Course – Lesson 3: III. Logic Guides Us
LESSON 3:III – Q+A WITH AUTHORS KATHERINE ORTEGA COURTNEY, PHD AND DOMINIC CAPPELLO
What is a logic model, and how is it used to guide the development and implementation of an experiment or initiative?
Katherine Ortega Courtney, PhD responds:
A logic model is a visual representation of your innovation that combines graphic elements and text to show the logical flow of your experiment, from problem identification and planning through implementation and evaluation. It helps organize the steps and participants in the process so that each component has a purpose, and needs are anticipated for each stage of the process.
What are the first steps in creating a logic model?
Dominic Cappello responds:
The life of a logic model begins with identifying the purpose and goals of the project. These are often statements identifying a need and should relate back to your assessment phase. A purpose is related to the broad problem you wish to address: “Improve access to behavioral health in my county.”
Goals are a little bit different. The goals stated in your logic model should be directly related to your purpose but also more specific. Using the example above, one might identify one goal as “Reduce staff vacancy rate in existing providers to 10%, sustained for a 12-month period.”
Logic model goals do not have to be so specific and measurable, however. They can also be items thought to contribute to the problem you are addressing in your purpose statement, such as “Increase employee morale” or “Create more office accountability and better communication.” It’s okay if you don’t yet know just how you’re going to do all of the above; that’s where the logic model comes in.
What are logic model "inputs" and how are they identified?
Katherine Ortega Courtney, PhD responds:
Put very simply, inputs are all the things (and people) that you currently have that are related to both the problem and the (potential) solution. Even when they are ostensibly linked to the challenge you’re addressing, inputs are still resources. For example, if you wanted to consider the development of a school based behavioral health care center, the following stakeholders would be considered “inputs.” School board, superintendent, school communities, behavioral health care providers and funders who might be city or county government.
What are logic model "activities" and how are they identified?
Dominic Cappello responds:
Activities are the specific steps/actions that move your experiment forward and should be performed by, about, with, or “to” the inputs. Each of the inputs in your logic model should be linked in some way to at least one activity.
An activity for the staff retention challenge discussed above may be an employee survey to discover areas of concern related to work conditions and office climate/culture. This may be created by research and development within the entity (Input A); administered by management within the entity (Input B) to the rest of the entity staff (Input C); and shared with CQI training coaches and colleagues (Input D). The team conducting the experiment may discuss the survey results within the context of an academic research article about similar issues (Input E).
What are logic model "outputs" and how are they identified?
Katherine Ortega Courtney, PhD responds:
Outputs are the specific things that result from Inputs and Activities combining together to create change. In examples shown in this course, the survey results are an output.
For experiments with the goal of implementing a new kind of practice in the office, that new practice will likely be the output of the activities. The outcomes that we hope to get from the new practice are a different component of the logic model.
It is always a good idea to look ahead and anticipate outputs so that you can be sure that the data you need for evaluation is somehow included. For example, if you will need data from a tracking tool to show how your experiment was applied, this data should be included as an output stemming from the activity of developing and using a tracking system (and by which inputs).
What are the different levels of outcomes in a logic model?
Dominic Cappello responds:
In a logic model, outcomes — short-term, intermediate, and long-term — are the sequential changes (and, we hope, improvements) that you expect will happen as a result of your activities and outputs.
Short-term outcomes generally should occur within 2-6 months (or less or more, depending on your project) and may be described as “individual-level” outcomes. These types of changes might involve staff, stakeholders, and/or clients, and probably won’t go beyond changes in knowledge, attitudes, beliefs, behaviors; or small changes in workflow, practice, or protocols. An example of a short-term, individual-level outcome (again using the staff retention example) might be: school board members agreeing to allow meetings discussing the importance of behavioral health care in schools.
Intermediate outcomes will likely take between 6 months and 2 years to appear and are sometimes called “agency-level” outcomes. With ongoing application of your experiment or initiative, you may start to see improvements in the “target numbers” related to your overall challenge at this point. Intermediate outcomes may also include consistent use of a new practice (an all-new “business as usual”); stronger relationships with stakeholders; and secondary gains (or unintended consequences) related to your innovation, such as improved collaboration or need for more resources. Using the example of a school based behavioral health care center being developed, it might be that a school principal has agreed to be a pilot site for such a center, and funding is being identified.
Long-term outcomes should relate most closely with the stated goal(s) of your logic model. These outcomes may take 2 years or more to materialize but should be well-established and consistent. At this point, if your experiment has continued apace, you should hope to see positive changes in your specific challenge area and possibly improvements in related areas. For the development of a school based behavioral health center a long term outcome would be that the doors are open in the center and students and families are receiving behavioral health care.
How are outputs different from outcomes?
Katherine Ortega Courtney, PhD responds:
Making the distinction between outputs and outcomes is easy in theory, but as you start to build your logic model it may feel murky at times. A good way to know for sure whether something is an output vs. an outcome is to ask this question: “Can I prove that this happened due to an activity we did, and not because of something else?” If the answer is yes (such as, you are holding survey results in your hand that came from the survey you administered), it is an output. If the answer is no (such as, your coworkers reported better job morale in a follow-up survey)… it is an outcome.
