The hardest part of leading a research team is not the research. It is figuring out what makes each person want to do their best work—and then building the conditions for that to happen.
When I became responsible for mentoring junior researchers at the University of Washington, I realized that managing people well requires the same rigor I bring to any other problem. So I did what I always do: I read. Daniel Pink’s Drive gave me the framework—mastery, autonomy, and purpose as the three pillars of intrinsic motivation. Deci and Ryan’s self-determination theory gave me the science behind it. Seligman’s Flourish and Authentic Happiness helped me understand what it means for people to thrive at work, not just perform. And Csikszentmihalyi’s Flow showed me what the optimal state looks like—the zone where challenge matches skill and time disappears.
These were not books I read for personal enrichment. They were diagnostic tools. I was trying to understand why some of my junior colleagues were disengaged, why talented people sometimes produced mediocre work, and what I could do about it. The answer, as it turned out, was not to manage them harder. It was to give them more of the three things that research consistently shows people need: a sense of growing competence, control over how they do their work, and an understanding of why it matters.
The most concrete expression of this philosophy was how I structured learning. When a research assistant joined the QUIC validation project—a complex effort to link Cultivate Learning’s new classroom quality tool with the established CLASS observation system—I did not hand them the data and say “figure it out.” Nor did I write the code for them. I wrote assignments.
The first assignment was just setting up RStudio. Install R, install the IDE, verify the installation works. The second introduced data aggregation: tabulate a variable, check for missing values, compute means within groups using dplyr. The third taught merging: use intersect() to find common identifiers, use merge() to join two datasets, verify the row count matches. Each assignment built on the previous one, and each came with hints rather than answers. I was explicit about why: the process of solving problems is the best training our research assistants can receive. Navigating challenges builds self-efficacy—the confidence that you can figure things out—and that is more valuable than any particular solution.
By the eighth assignment, the same person who had started by installing RStudio was now navigating three different SQL tables containing 180,000 observations, reasoning about which idRating from the QUIC system corresponded to which CLASS observation, drawing entity-relationship diagrams on whiteboards, and running regressions to validate that the linked data produced statistically meaningful associations. The gap between Assignment 1 and Assignment 8 was enormous, but because each step was calibrated to be just beyond their current ability—what Vygotsky called the zone of proximal development, what Csikszentmihalyi would recognize as the flow channel—it never felt impossible.
Scaffolding covers mastery, but mastery alone is not enough. People also need autonomy—the sense that they have genuine choice in how they approach their work—and meaning—an understanding of how their contribution fits the larger picture. I tried to provide both.
For autonomy, I learned which topics each person was drawn to and assigned them accordingly. If someone was interested in survey methodology, they worked on the survey. If someone cared about equity, they got the pay equity analysis. This was not just kindness; it was efficiency. People do better work on problems they care about, and the quality difference between assigned-interest and assigned-random is not marginal—it is dramatic.
For meaning, I made sure every research assistant understood how their specific task connected to the whole project. Not just “clean this dataset” but “this dataset will let us determine whether QUIC scores predict the same classroom quality that CLASS measures, which will tell the state whether their new tool is valid, which will affect how 3,000 child care programs are rated.” When people understand the stakes, the work changes character. A merge is no longer a merge. It is the link between two measurement systems that will shape policy for thousands of children.
The final piece was looking beyond the current project to the person’s career. For junior researchers in academia, publications are currency, and authorship order is the denomination. I fought for my mentees to be listed first on papers they contributed to substantively—even when convention or seniority might have suggested otherwise.
The clearest example is the “Early childhood educators’ pay equity: A dream deferred” paper, published in the Early Childhood Education Journal in 2023. Liu, a junior colleague, had done significant work on the analysis. I advocated for her to be first author and moved myself toward the back of the author list to compensate the rest of the team for the change. The paper was going to be published regardless of the order. But for Liu, first authorship meant something tangible on her CV—a signal to future hiring committees that she led the work, not just contributed to it.
This was not altruism. It was strategy informed by everything I had read about motivation. When people see that their leader is invested in their growth—not in the abstract, but in concrete, career-advancing ways—their commitment to the work deepens. The next dataset gets cleaned more carefully. The next literature review is more thorough. The entire team operates at a higher level because the incentive structure rewards genuine contribution rather than hierarchy.
The most ambitious expression of this mentoring philosophy was the 125-page ECE Research Manual I wrote for the team. It covered everything a junior early childhood education researcher might need: from formulating research questions and writing literature reviews, through hypothesis testing and causal inference, to R programming, effect sizes, power analysis, IRB protocols, and journal selection. It was a curriculum, not a reference—designed to be read front to back by someone entering the field, with each chapter building the foundation for the next.
The manual exists because I noticed the same questions coming up repeatedly across different mentees. Rather than answer each one individually—efficiently for me, but poorly for their learning—I wrote the answer once in a form that invited self-study. The manual was a hint system for an entire career: here is what you need to know, here is where to look deeper, here is why it matters.
- Scaffolded learning design — Progressive assignments calibrated to the zone of proximal development, with hints rather than answers
- Self-determination theory — Structuring work to maximize mastery, autonomy, and meaning (Deci & Ryan, Pink)
- First-author advocacy — Fighting for junior colleagues’ authorship credit to advance their careers
- Interest-matched assignment — Aligning tasks with individual interests and career goals
- Curriculum development — 125-page ECE Research Manual covering the full research lifecycle