Python has never attended a community meeting.
Neither has R. And Excel – bless its heart – has never explained why a program succeeded.
Yet somewhere along the way, we’ve started treating data analysis and evaluation as though they’re interchangeable. They’re not.
Data analysis is an essential part of evaluation, but it is only one part. It’s the instrument – not the orchestra.
Evaluation begins long before the first dataset is cleaned or the first regression is run. It begins by asking the right questions. What are we trying to understand? What does success actually look like? Compared to what? For whom? Under what conditions? And perhaps the most important question of all: Why?
This is where frameworks matter. A good evaluation framework does more than organize indicators – it provides a way of thinking. It helps us decide what deserves attention, what evidence matters, and how seemingly unrelated pieces fit together. Without a framework, even the most sophisticated analyses can become little more than interesting observations.
Theory matters. Programs are built on assumptions about how change happens. Evaluation tests those assumptions. Without theory, we may know what happened, but we’ll struggle to explain why it happened – or why it didn’t. And if we can’t explain the “why,” improving future programs becomes little more than educated guesswork.
Then there’s context. A program that thrives in one institution may struggle in another, not because the model is flawed, but because the people, resources, leadership, history, or environment are different. Evaluation that ignores context risks drawing conclusions that are technically correct yet practically misleading.
And culture. Culture isn’t an “add-on” or a checkbox. It shapes how people experience programs, situations, and places and how they define success, what they choose to share, and even whether our methods are appropriate. We don’t simply collect data from communities; we interpret evidence about people. That requires humility, curiosity, and cultural awareness.
Can anyone become an evaluator? Absolutely.
Just as anyone can become an architect, engineer, or a lawyer, evaluation is a profession that can be learned. It can be practiced. It needs to be done well. It requires more than mastering software or producing attractive dashboards. It requires reading, depth, understanding, and using methodology, theory, frameworks, context, culture, ethics, and the art of asking questions that matter.
When evaluation is reduced to “someone who can simply analyze the data and write a report,” we don’t just shortchange evaluators. We shortchange the programs seeking to improve, the funders making decisions, the profession itself and – most importantly – the communities whose lives those decisions affect.
The next time someone says, “We just need someone to analyze the data,” smile.
Then remember:
A spreadsheet has never walked into a community meeting.
A regression model has never built trust.
And a dashboard has not changed a life.
People do.
Evaluation helps us understand how.
