How does a theory of change differ from a logic model in guiding evaluation?

• A. A logic model defines long-term goals, while a theory of change focuses only on immediate outputs.
• B. A theory of change articulates the causal pathways and underlying assumptions; a logic model visually maps inputs, activities, and outputs but may not fully articulate the mechanisms.
• C. A theory of change is used only for program planning, not evaluation.
• D. A logic model includes verification of all assumptions through experiments.

Answer: (D)

At the heart of this idea is how we describe why change happens and how we test those ideas in evaluation. A theory of change lays out the causal pathways from activities to outcomes and, crucially, the underlying mechanisms and assumptions that link each step. It asks not only what will happen, but why it should happen, given the context, and what conditions must be in place for change to occur. That guiding model helps evaluators design questions and data collection aimed at confirming whether those mechanisms are actually operating and whether the contextual factors support them.

A logic model, on the other hand, is a clear map of the program’s components—resources, activities, outputs, and intended short- and intermediate outcomes. It shows what you plan to do and what you expect to produce, but it doesn’t automatically spell out the deeper causal mechanisms or lay out the full set of assumptions about how and why those links will work. Evaluation can test aspects of the model, but the logic model itself is primarily a schematic of sequence and components rather than a full theory of how change happens.

Therefore, the statement that a logic model includes verification of all assumptions through experiments isn’t accurate. Evaluation may test some assumptions, but a logic model doesn’t inherently function as a verification framework for every assumption. The correct distinction is that a theory of change articulates causal pathways and assumptions, while a logic model maps inputs, activities, and outputs with less emphasis on detailing mechanisms.