Framing of the problem defines the decision context and the dimensions of the analysis. The aim is that after this phase the problem is defined clearly and meaningfully and that all stakeholders interpret the problem similarly.
Image:
Issues to consider framing the problem
- What situation triggered the need for a decision?
- Who is the decision maker and stakeholders, and in what issues can the decision maker have an impact?
- What is the fundamental goal of the problem that you are trying to achieve?
- What are the scope and boundaries of the decision (Too narrow scope may miss important factors, but too broad can become unmanageable)?
- What is included (and what is not)?
- What are the different dimensions from which we should analyse the problem (time frame, spatial extent, included viewpoints of the stakeholders, level of governance, etc.)?
- Can relaxing the system boundaries give some new innovative viewpoints to the framing
The framing of the problem is often carried out iteratively with the phases of identifying objectives/criteria, alternatives and uncertainties. For example, the framing typically has an impact on what kinds of alternatives it is reasonable to consider, whereas the availability of alternatives has an impact on how the problem should be framed.
Innovative framing of the problem often requires out-of-the-box thinking and relaxing the system boundaries. For example, a problem of a forest owner of finding an optimal strategy for a certain forest stand may in a broader picture appear to be part of a larger system, when more efficient solutions could be found through designing an optimal regional forest management strategy for the whole area.
Outcome of the phase: Description of the problem including the boundaries of the decision making situation and the dimensions of the problem to be solved
Follow up Questions for the AI
Ask the following questions to get more information from the AI:
- Could you please describe the fundamentals of framing the problem in decision analysis?
- What issues should be considered when framing the problem?
- What are the typical dimensions to be considered when framing the problem in environmental decision making and can you give examples of these?
Image:
Image:
Benefits and Risks of Using AI
Image:
Benefits (++)
- Support for understanding what should be considered in the framing of the problem
- Summarising the consequences of different problem framing options
- Providing support for how to bring in transformative perspective into the framing
Benefit Scoring: ++ Medium
Image:
Risks (!!)
Hidden and framing bias
- AI systems are trained on existing data and patterns, meaning that they may favour conventional or dominant framings and underrepresent alternative perspectives (e.g., local knowledge, ecological values)
- AI is typically not very good in out-of-the-box thinking, but asking questions in a certain way (e.g., by asking AI to relax system boundaries) may enhance AI to use a broader context
Lack of context-specific understanding
- AI does not truly “know” your situation, but lacks detailed local knowledge and may not fully understand your personal values or goals
Oversimplification of complex problems
- AI is not being able to understand the framing as a whole but only focuses on certain dimension
Risk Scoring: !! Medium
Questions to Better Understand the Benefits and Risks of AI
Questions to make use of the benefits
- I have this problem: [description of the problem]. I want to find out transformative solutions for the problem. What kinds of options there are for framing the problem and what issues should I consider in the framing?
- What are the key system boundaries of my problem [description of the problem] and what kinds of framings can the relaxing [or tightening] of the system boundaries give?
- I have this problem: [description of the problem]. Please give me provide various ways to frame this problem and for each framing, list key objective, main
Questions to avoid the risks
- What are the risks of using AI to support problem framing in decision analysis?
- [After getting proposed framings] Critically evaluate the framing you proposed: what biases or limitations might it still contain?
- [After getting proposed framings] Explain why the proposed framing suits for this problem and why it does not suit for this problem?
Questions NOT to ask
- I have this problem: [description of the problem]. What is the best way to frame my problem?
These kinds of questions encourage premature closure and reinforce false objectivity.