What the Data Said About HOP

A few years ago, at a large global biopharmaceutical company, a piece of critical equipment was validated without the sample valves needed for product testing and release. Multiple functional groups had been involved. Subject matter experts had reviewed the work. Everyone was trying to do it right. And the error still happened.

That event became the catalyst for a Human and Organizational Performance implementation we led at the company. We tested new investigation methods, new risk practices on the floor, new training, and led a multi-year shift in how the organization thought about error. At the time, we believed it was working. We saw it in the deviation trends, in how investigators talked about events, in how leaders responded when things went wrong.

Belief is one thing. Data is another. This year, a Harvard University Master's thesis set out to test whether it actually happened.

The Question Behind the Program

The thesis, written by Christopher Mastrangelo through Harvard's Division of Continuing Education, asked a specific question: does implementing HOP principles measurably move an organization toward becoming a learning organization? Not whether people liked the training. Not whether leadership felt good about it. Whether the underlying data — deviation rates, root cause categories, CAPA types, and organizational maturity scores — actually changed in the direction HOP predicts.

To answer it, the study pulled five years of quality management system data and cross-referenced it against a Human Performance Assessment conducted annually from 2020 through 2024.

What the Numbers Showed

The deviation initiation rate declined steadily from 2020 through 2025, tracking closely against HOP implementation milestones. The rollout of Risk Ready Reviews, Pre-Job Briefs, Touch-Say-Do practices, and the shift away from "human error" as an acceptable root cause were all parts of this cultural shift.

More telling was what happened underneath that number. Root cause categorization shifted hard, away from Human Factors and toward Method-based causes. Over time, investigators became increasingly skilled at finding process and system explanations rather than stopping at "the operator made a mistake." CAPA data told a matching story: Change Controls, which fix systems, rose as a share of corrective actions while retraining, which targets individuals, declined.

Read together, this is the fingerprint of an organization that stopped asking "who did this?" and started asking "what let this happen?"

The Part We Didn't Expect

The most useful part of the study wasn't the growth. It was what happened when the program lost its sponsor.

Starting in late 2022, the HOP team was scaled down. In mid-2024, it was formally dissolved. The thesis captured that transition and measured it, and the data split cleanly. Training maturity and leadership engagement scores dropped, and Human Factors began creeping back up as a root cause category. Those were the parts of the program that depended on someone owning them.

But investigation methods, proactive learning, and floor-level risk practices held. Gemba walks were still happening. Event and Causal Factor charting was still the default tool. Workers still weren't describing themselves as blamed. Those were the parts that had become how people actually worked, not a program someone was running.

That distinction matters more than the headline trend. A culture shift that only survives while someone is actively managing it isn't a culture shift. It's a project. The parts of HOP that outlasted the team that built it are the parts that were genuinely embedded, and the parts that eroded are a fairly precise map of where sustained attention still needs to go.

Why This Matters Beyond One Company

Independent, peer-reviewed confirmation of this kind is rare in this field. Most of what gets said about Human and Organizational Performance is directional: case studies, conference talks, practitioner conviction. This is a Master's thesis built on five years of primary quality system data, testing a specific hypothesis against actual outcomes, at a scale most organizations never get to see analyzed this closely.

It confirmed what those of us who built the program experienced on the floor: HOP isn't just a philosophy that feels right. Implemented well, it changes measurable behavior, and enough of it sticks to survive real organizational disruption.

You can read the full thesis, Assessment of Efficacy of Human and Organizational Performance Methodology at a Large Biopharma Company, on Harvard's DASH repository: dash.harvard.edu/handle/1/42731250

--

Want to build HOP practices that survive leadership turnover, not just launch day?

The Operational Learning Institute works with organizations to design and embed Human and Organizational Performance programs built for the long run.

Learn more about our programs