On interviewing and building inclusive teams

When “consistency” backfires

Major tech companies are adopting “structured interviews”, which are shown to be more predictive of on-the-job performance than “unstructured interviews.”

On the surface, it sounds fair. You evaluate candidates by asking everyone the same questions in the same way, and setting consistent expectations across the board.

However, this approach seems incompatible with ongoing efforts to build more diverse and inclusive teams, where it is necessary not just to accept difference, but to celebrate it.

What’s the problem?

1. Celebrating differences requires that we first understand and acknowledge these differences.
2. People who are different will respond differently to different situations and communication styles.

Ay, there’s the rub!

How is it possible to communicate effectively with different people using the exact same language and communication style for each?

• Hypothesis 1. By discounting people’s unique communication preferences during structured interviews, the system selects for candidates with similar communication styles, and weeds out those with dissimilar communication styles (and ways of thinking).
• Hypothesis 2. By following a strict rubric, the process selects only for those who fit within defined boundaries of the role, and does not suitably evaluate the unexpected qualities of each individual. This makes it difficult for teams to become more diverse over time.

Personal perspective as an interviewee

Whenever I’ve been subjected to a “structured” process, I’ve come out of it lacking a sense of belonging.

• It felt like they wanted an automaton, not a human being. They failed to identify my key strengths, which told me that they wouldn’t value or capitalize on my unique capabilities on the job.
• A relationship is a two way street. I find it presumptuous to ask someone to prove themselves to you without also proving yourself to them. This means demonstrating that you are interested in them as a whole being, not just the subset of their capabilities that checks your boxes.
• They treated me how they wanted to be treated, not how I want to be treated. It’s a strange thing to be treated like someone you’re not. It makes it difficult to be yourself.

The problem with “consistency”

The approach of treating different people the same, and expecting the same results is a habit that’s pervasive in our industry, and detrimental to enabling both people and teams to reach their full potential.

While I appreciate the drive for consistency, I’m concerned we’ve gone too far. It shouldn’t be considered special treatment to treat someone in a way that makes them feel comfortable, and brings out their best qualities. It’s being human, and it’s good business.

Growing people is kind of like growing plants. Each person has different instructions for their care and feeding. The fun part is they aren’t written down, and are constantly changing.

As a simple example... If you’re an introvert, imagine being treated like an extrovert would want to be treated — it would feel overwhelming. Conversely, if you’re an extrovert, imagine being treated like an introvert — it would feel de-energizing. Does treating these two distinct people the same really qualify as a “celebration” of differences?

Treating people “consistently” is easier, it sounds good on paper, it sounds good legally, but it creates a false assurance that we are effectively promoting inclusivity and diversity.

What’s the alternative?

People are hard. Maybe we won’t be able to get to perfect, but how can we do better? Interestingly enough, Google already appears to have answered that question with Project Aristotle.

In the study, they went to their most effective teams, and studied why they were so successful. They identified a couple key qualities.

First, on the good teams, members spoke in roughly the same proportion, a phenomenon the researchers referred to as ‘‘equality in distribution of conversational turn-taking.’’
…
Second, the good teams all had high ‘‘average social sensitivity’’ — a fancy way of saying they were skilled at intuiting how others felt based on their tone of voice, their expressions and other nonverbal cues.

Or, more concisely, it comes down to psychological safety.

Within psychology, researchers sometimes colloquially refer to traits like ‘‘conversational turn-taking’’ and ‘‘average social sensitivity’’ as aspects of what’s known as psychological safety — a group culture that the Harvard Business School professor Amy Edmondson defines as a ‘‘shared belief held by members of a team that the team is safe for interpersonal risk-taking.’’ Psychological safety is ‘‘a sense of confidence that the team will not embarrass, reject or punish someone for speaking up,’’ Edmondson wrote in a study published in 1999. ‘‘It describes a team climate characterized by interpersonal trust and mutual respect in which people are comfortable being themselves.’’

Going forward

If we agree that psychological safety is an important characteristic for building effective teams, how can we apply these learnings back to improve the interview process? In particular…

1. How do we make it possible for the candidate to experience psychological safety during the interview itself, so we can get a glimpse of their true potential?
2. How do we demonstrate to candidates that the teams that they are joining will be psychologically safe for them?
3. How do we evaluate a candidate’s ability to create psychological safety for others on the team?

Proposal: semi-structured interviews

A few small, incremental improvements to help make structured interview processes more effective —

1. In addition to “job descriptions”, which focuses on defining a specific individual’s behavior, and encourages a “check-the-box” mentality… define a set of “gaps to fill” on the team. That’s the real problem to be solved.
2. In the rubric for structured interviewing, create blank spaces for “unexpected” attributes. Create a process for discovering the unknown unknowns. Oftentimes I find that discussions about these ambiguities help bring further clarity, and feed into the definition of the role itself.
3. Design choose-your-own-adventure style interview questions that have multiple valid answers, and can be asked in several different ways. This enables you to account for the interviewee’s preferences, which are difficult to determine until the interview itself.
4. As an interviewer, reflect on your own strengths and weaknesses. Ensure you ask questions in a way that you’re comfortable with while also keeping in mind a candidate’s personal communication style. If you can’t be your authentic self, the candidate is likely to mirror that discomfort.
5. Learn how the candidate resolves ambiguity, conflict, and disagreement. This will help evaluate their ability to create a psychologically safe environment for others.

Onwards!

This is a nuanced topic, so consider this a starting point, not The Solution™️. I hope you find my perspective valuable and thought-provoking, and I’d love to continue the discussion with anyone who’s interested.

Questions, compliments, complaints? Anything goes. Shoot me a note on Twitter. Would love to hear your feedback!

This concludes The Diversity Experiment v0.0.2 — if you found it interesting, be sure to check out The Diversity Experiment v0.0.1, which focuses on learnings from behavioral change research. The talk was tailored for the Node.js community, but should be generally applicable to most technical teams.