How to Manage: Understanding the problem

[ how-to-manage problem product ]

This is the first post about how-to-manage series, a future collection that hopefully will help managers in different areas. This one in particular is the base for decision making to truly understand the problem (and the problem of understanding the problem). Highly based on Microfoundations Of Strategic Problem Formulation by Markus Baer, Kurt T. Dirks, and Jackson A. Nickerson.

Before dealing with the problem

Ok, you have a problem, that can also be an opportunity or a crisis, something that you need to deal with or needs to be solved. The first thing that you need to do is to formulate it (or characterize), put some time and think carefully about it. The important point is: how you approach and solve the problem will be determined by how you formulate it¹.

Putting ideas together and developing all the details of a particular problem can be nontrivial, and this is also a problem. One that we are going to formulate. Think of it as the problem formulation of the problem formulation.

Note: We are not going to talk about solution generation! This may be done in a future article.

The problem of problem formulation

Before we start, there is something really important that I need to tell you, problem formulation is also a data science problem¹ (you collect data, find patterns, and use them to do a good guess, yes that's it). Also, there is another important thing, when you have a problem or a goal, you must simplify reality, and as a consequence, you have a distorted vision of the world, and that may be enough. Without this distortion, the problem may be too complex to do something about it. So imperfect models are still useful, they make decisions possible¹.

Back to the problem formulation. This can be done by an individual or team. Each has variations, advantages and disadvantages, it's a multivariable higher-order problem by itself. Something important to keep in mind is the variables in play and understanding them. Let's go deeper and see some different scenarios where problem formulation can happen:

Individual

Individuals have limited time, information, memory and cognitive resources^{2 3}. Can also have different interpretations of the vision, the mission, or the values⁴, be driven by self-interest and maybe operate with guile⁵.

It's important to know that complex issues reside in many different individuals perspectives⁶ with different experiences. Also, research suggests that individual problem formulation can be described as inappropriate⁵ or as a selection-bias problem⁷. But, the individual is part of the group, their view is important for independent symptom generation, useful for further integration and validation by others.

The problems of the individual are also their strength, no need to translate communication across different mental models, no discussions around terminology, assumptions and causality, quite useful for a first quick generation. Ok, with that in mind, it's important to know that for complex strategic problems, an individual approach can be limited, since it rarely spans the problem space⁵.

Creating a team

When selecting a group of people to do the formulation of the problem, it's necessary to select stakeholders and individuals (with different mental models) that are sensitive or aware of the problem and has information that is able to synthesize⁴. It's important that the actor has practical, intellectual, cognitive and emotional aptitudes⁸. Since we are talking about different mental models, a broader cognitive divergency may result in translation cost (terms, concepts, and assumptions differ), harder perspective structure (merge different perspectives into better formulations), conscious or unconscious tendencies to avoid or generate conflicts, and, lower comprehension if such gaps are not filled or minimized⁵.

One other problem to be aware is groupthink, where groups can be overly optimistic (taking more risk ideas), create mindguards, develop stereotype views, expresses disregard for arguments considered contrary of loyal members, creating self-censorship (accepting status quo) and illusory unanimity⁹. It's also important to note that if the group is over stress and time pressure, groupthink may happen as well¹⁰.

It's also important to not include only people that agree (consensus), if the team has a promotor fidei or create a dialectical inquire, it may improve the formulation of the problem with better generation of recommendations and underlying assumptions^{11 12}.

Note: The perception of a psychological safe environment may result in actors to speak up, express themselves, take inter-personal risks, admit failures and openly discuss errors / weakness¹³.

In the end, the idea is to purposely compose a team that captures the heterogeneity of ideas while simultaneously attenuating impediments⁵.

With that said, let's explore some types of groups.

Homogenous motivation and heterogenous cognition

When being part of a team, the problems of a single actor can be mitigated by different individual motivations and heterogenous cognition, however, it may not solve problems related to homogenous motivations and objectives, resulting in groupthink and its problems. This can happen when a group is cohesive and isolated⁹. The problems around this group formation is a subset of heterogenous motivation and cognition, as pointed in Creating a team, a series of problems may/will occur, but mostly from communication, integration and cognitive (conflict)⁵.

