What Is Integrative Capacity and Why Does It Matter for Research Leaders?

The team is assembled. Everyone is credentialed, motivated, and genuinely glad to be working together. The Request for Proposals seems like a great opportunity for joint funding.

Six months in, the timeline starts slipping. Not because anyone is uncommitted. The science each person is doing individually is still good and moving along. The problem, however, is harder to name: the investigators and their labs are doing parallel rather than integrated work. Data isn't crossing between labs when it should. A co-investigator runs a sub-study that overlaps with the analysis another site is already conducting. The steering committee meets monthly and makes decisions that don't seem to change anything at the bench.

Nobody is difficult. Nobody is checking out. The collaboration is just not working in a coordinated way.

This is one of the more common and underdiagnosed problems in multi-investigator research. When teams like this underperform, the explanations that surface first are usually about personalities: a communication style mismatch, a co-investigator who's hard to schedule, a trainee who doesn't ask questions. Those things happen. But more often, the underlying problem is structural. The team has expertise. What it lacks is integration.

That distinction matters because interpersonal fixes and structural ones require different interventions.

Two forms of capacity that both have to be present

The framework I published with Eduardo Salas and colleagues in American Psychologist in 2012 identified what separates scientific teams that produce genuinely integrated knowledge from those that produce parallel contributions that never quite connect.

The answer came down to two things.

The first is Capacity Within: What each team member brings individually. Their depth of expertise. Their ability to grasp work that lives outside their training. Their willingness to let a collaborator's framing change how they think about their own question. You can observe it in small ways. Does a scientist translate their own work for a non-specialist without being asked? Do they come to cross-disciplinary meetings with actual questions, not just updates? Can they articulate the team's collective contribution, not just their piece of it?

The second is Capacity Across: What the team builds together. Shared mental models. Coordination infrastructure. A decision architecture that enables distributed teams to work toward a common scientific goal without constant bottlenecks or redundant effort. It shows up in whether the team has a process for resolving scientific disagreements before they become resentments. Whether a new postdoc joining the project can figure out who decides what and where they can contribute. Whether the collaboration plan describes how the team will work together, not just what they intend to produce.

Both have to be present. A team with deep individual expertise but no integration infrastructure produces parallel publications. A team with strong coordination systems but shallow boundary-crossing produces activity without insight. The research on this has been consistent enough to appear not just in lab science contexts but also in mixed-methods teams, clinical research networks, and training programs for translational scientists.

Why Integration is Difficult in Academic Research

The systems that train and evaluate scientists are almost entirely focused on individual contribution. A CV documents an individual's output. A letter of recommendation assesses an individual's potential. A grant review scores whether each co-investigator is qualified for their specific role.

None of these standard measures assess whether the team will actually integrate.

This means a team can pass every individual quality threshold and still fall apart at the collaboration layer. And because the failure usually looks like slowness, drift, or duplicated effort rather than a clear breakdown, it often gets attributed to the wrong things.

Part of what drove the Collaborative Contributions Checklist work I've done with colleagues at UCI is precisely this gap: the absence of shared language for articulating what collaborative contributions actually look like, which makes it hard for people doing integration work to describe it and for institutions to recognize when it is taking place. The CCL is one tool for making that invisible work legible. Building Integrative Capacity is what makes it worth doing.

What Building Integrative Capacity Requires

Integrative Capacity is infrastructure, not chemistry. You build it the way you build any other infrastructure: deliberately, in advance, with attention to the specific places where things tend to break.

For a research team, that means three things.

First, getting clear about what the team is actually trying to integrate. Not just the scientific aims. The workflows, the data pipelines, the decision rights. Where does information need to move across sites or disciplines, and does the current structure support that movement? Where are the handoffs, and who owns them?

Second, building the relational infrastructure that enables cross-site work. In work with colleagues at UCI's Team Science Acceleration Lab, we identified several institutional leverage points that affect whether teams can sustain this kind of work: how credit gets allocated, how conflict surfaces, and how norms get established before problems arise rather than after. Laboratories and research centers that invest in this work before a collaboration starts typically function differently from those that try to build it during a funding cycle when the pressure is already on.

Third, developing the leadership practices that hold integration together over time. Managing your own lab well does not prepare you to lead a multi-site team. The skills are related but not identical. Coordinating across distributed authority structures, building trust with co-investigators who have their own labs and their own legitimate scientific priorities, staying oriented toward a shared goal when the components are moving in multiple directions at once: these require a different kind of capacity than the one that got you funded.

The question worth asking about your team

If you're leading a multi-investigator collaboration now, or preparing to build one, the most useful question isn't whether you have the right people. You probably do.

The question is whether the architecture that enables those people to work together is in place to drive joint work.

That assessment is what the Research Team Integration Diagnostic is intended to assess. If you want to go deeper, that's what the Research Career Accelerator is designed to support.

Funded projects deserve to actually work, and grand challenges deserve teams that can integrate to advance scientific solutions.