9 Lab Culture, Psychological Safety, and Team Science
Productive science depends as much on human interactions, communication, and group dynamics as on technical excellence. This chapter describes the culture we are building and the specific practices that protect it.
9.1 Psychological safety is the foundation
Psychological safety means members feel able to ask naive questions, admit they don’t know something, surface mistakes early, and disagree with the PI — without fear of looking foolish or paying a price. It is the single strongest predictor of whether a team learns and performs.
For a group where people are learning coding, statistics, and environmental-health methods at the same time, this is non-negotiable. We normalize “learning in public”:
- Questions are contributions: the person asking the “obvious” question is usually asking what three others were thinking.
- Mistakes surfaced early are cheap; mistakes hidden are expensive. We thank people for catching errors — including their own.
- Disagreement with the PI or senior members is welcome and expected. Good science needs dissent.
9.2 Leaders set the norms
Members emulate leadership behavior. The PI and senior members must model the culture they want — especially in the moments that test it: during authorship discussions, revisions, disagreements, and deadline pressure. If we want a team that is direct, rigorous, collaborative, and fair, leadership has to be visibly direct, rigorous, collaborative, and fair first.
Concretely, leaders:
- Admit their own mistakes and uncertainty out loud.
- Solicit input before stating their own view, so they don’t anchor the room.
- Resist dominating discussion; senior scientists can crowd out everyone else without meaning to.
- Give credit generously and specifically.
9.3 Team science, by design
We work across disciplines, institutions, datasets, and time zones. That requires intentional team-science practice:
- Shared narrative. The Community-to-Clinic-to-Cloud story is the operating narrative that lets people on different datasets feel part of one program.
- Clear interfaces. Pod charters and the project pod model define who owns what, so collaboration doesn’t dissolve into “who’s doing this?”
- Co-mentoring and distributed capacity. We draw on collaborators’ expertise rather than pretending any one person knows everything.
- Delegation as development. We deliberately hand off lab-meeting leadership, journal-club facilitation, figure ownership, codebook stewardship, literature synthesis, and dashboard building. Delegation builds trust and motivation — and trains leaders.
9.4 Meetings that invite participation
Meetings are where culture is most visible. Ours are designed so junior members speak, question data, and critique safely.
- Rotating roles. Facilitator, note-taker, and “first question” rotate so the same voices don’t always lead.
- Structured turns. We make space for students to present and to ask questions before senior members weigh in.
- Safe critique. We critique the work, not the person, and we praise good questions as much as good answers.
- Wins & roadblocks. A recurring segment surfaces both progress and stuck points without stigma.
(See Chapter 10 for channels, cadence, and agenda templates.)
9.5 Inclusion is part of the culture, not the branding
Leaders shape cultures that recognize and eliminate discriminatory or harassing behavior. This aligns directly with our mission to broaden participation in STEM. Inclusion shows up here in concrete practices: equitable recruitment, rotating visible roles, attention to who speaks and who is credited, bilingual communication, and a low-friction path to raise concerns (Chapter 18). We do not tolerate harassment or discrimination of any kind (see Code of Conduct).
9.6 We evaluate the lab as a human system
We already evaluate projects scientifically. We also evaluate the group as a human system. Each quarter we run a brief, anonymous lab climate check covering workload, clarity, belonging, mentorship quality, confidence, and barriers. Each year we run a fuller stop / start / continue review (Chapter 21). Self-awareness, feedback, and adaptation are leadership responsibilities, not optional extras.
Can the most junior person in the room disagree with the most senior person about the data — and have it go well? If yes, the culture is working. If you’re ever unsure, that’s worth raising.