Environmental Problem Solving

Decriptive decision making

characterising and explaining regulaties in choices people are disposed to

Prescriptive decision making

how people should make decisions

uses PrOACT cycle

PrOACT Cycle

define Problem

ollicit Objective

develop Alternatives

estimate Consequences

evaluate Tradeoffs

Scope

where and what time are we making decisionS

Scale

where are we making the decision

Frequency and timing

When are we making the decision

How long will the decision be in play for

Decision maker

usually one person

Trigger

cause of problem

Stakeholders

people or groups with interest in problem

they are affected or have power over decision

Real constraints

laws

env policy

Percieved constraints

Finances

budget

decision accpetability

Problem Statement

Decision

Trigger

Scale and scope

Frequency and timing

Stakeholders

Key decision maker

Potential Constraints

Objectives

Things stakeholders care about

Maximise/minimise

Thing that matters + direction we want it to go

Maximise employment

Maintain economic growth

Increase economical opportunities

Means objectives

Means to achieve a fundamental goal

Fundamental objective

Absolute goals or values

Process objectives

aspects that must be followed

tick boxes that one must jump through

Strategic objectives

Goal of an organisation as a whole

broader missions beyond any one decision

Targets

Desired level of objectives → conserve 100 parrots

lead to poor decisions with multiple objectives

reduce ability to make trade-offs

Make decisions less clear

Trade-off

how one objective may impact another

need to understand magnitude and direction first

Wishlisting

how to get stakeholder to reveal objectives

“what is wrong with the current situation?”

“what is the worst possible outcome? Why?”

“what is the best outcome? Why?”

“what if we did nothing?”

Means-end diagram

Linear, single fundamental objective

To get from means to ends, ask “why?”

To get from ends to means, ask “how?”

Keep asking until they say “because it just is important”

Objectives Hierarchy

Can have mutliple fundamental objectives

Branching

Fundamental on top, means below

Ask “what do you mean by this?”

How to identify hidden objectives

Ask “If we acheive all objectives here, would be satisfied?”

Ask “If our solution, action X, was implemented, would there be any concerns?”

If we only have means objectives and no fundamental objectives, ask “why?”

Performance Measure

Metrics to measure alternatives in respect to our objectives

Sick → P.M. could be your temperature

complete and concise

Natural, proxy, constructed

Natural Performance Measure

Directly reports acheivement

Best

maximise # of sick days → # of sick days

maximise revenue → $ of _ sold

Proxy performance measures

Correlate well with but do not directly measure objective

Use when performance not available

Maximise sustainability → # of EV’s sold

Minimise student boredom → # of yawns

Constructed Performance Measures

Constructed, relative scores

Rate an expirience

IUCN list catefories

Highway fire levels

Influence Diagram

Help understand potential actions

Graphically represents causal relationships

Unidirectional

Chance node = driver

Action node = action

Utility node = objective

Influence connector lines

Alternatives

Potential solutions to be compared by decision makers

Always include alternative of “Do Nothing”

Must address same aspects and have same timeframe

Must answer fundamental objective

Book End Alternatives

One book end: simple and feasible but not ideal

Other book end: extreme and challenging but perfect solution

Menu Board

Table with actions organised by degree of intensity

Leaverage points

Points on system we can influence

Influence diagram

Consequence Table

Alternatives and objectives

Help simplify problem

Conceptual Model

Help map out connections in the system

Influence Diagram

Predictive Model

Predict future

Systems, decisions trees

Good for estimating consequences

Expert elicitation

Dominated alternative

Contains consequences that are estimated to be the same or worse than a single other alternative across all objectives

Irrelevant objective

Contains consequences equivilent across all alternatives

Even Swap Method

Make consequence eqal by adjustive the values for a single alternative between 2 objectives

± for one consequence while ± to another at equal amount

Qualitative Data

For data hard to quanitfy into numbers

Inference from case studies

Quantitative Data

Uses numerical relationships to make inferences about a system

High theory

Widely accepted principle across the scientific field

Don’t need to test every timme

Gravity

Expert knowledge

Broad range

experience based

More holistic

Focuses on particular concerns rather than general categories

Traditional ecosystem knowledge

responds to nature variation

Flexibility and resilience

Role of experts

Predict consequences

Make consequence table

parameterizing models

Supplment when limited time/money/data

Steps for expert elicitation

1. Delphi process and get diverse group of experts

2. Investigate and ask questions in specific ways

3. Aggegrate answers

4. Discuss

5. Revise estimates, get closer to the truths

IDEA Protocol

Ivestigate → all experts answers questions and say why

Discuss → Experts shown anonymous answers and visuals of data

Estimate → experts make second and final estimate

Aggregate → mean of second round answers and discussion

Delphi Process

Step 1: Prepare

Step 2a: Intro meeting

Step 2b: Investigate (Round 1) → get first round estimates

Step 2c: Analysis and feedback → aggregate and display

Step 2d: Discuss

Step 2e: Estimate (Round 2)

Step 2f: Aggregate and visualise

Classes of decision analysis

Single objective, no uncertainty

Single objective, uncertainty

Multi-objective, no uncertainty

Multi-objective, uncertainty

Mutli-Criteria Decision Analysis (MCDA)

multi-objective

Outputs is Multi-Utility Attribute (MAU)

Needs weighted importance

Steps for MCDA

1. Make consequence table (simplify)

2. Add weights of importance to each objective

3. Combine Across Objective

4. Find largest MAU

How to get weights for MCDA

Estimate through discussion → quick and easy but may have bias

Analytical Hierarchy Process → compares pairs, good but complex

Swing Weighting → Ranks swings of outcomes, worst to best

Swing Weighting

Swing from worse to best

First identify best and worst outcomes

Make hypothetical scenarios for worse objective

Swing one value to best for each one

Rank which ones from worst to best

Then score based out of 100