‘ogram theory, the term program refers not only to something formally labeled as a ram (for example, in a corporate management hierarchy of programs, subprograms, components). It can refer to any intervention; a project, a strategy, a policy, a ing initiative, or an event. It includes interventions that are undertaken by a single nization, such as a direct service delivery project, and those that are undertaken by ple organizations, such as a whole-of-government policy. It refers to both lanned and tightly specified interventions, and broadly defined and emergent ventions. Program Theory agram theory is an explicit theory or model of how an intervention contributes to a of specific outcomes through a series of intermediate results. The theory needs to de an explanation of how the program's activities contribute to the results, not ly a list of activities followed by the results, with no explanation of how these are 4, apart from a mysterious arrow. We find it helpful to think of a program theory as 4g two components: a theory of change theory and a theory of action, One of the {fits of articulating program theory is being able to systematically review the quality a theory in terms of OlaUsT ERE BR CUA yM en. ane uly refers to the central mechanism by which change comes about for individuals, 28, and communities. For example, many health promotion programs are based on ory of change that behavior changes in response {o perceived social norms. It is ble to have more than one theory of change. There might be different theories of je at different stages of the program (for example, one about participants becoming ged with the program and one about their changing their behavior) or for different Xs of people (for example, some people might change their behavior in response to nformation about risks and benefits, while others might change only in response to ale incentives), agrams are usually, but not always, about change. Sometimes a program aims to ar reduce change or prevent something from happening—for example, maintaining nobility of a person with a disability who might otherwise develap restricted ment, or maintaining biodiversity despite pressures from agriculture and industry. In cases, the theory of change explains how pressure to change will be resisted or sted. Another way to think of this is that the program is changing a situation from it otherwise would have been. Theory of Action explains how programs or other interventions are constructed to activate their y of change—for example, what the program does te change social norms or make a Peat aos : Le oa sis WISI [6A6| 0| entcceze Mll| pe UeEKEg Jol eSCU Onycows (0 blognice [Ye [Us| IufeUgEwey meu Togram theory is usually displayed in a diagram called a logic model. The form that logic model takes can also affect the way we think about a program theory and can pe it. It is therefore important to understand the strengths and weaknesses of xrent types of logic models for different purposes and to achieve an appropriate ch between the type of logic model and the type of program and purpose. \ithough logic models can be drawn in numerous way, they can be grouped into four ad types: pipelines, outcome chains, realist matrices, and narratives. Pipeline logic Jels look like production line flowcharts or a series of building blocks, and they are n light on theory. They show the inputs used, the activities undertaken, and short- 1 and longer-term results, all without necessarily explaining how the activities bring ut the results. We refer to these as pipeline models because they represent the jram as a linear process where inputs go in one end and outputs and outcomes ye out the other end. Pipeline logic models are most useful when the activities are all ront, and then the rest of the results simply happen like a row of dominos (or at least \ the help of other processes which are not the focus of the program). They can 1etimes be a good starting point for organizations not accustomed to thinking about reporting in terms of outcomes and impacts, but the risk is that they will entrench an simplified and unhelpful view of the program, We have found that logic models in the 1 of outcome chains are usually more effective: these represent the intervention and sonsequences as a series of results. Logic models in the form of realist matrices 1s on identifying the conditions under which the theory will apply. Both outcome ins and realist matrices emphasize the importance of understanding and focusing on tionships and linkages as much as on the building blocks. Narrative representations rogram theory tell a story about how the intervention works, or the experience of icipants as they pass through. They can be simple narratives that follow a pipeline sal process or more detailed narratives that explain how change occurs. y forms of evaluative activity explicitly combine facts and values. Evaluation include ‘ont needs analysis that explores the nature and extent of a problem, existin rgths and capacities, and options for meeting needs, It includes process evaluatio ifferent types: documenting an innovation, checking compliance with specifications identifying issues and problems that need to be resolved in ongoing, formativ uation. It includes impact evaluation, analyzing the contribution of the intervention t cular results, and cost-benefit evaluation, where the costs of implementing a vention (including resources consumed and any negative effects) are weighe: nst the benefits nrndured