Excerpts from The Logic Model Guidebook

The Guidebook provides the practical support you need to create and use models. It will also enhance your understanding of the relationships between actions and results. Step by step, we describe logic models as both a tool and a process that resonates with learning and evidence-based design.

Logic models are a visual method of presenting an idea… Common synonyms for models include idea maps, rich pictures, action maps, and mental models.

A logic model, is no guarantee of logic. While many models do demonstrate some modicum of logic, a logical presentation does not equal plausibility, feasibility or success. There is some danger in seeing a display and considering it “true.”

It is critical to recognize the role of beliefs…Assumptions are often informed by knowledge. Sometimes assumptions are informed by experiences, habits or values that do not also reflect knowledge…Dogma, misinformation, ignorance and wishful thinking are hazards here.

Program logic models display what an existing idea, new program, or focused change effort might contain from start to finish. The elements in a program logic model consist of the recipe for a bounded investment of financial and social capital for a specified result. The level of detail increases so that the relationships shown by the model illustrate essential linkages needed to make a plan fully operational for each of the strategy strands identified in the theory of change. The primary elements for each strand of a program logic model include resources, activities, outputs, outcomes, and impact. Figure 3.2 is a template of the elements for most program logic models.

The process of modeling supports better thinking about a given idea or effort. It can establish routines wherein alternative possibilities are considered and explored. Modeling may be an important antidote to snap judgments…Sometimes errors in scale are an authentic oversight. Other times, the politics of context are the origin. It is quite common for agendas that include marketing or positioning to supersede what is feasible. Models can be very persuasive.

Models improve most dramatically when introduced to colleagues or external sources that are not directly invested in the models’ elements or their implementation…In this way “disinvested” parties can offer objective critiques which often uncover blind spots, identify weaknesses, flaws, leaps of faith, ambiguities, “cockamamie,” and fiction.

The logic model serves as the focal point for discussion about evaluation because it displays when, where, and how to look for the information most needed to manage the program and determine its effectiveness…It follows, then, that management is sufficiently evaluation literate and takes the time to reflect on the meaning and significance of evaluation data. For evaluation to be truly useful, its findings must be relevant, understood, and applied by program staff.

One important concept frequently overlooked is that in order to use evaluation to improve effectiveness, you need to engage in learning. ..Single loop learning does not question the assumptions and evidence…The original thinking is given and not examined…Double loop learning questions when activities do not seem to produce intended results. The strategies or the likely outcomes may be questioned or altered. Evaluation that serves double loop learning tests the theory of change…to improve the logic behind the design or implementation that drives implementation.

This chapter describes selected examples of logic model use. Through brief cases we present models as used in private and public sector organizations. The variation in format and content are intentional. These models, presented in the context of cases, are provided to enrich readers’ experience and experimentation with features of display.

• Identify variations in model format and style.
• Recognize that models reflect culture and intended use.
• Explore what will and will not work in your organization.
• Explain why logic models are highly interpretive.

Because logic models are socially constructed, perception, politics, and persuasion are all substantial influences on them. As a graphic display of the general approach to change or as a more detailed description of work, logic models reflect intentional choices of their authors. In reality, models can be compromised by the skills and experiences of their creators, along with the context and purposes they serve. Our own models reflect these influences. For example, sometimes clients do not ask for or want modeling (“improved versions”). Although Chapters 2–5 in this text have suggested quality features and selected principles for creating models, the examples here vary in adherence. As the use of logic models grows, it is possible that standards for them will emerge and be commonly used. Please note that the cases in this chapter include model examples that successfully served specific purposes. We identify variation of some key features as important context before sharing the six cases.

This chapter suggests readers consider the potent value archetypes can give to their own models. Archetypes are a tested, general template for an intervention, program, or strategy. Often, with modification, they can inform your planning, evaluation, communication, or other needs. Archetypes can also provoke new thinking and provide a quality check that improves ideas.

Why squander the knowledge we have about what works? Many important services, products, and programs have been built on the good efforts of others. For example, while automakers may change body styles year to year, they repeat great headlight designs that are cost efficient, aesthetic, and effective. Software programmers do this too. Once a particular code path is created that works well, it is often repeated as part of a subsequent routine. These examples demonstrate good use of prior knowledge with a highly positive effort to value (efficiency) and impact (effectiveness) ratio. Theory of change and program logic models can garner some of these benefits by using archetypes. They offer a substantive contrast to trial and error. Because archetypes are evidence based and tested, they can help jumpstart your own modeling.

As you start to think about how your planned work and intended results might look on paper, a blank page sometimes feels like a steep challenge. In many cases, there is no need to start with a blank page. Archetypes are a great remedy for “model block” or “display paralysis.” In addition to getting some shapes and words on paper, they can also contribute significantly to model quality. We define archetypes as commonly used templates that offer simple evidence-based guides for action. An archetype often looks and feels just like a theory of change or a program logic model. The main difference between an archetype and a logic model is that the elements, relationships, and outcomes specified in the archetype are tested or proven.