MRI Lecture 6

Practice oriented research, Interventions, Systems Dynamics

dr. Merel van der Wal

 PRACTICE ORIENTED RESEARCH, INTERVENTIONS, AND SYSTEM DYNAMICS

“Our plan is great for

everyone! We don’t know “We have made many
why they don’t say yes. changes to our business.
What can we do to convince We hired a Marketing
them?” manager. Things are going
great!
But now, he’s close to
having a burn-out. What
can we do to prevent
that?”
Example A.

This Photo by Unknown Author is licensed under CC BY-SA

LEARNING OBJECTIVES

Practice oriented research/ interventions, system dynamics

O1: The student is able to explain the value and the basic ideas of practice oriented
research and also is able to explain the principal aspects of system dynamics

O2: The student can read a simple system dynamics model and can formulate basic
answers on the questions following the model

O3: The student can reflect on the situations that are important for practice oriented
research and/or system dynamics. The student also can provide arguments for or
against the choice for SD in a given situation.

3
TOPICS

1. Theory oriented vs practice oriented research and interventions

2. System Dynamics (SD) to capture complexity and dynamics of problems

3. Group Model Building (GMB)

4. Types of SD models
1. THEORY VS PRACTICE ORIENTED
RESEARCH AND INTERVENTIONS
1. THEORY-ORIENTED VS PRACTICE-ORIENTED RESEARCH

• A research objective contains a knowledge (internal) and an action (external) goal.

• In theory oriented research, the knowledge problem is solved (answered), and it is

implictly assumed this will help solving the external action problem (f.e. theory gap).

• In practice oriented research, solving the action problem is a major part of the
research project (e.g. implementation of results in organisations).

6
?

How can you determine whether research

is theory or practice oriented?

a. You have to read the full paper to see what the authors say
b. You look at the objective of the research, and see what the external goal describes
c. You check for involvement of companies or other practice-partners
d. You check if the researchers use academic literature
TYPES OF PRACTICE-ORIENTED RESEARCH

Non-participatory research Participatory research

(action research)
Knowledge domain Knowledge domain

Action domain

Action domain

• Researcher dominated • Analyst / advisor: process facilitator

• Traditional methods and techniques • Own range of methods

 RESEARCH PERSPECTIVE OF PRACTICE ORIENTED ACTION RESEARCH

• Change through research

• Co-occurrence of (empirical) research, participation, and action (change)

• ‘Action’ research: knowledge is always gained through action and for action (Torbert,
2001)
• ‘Solution’ of the problem occurs in parallel with the research process

• Change in the role of the researcher: expert  facilitator

• Types of problems: messy, or complex, problems

9
INTERVENTIONS

• Argyris (1970) defined interventions as coming ‘between or among

persons, groups or objects for the purpose of helping them’.

• In messy problems, interventions are not about generating and choosing

solution alternatives: messy problems cannot be solved but are resolved.

• An intervention can generate consensus about what the problem is and,

sometimes, how to tackle it
(but we have to be aware of premature consensus).

10
 Research is always based on Interventions take place within the

Empirical cycle Intervention cycle

Theories/ Diagnosis
Theoretical
knowledge

Empirical analysis / Hypotheses

/expectations Evaluate Design the change
generalizations

Observations/ Implement
Collect the change
empirical data
THE CONNECTION

• Research produces knowledge; not solutions!

• Knowledge is helpful to find solutions for problems in reality

FOUR STAGES IN THE INTERVENTION CYCLE:

1. Diagnosis
a) Problem analysis: for whom is a certain situation problematic? Why?
b) Diagnosis: what exactly is the matter?

2. Design: what to do, in order to improve the problem situation?

3. Intervention/implement: Prepare and start the intervention

4. Evaluation: (how) did it work?

13
WHY ARE INTERVENTIONS AND ACTION RESEARCH MENTIONED SEPARATELY HERE?

In university teaching, we often focus on relationships between two variables in a ‘clean’

and non-political environment; for clarity, learning, understanding.

Research quality is determined through reliability, validity, consistency, etc.

But, as a consultant or within an organisation, we are continuously confronted with

multiple relationships in a complex and very political environment

‘Quality’ can include many other criteria in a real organisation

14
EXAMPLE A “We have made many
changes to our business.
We hired a Marketing
manager. Things are going
Initially, this organization wanted to know how they
could proceed into a more professional organization great!
with growth and development in business and But now, he’s close to a
product. burn-out. What can we do
to prevent that?”
• What would this organization want?

• When is the research any good to them?

• And to the marketing manager?

