A post by Dr Cristina Vasilica @cristinavas
I become acquainted with the logic model back in 2009 during a project which involved evaluating the impact of digital media on engaging young people in looking after their sexual health. Evaluation was something I never experienced before, hence it was certainly clear that I had to pick a model that resonated with my interests and skills. The logic model seemed ‘logic’ to me as it explained the relationships between what ‘we do’ and what ‘we get’ as a result of our actions, likewise in system planning and development (my field of work).
The logic model is as much an evaluation method of efficiency and effectiveness of a project as a way to provide a common roadmap or purpose to all the members involved. The approach is situated within the theory driven evaluation. The theory-based approach is an umbrella term for two types of evaluation: realist evaluation and the theory of change. The former is concerned with identifying mechanisms (how change occurs) that work within a specific context (external constraints or setting) to achieve an outcome (results wanted) (Figure 1).
In this blog, I am going to focus on the ‘beauty’ of the theory of change, a concept that in research is often used synonymously with the logic model. There are various logic models developed including the Weaver’s Triangle (Figure 2) and the Wisconsin Model (Figure 3). The Weaver Triangle is visually limited in showing the relationships among aims – outcomes – outputs, therefore used in smaller projects with a limited number of activities.
The Wisconsin model is the most common logic model used.
According to the model, programmes have Resources, Activities, Outputs and Outcomes, explained below.
Personally, I used the Wisconsin model because it enabled me to see the wider picture but also draw relationships. For example, I used it to compile a toolkit for a charity to evaluate the impact of their digital services. It provided a clear guide on how to plan and generate evidence-based data. Furthermore, I used the model as a blueprint for a realist evaluation on the impact of social media on patients’ information provision, networking and communication. It enabled me to find some logic and relationships amongst disconnected Context-Mechanism-Outcomes.
Creating the logic model can be difficult and time consuming. There is a fine line between oversimplifying relationships and not adding enough detail. Furthermore, researchers argue that the logic model – unlike realist evaluation – focuses mainly on ‘what works’ not capturing the pitfalls of projects.