Mitigate Biases with Rubrics

When I first started working at Quora and I was told to interview people, they just threw me into a room and told me to come out with a yes or no answer. Which is very standard for startups. It was like, just go. There's no process yet. But much better practice would be actually having rubrics and having criteria for what you're looking for that helps to standardize. It also helps remove a lot of bias. So it's been interesting academic research on this. If you don't ask people to specify what they're looking for up front, they'll let their biases come in, they'll justify it later. There's a really interesting study about hiring for police chiefs where people were shown resumes like one of a man, one of a woman. And in one scenario, the man had more experience, the woman had more academic background, and the people choosing would choose the man and they would justify it by saying, oh, we want somebody with more work experience. That's what this job needs. And then the other scenario, the man had more academic experience, the woman had more work experience, they still chose a man and said, well, we just really need more academic experience in this case. So they would back into their choices with these justifications. But when the people choosing were told to write down their criteria ahead of time, they would be consistent with that. So if they said they wanted to have the academic background, then when they were presented with the two resumes and the woman had more of an academic background, they would choose her. So there's actually a lot of really great research about how do you mitigate your biases? And some of these things, they just make more sense. It's not just about diversity and inclusion, it's just making these processes better. So writing down rubrics around not just hiring, but also promotion, very valuable. A lot of times in companies when you're thinking about who's going to lead some new projects or who's going to work on this interesting new opportunity, there's not a lot of process around it. And people would just think, who do I know who might be interested in this, who might be good at this? And we tend to fall into the same biases there. It's written out when we're choosing a tech lead, these are the things we're looking for and this is our process. We're going to look at everybody who's at a certain level and has been at the company for a certain amount of time and consider them. Then you actually end up with a much better decision. And it has a side benefit of helping diversity and inclusion people from the underrepresented backgrounds.