 PRACTICAL PROBLEMS = MESSY PROBLEMS

• Complex (so: not simple and not routine)

• Analytical complexity: too many variables and relationships
• Dynamic complexity: existence of feedback
• Social complexity: multiple actors, multiple perspectives, conflicting
interests, interdependencies

• Important intangibles
• Key uncertainties
• Is there ‘one’ problem, and what is it?

• We use interventions to help organizations with their problems

16
2. SYSTEM DYNAMICS
18
19
3. SYSTEM DYNAMICS (SD) TO CAPTURE COMPLEXITY AND DYNAMICS OF PROBLEMS

www.emsd.eu

https://www.ru.nl/en/education/masters/business-analysis-and-modelling

20
WHAT IS A SYSTEM?

• A system is a set of interacting or interdependent component parts

forming a complex and intricate whole;

• So it concerns
• Variables (Concepts);
• Relationships;
• The system as a whole

21
Objectives & Uses of System Dynamics

• Understand complexity and dynamics of systems

• Complexity : Understanding interconnections, feedbacks, &

delays
• Conflicting actors : Modeling what is on people’s minds
(mental models)

• Models are tools for problem-structuring & decision support

• Causes & Consequences :

• Visualization & Analysis of problem causes
• Exploration of decision consequences

• Dividing and structuring complexity into manageable pieces

Building Blocks of System Dynamics

CAUSALITIES & POLARITIES

• How X and Y are related? We are interested in the direction of change

• Causality: X affects / influences / causes change in Y

X Y

• Polarity: Causality can be Positive (same direction) or Negative (opposite direction)

+ -
product quality sales product price sales
Reinforcing feedback loops: Why are systems so complex?

System elements are interconnected

• Feedback: First system element affects second system element (X -> Y),
then second system element affects first system element (Y -> X)

• Time is crucial in understanding and seeing feedback

• Feedbacks loops leads to non-linear & counter-intuitive behaviors

Reinforcing feedback : Why are systems so complex?
Reinforcing feedback loops
Balancing feedback loops
 Feedback loops

• A loop is a closed path of cause and effect.

• We tend to think linearly and overlook these indirect effects on systems. That is why
feedback loops can lead to counter-intuitive behavior.

• Complex systems: There can be different feedback loops affecting a system element.
Then, which of the effects will be dominant?
Exercise: Is it reinforcing or balancing?
 Is this a Reinforcing or a Balancing loop?

• Walk through the loop : Always check your narrative!

• Start with an increase of a system element:

• If it increases in the end again after walking the loop: Reinforcing
• If it decreases in the end after walking the loop: Balancing

• Odd or Even number of minuses

• Odd number of minus(es) in the loop: Balancing
• Even number of minus(es) in the loop: Reinforcing
 Underlying idea of System Dynamics

• Endogenous view: Behaviour is a consequence of system structure; not external factors

• System elements are interconnected

• Feedbacks & non-linear behaviours

• Structure drives Behaviour; (= the outcome of all these accumulating forces)

• That is why, we use System Dynamics to study complex systems’ behaviour.

3. GROUP MODEL BUILDING
 Group Model Building
• Participative workshops to build System Dynamics models.

• We invite relevant stakeholders to model the system

• No single person can be an expert in all of the influencing forces in a system

• Each expert holds a piece of the puzzle

• Conflicting goals

• Each actor has an idea, stake, and interest for the system (mental models). They work to create their
desired change in that system. This can lead to conflicting forces.
• Making mental models explicit.

• Creating consensus & commitment

 Role of the models in SD/GMB

• Using models as learning and communication tools

• Structuring complexity one piece at a time

• Mapping mental models

− How is the system perceived by stakeholders
− Making implicit views of stakeholders explicit

• Tool/Language for understanding complex interconnections

• Initiating dialogue/discussion between stakeholders

• Causalities are neutral and void of conflict

• Cannot deny a systemic effect if we built those causalities together
Steps to create System Dynamics models with a group

1. Define ‘the’ problem variable of the intervention

• Reference mode of behaviour: Showing or drawing the behaviour of the problem
variable over time.

2. Identify any variables somehow related to the problem

3. Identify causal relationships between variables and their polarity

(positive or negative, also: delays)

4. Identify feedback loops that explain the behaviour of the problem

variable

5. Build insights/narratives/forecasts/scenarios by using the model on how

to address the problem
What could possibly go wrong in interventions?

• Commitment form stakeholders

• Access to stakeholders, and/or data
• Modeling does not mean implementation
• Why do they want an intervention?

