What is scenario planning?
Herman Khan (1967) defined scenarios as “hypothetical sequences of events, built with the intention of attracting attention to causal processes and points of decision” The storyline products of a ParEvo fit this description
How scenario planning differs from other related approaches
As explained by Mathew Spaniol here:
The storyline products of a ParEvo fit this description of a scenario
Typologies of approaches
There are many different ways of doing scenario planning, and variants of scenario planning (futures exploration/research, horizon scanning, strategic foresight, etc.).
Here is one recent review: Bengston, D. N. (2019). Futures Research Methods and Applications in Natural Resources. Society & Natural Resources, 32(10), 1099–1113.
In the above paper, Bengston describes three axes that can be used to differentiate different types of scenario planning methods:
In terms of this framework, ParEvo is on the qualitative, imagination and participatory edge of this space. But within any ParEvo exercise, there is room for expert input, quantitative data, and real-world evidence. But as Bengston notes “…there are limits to evidence-based approaches to studying a future that does not exist. Bell (1997, 148) observed that “The future is nonevidential and cannot be observed; therefore, there are no facts about the future.” The creativity- and imagination-based dimension is what sets futures methods apart from most other social science research methods”
Another perspective has been provided by McBride et al (2017). These are more evaluative criteria. For their application to ParEvo, see below.
Seeve (2018) highlights the difference between deductive and inductive scenario planning approaches. I will quote his thesis at length…”
“In the general sense, deductive scenario planning methods can be considered as processes that build the scenarios in a top-down manner. In other words, in deductive methods, the big picture of the developed scenarios is sketched out in an early phase of the process. McBride et al. (2017) specify deductive approaches as general-to-specific techniques to identify key uncertainties. Similarly, Bowman et al. (2013) describe deductive processes beginning with the broad framework of the scenarios, then refining and inserting data in these scenarios.
In practice, deductive scenario planning approaches most often utilize a specific 2×2 scenario matrix technique [Bradfield et al. (2005)], originated by Shell [van der Heijden (1996)]. In this technique, two key uncertainties are represented by so-called scenario axes, the extremes of which correspond to the two possible outcomes of these uncertainties. Then, four skeletal scenarios are developed as combinations of the outcomes of these uncertainties [Schwartz (1991)]. The selection of the two key uncertainties from a large number of factors is based on their uncertainty and relevance. Ogilvy and Schwartz (1998) present the so-called poker chips narrowing exercise as an appropriate method for selecting the two key uncertainties. In this method, the participants of the scenario workshop assign, e.g., 25 poker chips on different factors, based on their uncertainty and relevance. Then, the two factors with the most chips are selected for building the four skeletal scenarios in the 2×2 scenario matrix. These scenarios are further developed by considering what outcomes of each of the remaining uncertainty factors are deemed plausible in the four quadrants of the matrix….
The deductive scenario development approach has several limitations with respect to the aspects in Table 2.1. With respect to the explorativeness aspect, the deductive 2×2 scenario matrix approach is limited in that (a) focusing on the extremes of axes can drive unnecessary polarization in thinking and (b) limiting the number of uncertainty factors at an early stage may pre-emptively restrict the exploration of the future possibility space [Wright et al. (2013), Lord et al. (2016)]. With respect to the trustworthiness aspect,
the deductive approach can be limited as well. More specifically, reaching consensus on the selection of the two scenario axes may be difficult, which may lead to a lack of trust by the process stakeholders in the scenarios resulting from these axes [van der Heijden (1996)]. For example, van’t Klooster and van Asselt (2006) observed controversy in the selection of the scenario axes in the scenario projects they followed. Stakeholders of scenario workshops felt a lack of trust in the selected axes, their critiques including, for
example, that ‘the scenario axes represented a very classical scheme’, and that the selection criteria of the two key uncertainties were not transparent.
In contrast to deductive approaches, inductive scenario planning methods use specific-to-general techniques for building scenarios [McBride et al. (2017)]. In other words, inductive scenario development processes identify scenarios in a bottom-up manner such that the uncertainty factors are not restricted to a small manageable amount. Instead, inductive approaches recognize the scale of the scenario planning problem and seek to include all the relevant aspects of the problem at hand throughout the whole scenario exercise. Hence, the inductive approaches are less prone to restricting explorative thinking, which is evidenced by many successful inductive scenario planning studies [Ritchey (2011), Bowman et al. (2013), Vilkkumaa et al. (2018), Johansen (2018)]. In these studies, entire solution spaces and all possible solutions of scenario problems have been systematically explored [Johansen (2018)]. Consequently, broad thinking and the retaining of an explorative mindset have been supported [Wright et al. (2013)], because the point of view to the future is not narrowed down to just two key uncertainties of change….
