And the second approach that we´ll be discussing today is network analysis.

And network analysis allows you to understand

who are the actors in a network.

These can be either individuals or organizations.

How they are related, and what are the characterizations of their relationships.

In addition, network analysis allows you to understand

how different actors share information or resources among them.

And to read a network diagram, I'll introduce you to the key network elements.

These are the nodes, and

these are usually represented by circles or some sort of geometric form.

The edges, which are the lines in between them, and both nodes and

edges will have different attributes which will be represented by size and color.

So you can make the nodes bigger depending on, let's see, characteristics

such as population size if we're talking about a network of countries.

And you can color code the different relationships that might be available.

You can also denote directionality, so for example,

in the bottom part of the diagram, you see that from the white circle,

there's an arrow pointing to the gray circle.

So this denotes that the white circle shared something with the gray circle,

or the white circle defined an existing relationship between them and

the gray circle.

So these are the main components if you want to read a network diagram.

In terms of the data that is used for network analysis,

there are two types of data that are worth mentioning.

One is egocentric data, and basically this is obtained by going to one person,

in this case, person a, and asking them to provide you information about

all the other persons in their network.

So this is pretty open ended, and there's no confirmation from b,

c, d, and e about what person a said.

The second type of network data is sociometric data.

And in this case information is obtained from all the actors in the network.

So you begin by drawing clear boundaries around the network of interests, and

then collecting data from each point about all the other points in the network.

And as you can imagine, the sociometric data is much more complete.

It can also be more difficult to collect than egocentric data.

But depending on how well you know your network boundaries,

you'll decide which one fits the question of interest better.

And once you have your network data, there are several questions that you can answer,

and several measures that one can compute, and I won't go into these in great detail,

but I'll just highlight the main ones.

So, for example,

you might be interested in which are the most important individuals.

So, for example, you might be interested in who

are the most important individuals in your network.

And you might want to know this, because in an epidemic for example,

you want to know who are the people that are spreading the disease the most

in their network, for example.

You might also want to know how tight is the network, how dense it is.

You might be interested in how many degrees of separation there are between

two actors.

Because that has a role in how information is shared and how long that takes.

In that case you might be interested in the distance.

And you also might be interested in knowing to what degree similar actors

stick together.

Are people of the same age sharing more information than otherwise?

And you might also want to know the extent to which two actors

reciprocate each other's interactions.

So if I share resources with someone, there's that frantic knowledge that

I shared resources, and do they share as well?

So, these are some of the questions that might be interested in asking.

And I didn't go through these in very good detail,

so I'm referring you to a couple of resources here.

One is a textbook that is available digitally, and the other one is a video.

And the final tool or approach that we will be discussing today is

the Participatory Impact Pathways Analysis.

So we'll call it PIPA for short.

And this is an approach that can be used as a project planning and

both as an M&E approach.

And what it does is, as you can tell from its title,

it brings together key stakeholders, and it facilitates them through

a process to identify a vision, usually for one and a half years in advance.

Because it might be more difficult to predict it longer term.

Then it goes through identifying actor types, as well as actor relationships.

And then there's a process of developing network maps.

So mapping these actors in relationships, both now and

in this future where the vision is realized.

And the facilitated discussion that follows is around identifying and

breaking down the pathways that are needed for the change to happen.

And there are similar tools, and that map is another approach that is quite similar,

but maybe has less emphasis on the M&E and outcome pathways pieces.

So, you can see this is the NOW map that was produced by this particular project,

and this is the FUTURE map.

And it's quite messy, and it's, to a foreign eye,

one that wasn't involved in the process, might not make a lot of sense.

But what happens after this is drawn is that each of these relationships

is then put into UCINET, which is a software for analyzing network data.

And based on that, you won't be

necessarily developing quantitive measures, but you can use the visuals

in a more easy fashion to understand how the pathways have changed.

So, in this case, we can see the central point, the white point, was a receiver of

funding from the red dots, and then a giver of the funds to all the other ones.

In the future, we see that the white organization still receives the funding,

but that there are more of the blue dots,

the blue organizations, that receive funding.

And the vision for this project was that eventually the blue dots would be

the ones serving the role of the white dot.

And this approach, actually there has been quite a lot of documentation on it.

It has been used a lot in agriculture, and I'm providing you with two resources here

in case you want to learn a little bit more.

So in conclusion, we'll finish up with some summary or key points.

And that is to highlight that stakeholder engagement is very important, especially

when you don't know who the stakeholders are, or what their interests are.

And also, when you might face stakeholders with diverse perspectives and

different mental models, both which make it difficult to obtain buy-in,

to build trust, and to reach consensus.

Approaches that you use to engage with stakeholders depend on the purpose and

the problem that you have at hand, as well as the resource and research constraints.

And we'll talk a bit about those in a minute.

But the three main approaches that we went over today, were stakeholder analysis,

which is used to identify stakeholders and

anticipate their reactions to a problem policy or system change.

We discussed a bit network analysis, which is used to identify key stakeholders and

quantify their relationships.

As well as if done more than once, you can also model these relationships over time.

And finally, we discussed the PIPA approach,

which is a participatory approach that can be used to identify stakeholders and

link them with pathways of change that can then be monitored.

So it's very much an ongoing engagement to monitor change.

And finally, I made brief mention of group model building,

which will be covered in subsequent lectures.

Which helps you bring together stakeholders to collaborate,

and works towards a common model.

And through using systems diagrams, and system dynamics models, or agent-based

models, it helps stakeholders together decide on and discuss different scenarios.

And what's important to remember is that all of these

stakeholder engagement approaches, there are different

dimensions to consider when deciding which one to apply and how to apply it.

So, for example, these approaches have different resource needs,

both in terms of time and and financial resources.

They also involve different levels of participation.

Though, for example, the network analysis,

especially if you're doing a sociometric network analysis, it can be quite high,

perhaps because it needs somebody knowing how to collect the data,

how to analyze it, and it might need some sophistication or expertise.

It can be very interesting, but it doesn't require any participation.

On the other hand, group model building,

it's the most participatory way to engage with stakeholders among these tools.

But that resource is necessary to stakeholders over a certain

period of time.

PIPA stakeholder analysis are maybe similar but

a little bit less resource intensive.

But what's important to keep in mind is that although stakeholders

can start blurring the boundaries of your problem at hand,

and can require more resources to be engaged,

to achieve trust and buy-in in the long run,

it's important to invest the necessary resources.

And I'll leave you with a note from Peter Senge's Fifth Discipline.

His 11th law says that faster is slower.

And what that means is that sometimes,

especially when it comes to engaging stakeholders, it's important to

devote the time and necessary resources to achieve the goals in the long run.

Because if you do it faster, and if you cut corners,

it might actually overall take longer to achieve your goals than otherwise.

So these are just a few resources.

I hope you'll be interested to read more about the various methods that we

went over today, and as I mentioned before,

group model building will be discussed in the next couple of lectures.

[MUSIC]

Lecture 2B: Network Analysis, Participatory Impact Pathways Analysis, and Summary

Systems Thinking In Public Health

Meet the Instructors