• Conflict & Power

• Actors’ stakes
• Heated discussions vs Group think

• Inclusiveness of the modeling

• Modeling system vs problem
− Boundary of the model
• Who are involved? What is my bias?
− How would results change if participants would have
changed?
Three major advantages of Group Model Building:
1. Creating new insights by using the role of systems thinking
2. Effective group facilitation;
3. Creating consensus and commitment

Quality of this type of research:

Internal and External Validity?

Reliability?

Useful, relevant, appropriate?

38
4. TYPES OF SYSTEM DYNAMICS MODELS
5. Different types of System Dynamics models

• Causal loop Diagrams (CLD)

• Accessible for non-experts in modelling
• Model denotations: We only use variables and polarities while building models
• Time-sensitive variables are not explained (or they are simplified)

• Stock and Flow Diagrams (SFD)

• Provide more accurate models than CLDs because they differentiate time-sensitive
variables :
− Stocks represent accumulating variables
− Flows represent the influence through which stocks (accumulating) elements change
• This is why SFDs can be hard to understand for certain audiences.
• Model denotations: We use stocks and flows as well as variables and polarities (example
in next slides)
Different types of System Dynamics models
STOCK AND FLOWS
Stocks are accumulations. They characterize the state of the system and generate
the information upon which decisions and actions are based. Stocks give systems
inertia and provide them with memory. Stocks create delays by accumulating the
difference between the inflow to a process and its outflow.

42
System Dynamics Society: www.systemdynamics.org
Inflow and outflow

44
Example

Behavioral risk for

Being infected
infection

Behavioral risk for

Being Exposed Being infected
infection
Example

Community coronavirus model, Tom Fiddleman.

https://metasd.com/2020/03/community-coronavirus-
model-bozeman/
 STRUCTURE AND BEHAVIOUR OF THE SYSTEM

The behaviour of a system arises from its structure. That structure consists of the
feedback loops, stocks and flows, and nonlinearities created by the interaction of
the physical and institutional structure of the system with the decision-making
processes of the agents acting within it.

Fundamental Modes Of Dynamic Behaviour

47
1. Exponential Growth
Exponential growth arises
from positive (self-reinforcing)
feedback.

The larger the quantity, the

greater its net increase, further
augmenting the quantity and
leading to ever-faster.

48
1. Exponential Growth
Exponential growth arises
from positive (self-reinforcing)
feedback.

The larger the quantity, the

greater its net increase, further
augmenting the quantity and
leading to ever-faster.

49
 Positive feedback loops generate
2. Goal Seeking growth, amplify deviations, and
reinforce change.

Negative loops seek balance,

equilibrium, and stasis.

Negative feedback loops act to bring

the state of the system in line with a
goal or desired state.

50
 Positive feedback loops generate
2. Goal Seeking growth, amplify deviations, and
reinforce change.

Negative loops seek balance,

equilibrium, and stasis.

Negative feedback loops act to bring

the state of the system in line with a
goal or desired state.

51
4. S-Shape Growth
Growth is exponential
at first, but then
gradually slows until the
state of the system
reaches an equilibrium
level.

52
4. S-Shape Growth

+
- +

+ +
+

Sales of product with certain life span

https://thesystemsthinker.com/step-by-step-stocks-and-
flows-converting-from-causal-loop-diagrams/

53
Nature of the S-shaped curve
Another example: viral infections
Two ways of using two types system dynamics (CLD and S&F)

1. As a researcher, analyst, etc. As an expert, from paper or data or theory

or…
2. In a facilitated mode, participatory modeling: group model building.

Both can create CLDs and/or S&Fs in various steps.

CLD S&F
Expert mode From theory, articles, From theory, articles,
existing (qualitative existing (quantitative)
and quantitative) data data
Facilitated mode GMB GMB + quantitative
data
Example A
Organisation was in an exponential
growth situation, evaluation showed: Intervention cycle
everything is going great! More Diagnosis

visibility created more assignments.

Evaluate Design the change
Other effects (unexpected or
unforeseen) now hindered the
handling of the amount of generated Implement
the change
work: f.e. communicate about
products, need for growing # of staff,
clearly discerned responsibilities, use
of CMS, product diversity.
Practice oriented research/ interventions, system dynamics
Next week:
O1: The student is able to explain the value and the basic ideas of
practice oriented research and also is able to explain the principal
aspects of system dynamics Final week!
O2: The student can read a simple system dynamics model and can
formulate basic answers on the questions following the model Logic of Inquiry

O3: The student can reflect on the situations that are important for
practice oriented research and/or system dynamics. The student
also can provide arguments for or against the choice for SD in a
given situation.

58