Updated: Oct 23, 2023
In this article I will share 3 tips on how to design better Theories of Change (ToC). The ToC is the organisation’s critical assumptions on how their interventions bring about change (Valters, 2015). You can read more on the definition of the ToC and the background in my previous blog post.
A ToC usually consists of a diagrammatic representation and a narrative. Most persons are familiar with the former. It is the visual representation with the boxes, circles, clouds, arrows, trees, woodland creatures, rainbows, pixie dust…oh, I got carried away (similar to the way some organisations get preoccupied with ‘artwork’ of the ToCs).
The visual representation is usually accompanied with text (the narrative) that gives more details on the wider context, assumptions, indicators, rationale and the elements shown on the diagram.
This blog post focuses on the diagrammatic representation.
That is, suggesting tips on how to make a ‘good’ visualisation of a ToC. My subjective definition of ‘good’ refers to a visualisation that is communicable to a wide audience, while being evaluable at the same time (learn more about evaluable ToCs from Davies (2018)).
Pondering on what makes a ToC ‘good’
Now, while in the literature there is no consensus on whether a logical framework or logic model qualifies as a ToC or not, what is apparent is that at the very minimum, a ‘good’ ToC should have an outcomes pathway WITH causal links.
The outcomes pathway depict how one action/intervention (outcomes) leads to another until the ultimate change (impact) is achieved. Think of it as a ‘road map’ if you will (or as a ‘compass’ (Valters, 2015)) that navigates the way to arriving at the destination of the achieved goal.
The visualisation may be enhanced in the following ways.
Tip 1: Describing and labeling the connections
The outcome pathway in most ToCs are usually presented as the one in Figure 1. Text boxes are connected by arrows that flow in a linear way towards the end result(s).
Figure 1 The Theory of change of vegIMPACT
However while it is useful to see where the interventions are connected, the person looking at the ToC would not gain insight on the nature of the connections nor on the evidence for the particular causal link.
For example, what is the scale, timing or duration of the connection? Or why does this action trigger or cause the next action to occur. Just drawing an arrow from one outcome to the next does not mean that the next outcome will automatically happen.
As such, a good practice is to label the connections using colour codes and different text to describe the relationship between each outcome (more on that later).
Figure 2
The diagram shown in Figure 2 has numbers on the casual links (arrows) which directs the reader towards additional information on the nature of the connection. It could be that connection labelled number 9 (women attending classes in non-traditional skills) has more significance or weight to the achievement of the long term employment of women (ultimate result) versus the connection labelled 6 (counselling and practical support). The ToC in Figure 1 did not have any label or code to give further information on particular causal links/connections.
Tip 2: Inclusion of feedback loops
The inclusion of a feedback loop on the visualisation of the ToC suggests that the developers are cognisant that the process of change and real life is a complex and non-linear. Shifting contexts, priorities and actors all affect and act upon each other during the course of an intervention.
For example, an output of ‘training women in non-traditional skills’ leads to an outcome of ‘more educated women’. After these women graduate from the programme, they may become trainers themselves and ‘feedback’ into the output of ‘training [other] women in business skills’.
Feedback loops are one way to capture this non-linear change process (read this interesting article for more information on feedback loops).
However, from the research it is seen that of the limited number of ToCs that do have feedback loops, these loops are usually positive and not negative (Davies, 2018). So keep in mind that feedback loops can be also be negative. Remember also to have a ‘reasonable amount’ of feedback loops so that the ToC does not become a tangled incoherent web that is hard for the reader to decipher.
Figure 3: The Theory of Change of Rehabilitation Intervention for people with Schizophrenia in Ethiopia.
Figure 3 shows a ToC with codes and labels for the connections as well as feedback loops.
Tip 3: Highlight the relative significance of each pathway
Another tip to enhance the visualization of the ToC is to differentiate between the relative significance of the elements and the connections. As mentioned earlier, not every connection has the same weight within the outcomes pathway. Yet some ToCs present all the elements equally with the same size boxes and same thickness of the arrows. This implies that no outcome is more relevant than the other.
In reality, some outcomes are more critical than others. For example, a change in the attitude of employers towards hiring women is more critical to achievement of long term employment for women than an outcome of women attending peer-to-peer counselling.
Other outcomes may be easier to achieve in the shorter term (for example, women starting their own business may be easier to achieve than changing the cultural mindset towards economically emancipated women in a patriarchal society).
This differentiation on the significance of the outcomes may be done by varying the thickness of the arrows/ lines connecting the outcomes and/or varying the colour, shape and size of the boxes. See Figure 4 for an example.
Figure 4 reproduced from Dhillon and Vaca (2018)
Dhillon and Vaca (2018) provides a nice summary of the features of a ‘good’ strong ToC versus weaker ones (see Figure 5). You could consider these points as you design the visualisation for your own ToC.
Figure 5. Reproduced from Dhillon and Vaca (2018)
Lastly, figure 6 shows a ToC that has all the elements discussed in this blog post; the feedback loops, varying the colour of the lines, labelling some of the connections etc. On the surface this visualisation is a ‘good’ one (we don’t know about the plausibility and evidence that the outcomes and linkages identified will lead to the change (s)).
For the purposes of this blog post this is not important to scrutinise. Hence, Figure 6 does not need to be legible. What is more important is to see how the elements are presented with the feedback loops, colours etc.