Heterogenous motivation and cognition

This group formation helps to mitigate a couple of problems from Homogenous motivation and heterogenous cognition and also pointed in Creating a team. With the heterogenous motivation, groupthink (single-interest framing) is unlikely to happen, conflict interest will probably be exposed, assumptions justification may increase and, as consequence, representational gap will be reduced. But, the same problems may happen by different reasons. It's possible to have a dominant frame by high stakeholders^{5 14}. Information may not be shared, distorted or emphasized, in order to manipulate the problem formulation for personal benefit⁵. The team may jump to a solution, to reduce time in the problem formulation or to engage in dominant activity, if an actor shares a particular solution prematurely, it may affect the search and formulations of alternatives⁵. There is also the possibility of other actors to engage more in dominance and alternative proposals by noticing that someone else is behaving in political interest⁵.

Homogenous cognitive and homogenous information

Teams created by low diversity in cognitive (expertise, assumptions, problem-solving methods) and information, may add little beyond the individual problem formulation and less compared with the heterogenous cognitive and information. The main difference compared with other groups are time spent in translation and integration, without such differences, the group may operate as in an echo chamber, where participants can amplify and reinforce beliefs with the possibility of increase social behaviour.

How to formulate the problem

The following is proposed by Microfoundations Of Strategic Problem Formulation by Markus Baer, Kurt T. Dirks, and Jackson A. Nickerson (in my vision at least), the problem formulation should be done in two phases, framing and formulation:

Framing:

1. Define an initial symptom to trigger the actors.
   • E.g: Users replace Raspberry Pi with other hardware for autonomy, image processing, AI, etc.
2. Set the rules: Only symptoms and evidences, no causes or solutions.
   • Avoid directing points of view or interest by discussing solutions and causes.
3. Actors write down all symptoms with supported evidence individually.
   • Mitigate influence, dominance and early consensus while avoiding opinion or personal bias.
   • E.g: Support tickets, forum posts, custom integrators, competitors mentions, etc.
4. Each actor takes turn to reveal one symptom until all symptoms are revealed.
   • Again, mitigate influence, dominance and early consensus.
5. Each symptom and evidence is then discussed to be excluded or included.
   • Improve comprehension and information sharing.
6. The team compiles a document with all symptoms and evidences (web of symptoms).

Formulation:

1. For each symptom, the members write down the causes individually (no solutions).
   • Create a web of causes avoid prematurely elimination or consensus.
   • E.g: The Raspberry Pi does not have enough compute power for intensive tasks.
2. Each person reveals one cause at a time until all causes are considered.
3. Discuss whether each cause plausibly explains the symptom.
   • Improve comprehension and information sharing.
4. Compile the accepted causes and supporting reasoning in a document.
5. Share the compiled document with the team and external stakeholders to detect missing points and share comprehensiveness.
   • Avoid blind spots and improve comprehension and information sharing.

The important point of this approach is how it mitigates the possible problems described previously related to individuals and groups organization. An important point to take is that is not necessary to use the proposed process, but to understand and be aware of the problems and the idea behind the steps that could help to mitigate and improve problem formulation tasks.

Personal final notes

First, this article was mostly based in my personal search around the matter, many days in a café and a lot of chai latte. As consequence, think of it as the work of an individual, with all the consequences that was previously mentioned here.

The following notes are just things that will not be further explored here but are important to have in mind:

• Conflict is something that I personally believe that's necessary, don't think of it as people fighting, but as people discussing the problems of the proposal and exploring the span of consequences to finally calculate trade-offs (empirically)¹¹.
• Neurodiversity may play an important role in problem formulation tasks, where such cognitive outliers may provide a broader problem formulation perspective with non-linear thinking and pattern-resistant minds, things that a conventional AI cannot replicate. ^{15 16}.

References