Leveraging Complexity Science to understand Human Behaviour

Part 5 | Rethinking Problem Solving using Complex Systems Solutioning Approach

By: Alok Gangaramany and Anushka Ashok

In the Complexity Series, we have articulated a number of issues related to linear problem solving approaches, frameworks and methodologies. In this final post of the series, we will focus on principles and processes that could be used by programs and projects that allow for a more adaptive approach towards addressing complex behavioral problems. In addition to our own experiences and observations, we are inspired by the work of Douglas O. Norman and Yaneer Bar-Yam who have written extensively on Complex System Engineering. We will also leverage ideas from the frameworks and principles from public health focused literature such as the recent framework developed by Medical Research Council

Before we dive into the principles and approach, let’s quickly recap the need for it.

• Development sector programming (particularly in public health) usually begin with comprehensive advance planning which limits the entire program design with the knowledge , theories or hypotheses at the start of the program
• Adapting the program processes (e.g. new experiments) based on learnings and ideas that emerge during the course of projects is difficult if not impossible
• Due to the limitation of our own imaginations and individual cognition and our need for certainty, program designers prefer a linear problem solving with discrete distinctions between problem framing, research, design and implementation phases.
• Even the multiple design approaches (e.g. double diamond 4 D’s) articulated in design literature, tends to suffer from this at least during implementation. While these approaches may have been initially developed on small scale projects with an iterative mindset, they are being applied on significant large scale complex engagements with high level of uncertainty and ambiguity without significant alterations.
• Overall, the current approach tends to promote a deductive and a reductionist approach to problem solving that may be appropriate for Simple or even Complicated problems but not for complex problems. As a result, it limits the potential of development of emergent and non-linear solutions.

Principles

Lets first layout some key problem solving principles that could serve as guideposts for a project:

1. Evolutionary driven Approach: Implement a dynamic and evolving Problem Solving Approach that allows of significant variation in the individual process elements (e.g. removal or addition of primary research efforts) with the assumption that the context is also continuously changing.
2. Team of Teams mindset: Allow for parallel execution by cross-functional teams (e.g. research teams, pilot and implementation teams) that may be part of multiple organizations and communities working in the same project environment.
3. Multiple Scale (or Scope) Operation: In addition to parallel efforts, we may want different teams to operate with different scale of population. Example, population level tools or research approaches (e.g. segmentation) vs community level solutions. This would allow for building on local understanding and implementing a non-linear problem solving approach that imposes variety at multiple scales of the program.
4. Engagement guidelines vs comprehensive requirements: Instead of detailed specifications that limit emergence of solutions, articulate overall project constraints / boundaries, internal and external communication and risk mitigation processes.
5. Research in Development (RinD): Support localized research in support and embedded within-on-going development process
6. Ongoing Learning: Ensure a learning process about effective solutions through direct feedback from the environment is place.
7. Living Theory of Change: Develop and continuously update a living Theory of Change focused on contextual elements that may be driving the behavior of the overall system and its stakeholders
8. Platform for multiple interactions: Plan for an overall Coordination team that focuses on interaction aspects (how are people going to share and coordinate, how will new services be added etc) instead of minute details.
9. Interventions as events: Re-conceptualize the notion of intervention , as an event/s in the history of the system that may allow for the system to self-organize and change the future trajectory of the system’s dynamics.
10. Scalability of function: Instead of the conventional view around scalability of interventions where the form has to be standardized and replicated, focus on the function of the intervention and adapt it for different population context and complexity.

Organizing a new Problem Solving Process

Lets now focus on the structure and process that could support the implementation of the principles mentioned earlier.

The image below provides an overall organization of a complexity driven approach along with the tasks and activities that may be embedded within them.

Unlike usual project engagements where teams and tasks operate sequentially, we propose a parallel team effort. As visualized above, we imagine 3 teams (Theorizing, Learning and Solutioning) that may operate independently with ongoing formal and informal interactions within them and a fourth team responsible for overall coordination of the efforts.

a. Theorizing

As part of theorizing, we are referring to efforts such as problem framing, strategy design, developing a theory of change / logic maps etc. In many ways, the primary purpose of Theorizing team is to hypothesize and synthesize the variety of learnings that would be generated over a period of time.

The complexity driven approach is expected to develop a wide variety of learnings. These could come from different research, solution and evaluation efforts. Using these learnings, the theorizing team may be in a position to describe (and avoid overly prescribing) various solution spaces (e.g. policy, provider, community) that can be leveraged for different solutions. The team could also help build a database of various insights, failures and ideas that the program experienced over a period of time. A key task would be to ensure that the Program Theory of Change stays current and is regularly updated based on field learnings.

Ultimately, the team may help in disseminating models and frameworks that are grounded in real life experiments, capture the contextual constraints and provide guidance on replicating / scaling the principles and ideas in other geographic and problem contexts.

b. Learning

Conventionally learning is associated with research efforts that may be carried out at a population level and sometimes at the supply side (e.g. provider behavior). In addition, to these efforts we believe the learnings that come from Monitoring and Evaluation related tasks can be part of this team as well. One of the key reasons to do that is to break the mental model of an external M&E team responsible for only evaluating outcomes of a program to one that is integral to the program design and contributes towards overall learning and continuous improvement.

As such the overall purpose for this team would be to generate population / aggregate level learnings. This could be executed by analyzing available public health data, running qualitative and quantitative population level research, working with a panel to generate longitudinal learnings and closely working with the Solutioning teams (as part of evaluation efforts) to capture process and outcome level learnings.

c. Solutioning

The Solutioning team is equivalent in some respects to solution design and implementation efforts. The additional layer here is to also allow of localized research that would otherwise be missed when one is conducted at population levels. Their purpose essentially is localized problem solving.

In order to do that, the team may engage in rapid ethnographic observations, learn about culture and community structures from local leaders and influences and use these learnings to develop and test ideas. In the design language, this may encompass prototyping and piloting efforts. In addition, the team would be ultimately responsible for local implementation efforts.

The Solutioning team may be the largest teams and may not be limited to experts. In fact, it may be imperative that this team includes a fair representation and even lead by local communities the program is working with.

The most important role of the team would be to generate and test multiple ideas. Rather than conventional intervention of standardized and replicable intervention across sites, we need to more adaptive interventions, and at an ideal/north star work towards an n=1 intervention. These ideas and feedback on them could then feed into the population efforts of the learning team and influence theory of change and other syntheses of the theorizing teams.

From an evolutionary point of view, with multiple teams learning, pro-typing and iterating in small parts of the system, the “right intervention” in the system could chosen through “(natural) selection” process based on performance in the real context.

d. Coordinating

Its evident that the teams mentioned above would have to frequently feed into each other’s knowledge and experience throughout the course of the program. As an example, the Solutioning team may be able to identify a specific gap (e.g. a need for Discrete Choice Modeling to understand product preferences) that could then influence the efforts of the Learning team. Similarly, Theory of change would need continuous inputs from M&E team as they are evaluating different hypotheses and field implementations.

The role of Coordinating team is help facilitate this dynamic interaction. The team may have to engineer formal interactions between the teams and also help set up forums and other avenues for informal unplanned interactions. Part of coordinating function would also be about identifying ideas and concepts that could be amplified / distributed and those that may need to be dampened / rejected. The Coordinating team could be comprised of the representatives from the other teams who are able to sense the need for interactions.

In addition to coordination, the team may also serve the role of systemic risk managers. Projects in public health tend to carry multiple levels of political, financial , physical and ethical risks that requires strong vigilance and action. Since the coordination team is likely to have a wider view of the program across all teams, they may well position to understand and mitigate risks as they come up.

Lastly, the coordination team could also serve the role of disseminating the program learnings with the external audience. In a way, they could act as an interface between the program team and external stakeholders (e.g. funders) who may be interested in learning and contributing the program over a period of time.

The process outlined above is expected to lead to some level of redundancy / duplicity of efforts between teams. To some extent we think that is needed and even potentially beneficial. An overly optimized approach risks losing important learnings that are gained when ideas are repeated or tested in multiple contexts. We also expect the teams will overlap in skills. For example, someone with a design background might help build a more human-centered theory of change or behavior change strategy as part of the Theorizing team. In general, we see a multidisciplinary group across each team with some level of specialization (e.g. M&E, Implementation) that may be specific to a particular team.

Conclusion

In this 5 part series , we have attempted to highlight issues with conventional problem solving approaches, introduced tools and concepts that we have found useful in our work with complex behavior problems, and shared a high level principles and processes for implementing a complexity-aware problem solving approach. We have relied on the work of many scientists and practitioners to inform these perspectives. At the same time, we have attempted to make new connections and evolve relatively novel ideas. In the spirit of complexity, we expect to continue to test and iterate these approaches and report back our learnings with the wider world.

Final Mile brings unique and proven capabilities in addressing complex behavioral challenges. As one of the first Behavioral Science & Design consultancies, Final Mile has had the opportunity to bring these to practice in a wide variety of sectors and contexts. We have executed highly complex behavior change projects across a wide variety of areas covering Global Health (HIV, TB, Maternal Health, WASH), Financial Inclusion, Safety across Africa, Asia, Europe, and the US.

Final Mile is also building a pandemic playbook that can be used as a potential toolkit by policymakers and implementors in mitigating Covid19 and future such pandemics.

Reach out to us at contact@thefinalmile.com.