Thursday, November 25, 2021

Choosing between simpler and more complex versions of a Theory of Change

Background: Over the last few months I have been involved as a member of the Evaluation Task Force, convened by the Association of  Professional Futurists. Futurists being people who explore alternative futures using various foresight and scenario planning methods. The intention is to help strengthen the evaluation capacity of those doing this kind of work

One part of this work will involve the development of various forms of introductory materials and guidelines documents. These will inevitably include discussion of the use of Theories of Change, and questions about appropriate levels of detail and complexity that they should involve.

In my dialogues with other Task Force members I have recently  made the following comments, which may be of wider interest:

As already noted, a ToC can take various forms, from very simple linear versions to very complex network versions. 

I have a hypothesis that may be useful when we are developing guidance on use of ToC by futurists. In fact I have two hypotheses:

H1: A simple linear ToC is more likely to be appropriate when dealing with outcomes that are closest in time to a given foresight activity of interest. Outcomes that are more distant in time, happening long after the foresight activity has finished, would be better represented in a ToC that took a more complex network (i.e. systems map type) form

Why so?: As time passes after a foresight activity, more and more other forces, or various kinds, are likely to come into play and influence the longer term outcome of interest. As a proportion of all influences, the foresight activity will grow progressively smaller and smaller. A type of ToC that takes into account this widening set of influences would seem essential 

H2: This need for progressively more complex ToC, as the outcome of interest is located further away in time, can be moderated by a second variable, which is the social distance between those involved in the foresight activity and those involved in the outcome of interest . [Social distance is measured in social network analysis (SNA) terms by units known as  "degree", i.e, the number of person-to-person linkages needed for information to flow between one person and another]. So, if the outcome is a change in the functioning of the same organization that the foresight exercise participants they themselves belong to, this distance will be short, relative to an outcome relating to another organisation altogether - where there may be few if any direct links between the exercise participants and staff of that organisation

The implications of these two perspectives could be graphically represented in a scatter plot or two-by-two matrix e.g.

On reflection, this view probably needs some more articulation. Social distance will probably not be present in the form of a single pathway through a network of actors. Especially given that any foresight activity will typically involve multiple participants, each with their own access to relevant networks. So there may be a third relevant dimension here to think about, which is the diversity of the participants. Greater diversity being plausibly associated with a greater range of social (and causal) pathways to the outcome of interest. And thus the need for more complex representations of the Theory of Change.