Typically, inductive methods for scenario development are relatively unstructured. This is because the strengths of the inductive approach have been seen to arise from the process being open-ended and exploratory, such that scenarios emerge from in-depth discussions about individual events, and the more broad scenario storylines are then developed organically [McBride et al. (2017)]. McBride et al. (2017) note that by having a broad range of plot elements available, the inductive approaches yield compelling plot lines that can focus on the relevant strategic decisions at hand, depending on the particular case study. Moreover, they explain that by having direct connection to plausible events, unstructured inductive methods can effectively link the developed scenarios to relevant strategic decisions in the present [van Vliet et al. (2012)]
McBride et al. (2017) acknowledge that inductive methods, such as the Emblematic events approach, are effective with respect to the explorativeness aspect in the selection of scenario development technique, as presented in Table 2.1. This notwithstanding, they preferred the traditional deductive 2×2 matrix approach, because of the limited timeframe of their scenario exercise and concerns that the inductive process might be unsuccessful due to its unstructured nature. If the inductive scenario development exercises are carried out in an unstructured manner [as in, e.g., Bowman et al. (2013)], the process can become more opaque and dependent on the creativity and imagination of the participants. This unstructured nature of inductive processes can cause greater time and facilitation demands and even risk the success of the scenario exercise [Volkery and Ribeiro (2009)]. Thus, inductive scenario development approaches can be inefficient, having limitations with respect to the third aspect (efficiency) of scenario technique selection in Table 2.1. Moreover, opaqueness in building scenarios can undermine trust in the developed scenarios, and thus an unstructured inductive process can have limitations with respect to the second aspect (trustworthiness) of Table 2.1 as well.
in my opinion, ParEvo is an inductive process but without some of the weaknesses described above: (a) The time frame can be carefully circumscribed by a facilitator, (b) ParEvo uses a very structured process, (c) the process is transparent, (d) can be efficient, (e) and trustworthy. But, it is still true that the process is dependent on the creativity and imagination of the participants.
The Futures Wheel
The most similar approach I have found is called the Futures Wheel or Implications Wheel. Like ParEvo, this starts with a common text, it then involves multiple parallel additions to that text by different participants, through a series of iterations. Here is a description: Bengston, D. N. (2016). The Futures Wheel: A Method for Exploring the Implications of Social–Ecological Change. Society & Natural Resources, 29(3), 374–379.
Here is a simplified example from the above paper:
“A complete wheel typically has about five second-orders for each first-order, and five-third-orders for each second-order. Starting from the center the group is asked, ‘‘If this occurs, then what might happen next?’’ “Both positive and negative first-order consequences should be identified, and the process should be open to even low-probability consequences— the idea is to identify possibilities, however remote”…” Once the group has identified the most significant first-orders, the process is repeated to identify a set of possible second-order consequences. For each first-order, the facilitator again asks, ‘‘If this occurs, then what might happen next?’’ Positive and negative second-order consequences should be identified for each first-order. It is important to complete all first-orders before going on to the second-orders, and all second-orders before adding any third-orders.”
Unlike ParEvo the process tends to end with the identification of third-order implications because “going beyond this level becomes too tenuous.” and probably also because the amount of time required for each level tends to increase geometrically. This does not happen with ParEvo : (a) because the number of contributions per participant per iteration is limited to one and (b) because the selective extension of existing storylines – only those extended in the last iteration can be added to in the next. Thre is no geometric growth in content.
“Some approaches to the Futures Wheel use the groups to subjectively rate each of the consequences in terms of their importance, uncertainty, and other factors” But with ParEvo such ratings are applied to whole surviving storylines, not individual contributions. However, the Comment facility in ParEvo does allow a more free form of evaluation of individual contributions.
There are other common features in how the results of the two types of exercise can be analysed. Each of these possibilities for the Future Wheel can also be pursued within ParEvo:
“Inductive thematic analysis may be performed to identify broad themes (Benckendorff 2008), but more in-depth analysis aims to discover:
- Highly desirable, low-likelihood consequences (and policies or management
actions designed to increase their likelihood).
- Highly undesirable, high-likelihood consequences (and policies or management
actions designed to decrease their likelihood).
- Surprising consequences, including those that could have catastrophic or extraordinarily positive impacts.
- Differences in scoring from alternative points of view.
- Information and monitoring needs for developments that are highly uncertain.
However, because of the Future Wheel’s less structured approach to participation ParEvo enables analysis of the process as well as the content of an exercise. Notably, how the different participants contributed, to each other’s contributions and to different storylines
2020 04 23: Another point of difference: From what I can see, the size of the text contribution made within each node of the Futures Wheel is much smaller than the equivalent seen in a ParEvo tree structure. In the Futures Wheel the contents seems limited to one sentence at the most. In ParEvo it can be one paragraph, or more (depending on the word limit set for the exercise).
See these examples of online Futures Wheel software: