." It's subconscious, something you're not aware of. Bias -- those stereotypes and attitudes that we all have about certain groups of people. So, implicit bias is what lurks in the background of every decision that we make. So how can we fix it?
让我们从“隐性”开始说起， 它是潜意识的，你察觉不到的。 而偏见 -- 就是对于特定人群的 刻板印象和态度，这是我们都有的。 所以，隐性偏见指的就是潜藏在 我们做出的每一个决定的背景因素。 那我们该如何解决这个问题呢？
I have a promising solution that I want to share. Now, in almost every state, there are high numbers of black kids going into foster care. But data revealed that Nassau County, a community in New York, had managed to decrease the number of black kids being removed. And in 2016, I went into that community with my team and led a research study, discovering the use of blind removal meetings.
我有一个很有前景的解决方案， 想和大家分享。 现在，几乎每一个州， 都有着不胜其数的黑人儿童 被送进了寄养所。 但数据显示，拿骚县， 纽约的一个县城， 已经成功减少了许多将黑人儿童 送去寄养所的案例了。 2016年，我和我的团队走进拿骚县， 领导一项调查研究， 看看隐去孩童个人信息后 再进行寄养汇报的成效如何。
This is how it works. A case worker responds to a report of child abuse. They go out to the home, but before the children are removed, the case worker must come back to the office and present what they found. But here's the distinction: When they present to the committee, they delete names, ethnicity, neighborhood, race, all identifiable information. They focus on what happened, family strength, relevant history and the parents' ability to protect the child. With that information, the committee makes a recommendation, never knowing the race of the family. Blind removals have made a drastic impact in that community. In 2011, 57 percent of the kids going into foster care were black. But after five years of blind removals, that is down to 21 percent.
这个汇报是这样进行的： 社会工作者受命调查一起虐童案件， 他们前往受访家庭， 但是在决定让孩子搬去寄养所前， 这名社会工作者必须先回到办公室， 汇报他们的调查结果。 但与以往不同的是， 当工作人员向组委会汇报结果时， 他们隐去了受访家庭的姓名， 民族，所住街区，种族， 以及所有能透露身份的信息。 他们把重点放在事件的经历， 家庭的优势，相关的历史， 和父母保护孩子的能力上。 在听完汇报后， 组委会将提出一项建议， 他们至始至终不会知道 受访家庭的种族。 这项举措对拿骚县造成了 巨大的影响。 2011年的时候，有57%被送进 寄养所的都是黑人儿童， 但在隐去孩童的个人信息后， 这个数据降到了21%。
(Applause)
（掌声）
Here's what we learned from talking to some of the case workers. "When a family has a history with the department, many of us hold that history against them, even if they're trying to do things differently." "When I see a case from a certain apartment building, neighborhood or zip code, I just automatically think the worst." "Child welfare is very subjective, because it's an emotional field. There's no one who doesn't have emotions around this work. And it's very hard to leave all of your stuff at the door when you do this work. So let's take the subjectivity of race and neighborhood out of it, and you might get different outcomes."
通过与社会工作者的对话， 我们了解到， “一个家庭如果曾经 在社会工作部门留下不良记录后， 我们大部分社会工作者 就会以此否定这个家庭做出的努力， 即便他们已经在尽力改变了。” “当我只要看到这个案件发生在 某个特定的公寓， 街区，或邮编， 我就会不由自主往最坏的方面想。” “儿童福利是一项非常主观的工作， 因为这个领域牵动着我们的情绪， 从事这项工作的人， 没有一个是可以没有任何情绪的。 在你执行工作之前， 也很难让你将所有的情绪留在门外。 所以，让我们将种族和社区这些 主观因素排除在外， 你所得出的结果也许会有所不同。
Blind removals seem to be bringing us closer to solving the problem of implicit bias in foster-care decisions. My next step is figuring out how to use artificial intelligence and machine learning to bring this project to scale and make it more accessible to other states. I know we can transform child welfare. We can hold organizations accountable to developing the social consciousness of their employees. We can hold ourselves accountable to making sure our decisions are driven by ethics and safety. Let's imagine a child welfare system that focuses on partnering with parents, empowering families, and no longer see poverty as failure. Let's work together to build a system that wants to make families stronger instead of pulling them apart.
隐去孩童的个人信息，似乎将我们 更紧密的联系在一起， 去解决这个影响着寄养决定的 隐性偏见的问题，  下一步，我要找出办法， 使用人工智能和机器学习 让这个方法更加容易 大规模的运用到别的州上。 我知道，我们是有能力变革 儿童福利体系的。 我们要让相关的机构负起责任， 对他们的社会工作人员 进行社会意识培训。 我们应该让自己承担起相应的责任， 确保我们每个决定 都符合道德规范，为安全着想。  让我们想象一下这样一个美好的 儿童福利体系：致力于与父母合作， 给家庭赋权， 不再视贫穷为失败。 让我们一起建立一个 让家庭更加团结，而不是让其 支离破碎的儿童福利体系吧！
Thank you.
谢谢。
(Applause) (Cheering)
（掌声）（欢呼）