Type to search

Module 4: System Mapping

Tip Sheet: How to Select Elements for a Final Domain Map/List

After a group has developed a domain map, or several versions of a domain map, it can be helpful to reduce it to the most important elements (e.g. a manageable set would be 5–10 elements max, not 20). This can make it easier to work with and ensure that participants are clear about the scope of the study. At Horizons, where the Metascan projects often have a very large scope, this step is often necessary. For projects with a smaller scope, it may not be needed.

In the scanning phase, a reduced list of elements provides focus for participants as they scan for signs of disruptive change affecting the domain. This paring down of elements is also necessary for later stages of Horizons’ foresight method, such as selecting change drivers (module 5) and developing scenarios (module 6).

To reduce and finalize elements, it may be helpful to:

1. Ask participants to indicate which elements of the domain they consider to be the site of most significant disruption or surprise. At the end of the domain map exercise (guide #1 or #2), there is an option for participants to weigh in on this question. Specifically, participants were asked:

  • Which elements are expected to change most over the period of the study (typically 15 years)?
  • Which elements would be most disruptive for policy and policymakers, should they change significantly?

2. Collapse several elements into one, where appropriate. It may be necessary to identify several possible broad categories of elements and compare them for their relative merits. (Do this in a discussion with others in the core facilitation team or with participants). After broader elements have been selected, more specific elements can be incorporated into them.

3. For each element, write a statement to define what it includes. For example, if “equality” is selected as an element, would this include the social, political or economic dimensions of equality, all of the above, or something else entirely for this study?