What counts as a significant change to an evaluation plan?
A significant change to an evaluation plan is generally a change to the evaluation questions, the outcomes being measured, the methods, the people being asked, or the budget. Your funder decides what needs approval, so check with your program officer before you make one.
Funders draw these lines in their own way. NSF's award rules are one example. They say a project's objectives or scope may not be changed without NSF's approval in advance, while a significant change in methods or procedures must be reported to the right officials at your institution and to NSF. Your program officer can tell you which kind a particular change to your project and its evaluation is. Other funders may set different rules.
Smaller adjustments come up too, such as a data collection date that has to move with the program's schedule. At Grantibly, we adjust evaluation schedules when a project's timeline changes and keep the evaluation in line with what the funder requires.
How do we document what changed and why?
Document a program change by recording what changed, why, and what happened as a result, and update the logic model and the evaluation plan together. Record the change when it happens, while the reasons are still fresh. Your evaluator can help you decide what to capture and can fold it into the evaluation, so the record becomes part of the evidence about how the program was carried out.
A logic model written for a proposal describes how the project was expected to work at the time. Revisiting it when the program changes keeps the evaluation questions connected to what the project is doing now. What's a logic model, and does my proposal really need one? explains what a logic model is and what it's for. The change may also need to appear in your next report to the funder, and What does my funder's annual report actually need from the evaluation? covers that report.
What happens to the outcomes we promised in the proposal?
If an outcome you promised in the proposal no longer fits the program, talk with your program officer. The evaluation should document the change and help explain it.
That explanation matters because an outcome measured against a program that no longer exists can be hard to interpret. If a recruitment strategy changed, for example, an enrollment outcome written for the original strategy may no longer describe what the project is trying to do. The evaluation can show what changed, why, and what the project is now working toward. Depending on your funder, changing a promised outcome may count as a change in scope that needs approval in advance, so don't change it on your own.
Whether results that fall short could affect your funding is a separate question, and Will an evaluation report that shows our program isn't working cost us our grant? takes it up.
The Grantibly perspective
Lauren's perspective
A logic model created during proposal development is a plan. It isn't a prediction of exactly what will happen over the next three, four, or five years. Programs evolve. Staff change, partners change, recruitment strategies change, and students or participants respond differently than expected. A strategy that seemed promising during proposal development may need to be redesigned once implementation begins.
Evaluation should document those changes and help teams understand them, rather than treating every adaptation as a deviation that needs to be defended. The questions that matter are: What changed? Why? What happened as a result? What did we learn? That's also why we want clients to call us when they're considering a change, and not only after it's made.
