/ 25 Westcott Road
Harvard MA

ENRAP

Knowledge Network ONA:

Report

November 2008 Updated: February 2009

Client: / Shalini Kala, IDRC
Apoorva Mishra, IDRC (Project Lead)
Summary: / Review and analysis of Organizational Network Analysis conducted for the ENRAP Knowledge Network
Net Work
Consultant: / Patti Anklam

Date: / February 21, 2009

February 21, 2008 Page 2

Executive Summary

In the fall of 2008, IDRC requested a network analysis of the IFAD-funded projects that are participating in the ENRAP knowledge networking program. IDRC is in Stage 3 of the IFAD-funded project ENRAP. The goal of ENRAP is to increase the flow and exchange of knowledge across IFAD projects in 18 countries in the Asia/Pacific region. The project has three thrusts:

·  Promote the use of Information Communication Technologies (ICT) to support communication, collaboration, and sharing of ideas and lessons learned as well as to promote outreach across countries in the region to tap into each other for advice and knowledge transfer.

·  Grow the capacity of IFAD staff, project directors, project staff, consultants and NGO partners in the region to be effective knowledge-networkers, such that reaching out and sharing becoming a normal and accepted part of their way of working. This is being accomplished through training and facilitated networking sessions

·  Expand the knowledge base available to the IFAD community in the region by leveraging the technological and social capital being developed as part of this project.

This project surveyed the IFAD staff, country program managers, and IFAD project teams that collaborate across the region to develop a baseline set of metrics against which to measure progress. This report describes the methodology used as well as detailed results of the survey.

Summary Findings

The primary guiding questions for this project were:

·  How do ENRAP participants currently communicate? What are the most frequently used means for communication?

·  How well do participants know each other and interact? Are there sufficient linkages across country boundaries to ensure that information can flow to the people who need it? Are people sufficiently aware of what skills and experience others have to offer so that they can call upon each other for assistance as needed?

·  What value do the participants experience through participating in the ENRAP network?

The response rate to the survey averaged 70% across the different constituencies, with the lowest response rates from Mongolia and the Maldives, Pakistan, Philippines and Sri Lanka. The ENRAP team must ascertain if the response rate is related to English fluency.

Current Communication Means

  • Participants obtain information from a wide variety of sources, most predominantly through email and interpersonal communications, as well work documents via shared files, the IFAD web site and other Internet sites.
  • Participants use all available means for communication – email, phone, face-to-face, and the Internet. Person-to-person interactions are multi-purpose: people seek each other to share ideas, solve problems, ask questions, and obtain resources.
  • There is some use of the ENRAP collaboration sites. There is room here for improvement over time.

Connectivity of Participants

  • The network overall shows no significant disconnections. There are areas of sparse connectivity in remote countries, but the activities of the IFAD and IDRC staff have a definite impact on ensuring that people are connected to the network.
  • Country team members interact with each other on average less than once a month. The groupings of project teams show similar inconsistent contact. It is important for the ENRAP team to determine if greater interaction is desirable and if so, what means are available to support interaction and what motivational initiatives need to be put in place.
  • The analysis of the data show an emerging set of people who are on critical paths of information that cross country boundaries; these people can influence the flow of knowledge. There is perhaps an opportunity to provide additional support to these people, or to acknowledge their roles in a way that will increase the flow of information and knowledge even more.
  • We asked participants to indicate the people with whom they interact most frequently. The respondents to this question tend to have diverse networks, with 35% of the people in their network being from their same group, 68% from the same country. 31% are outside of the IFAD/IDRC/ENRAP network.

It is possible that the connectivity numbers are skewed by the language issue. That is, people who speak English may perhaps be more connected outside of their own countries, whereas the non-English speakers who may not have taken the survey would tend to be more disconnected.

Value of ENRAP to IFAD Project Projects and Country Teams

  • The averages of responses to statements about the value of ENRAP range from 4 (neutral) to 4.9 (slightly agree). Agreement is higher among the countries that began participating in ENRAP in Phase II.

The details for each of these areas of analysis are provided following the next section, which details the methodology used for the network connectivity questions.

Methodology

For the connectivity questions, the IDRC/ENRAP team decided to use the Organizational Network Analysis (ONA) methodology, which entailed a survey that asked people in the network specific questions about whether they knew and/or interacted with the other people in the network. The survey produces metrics that can be used as a baseline against which to measure improvements over a time period to be determined (nine months to one year). The baselines we will use for comparison include:

·  Expansion in ways that people communicate across boundaries, specifically if there is greater use of the IFAD web site and collaboration spaces.

·  How well the ENRAP project is meeting its goals of connecting people across country and cultural boundaries.

·  Whether individuals have enriched their personal networks of contacts.

In doing such ONA analyses at different points in time, it is possible to show both diagrams illustrating the changes.

Principles of Organizational Network Analysis

ONA is based on the assumption that we can detect patterns in the interactions of people in an organization, group, or network and that these patterns can help to diagnose where there are:

·  Too few or nonexistent connections between groups where connections may have been assumed to exist or where connections must exist in order for the organization to be successful

·  People who are overly central to a network, leading to bottlenecks or overload

·  People who are key to moving information and knowledge around the network

·  The overall connectivity of the network for comparison to benchmark metrics


The use of ONA is predicated on a small number of principals:

  • It is possible to obtain the data that shows interconnections
  • The data collected represents a single moment (a “snapshot”) in time
  • The data and maps produced by the data provide insights into problems and opportunities
  • The insights can be probed by asking good questions and seeking the answers to those questions
  • The data produced can be used as a baseline

How an Organizational Network Analysis is Conducted

A typical Organizational Network Analysis (ONA) begins with data collection. Data can be collected:

  • With surveys
  • Via interviews
  • Mining email and other online interaction data

For this project, we determined that surveys would be the most efficient method of reaching a large number of people quickly. An ONA survey asks people questions about their specific relationships with others, for example the screen shot below illustrates how a person might see a specific survey question on the online survey form:

Each question represents one way that people have of describing their relationship with each of the other people. Choices are presented in menus; each response equates to a numeric value. We collect demographic information so that we can distinguish different subgroups or functions in an organization. The data collected from the survey is analyzed by special tools that create map; the numeric value produce metrics.

Looking at an ONA Map

One potential map that might result from this analysis is shown below:

This map illustrates a number of concepts about ONA.

1.  Each line in this drawing shows responses of “Strongly agree” (value of 5) to the question, “How well do you know this person’s knowledge and skills?”

2.  An arrow shows who answered the question about whom. For example, MA says that he knows the knowledge and skills of KM very well. CMA and JS each answered “strongly agree” when asked about each other. This map does not indicate whether people are friends or colleagues – it only shows those relationships in which a person says that they know the other person’s knowledge and skills very well. Colors indicate subgroups within the organization.

3.  The size of a circle (a “node”) can be used to indicate a level in the hierarchy, or can be used to represent some other value.

4.  We can ask good questions about relationships by looking at the map. It is easy to see that“VM” must be a key person in the network, because many people know him. We can ask the question, “What is his role?” We notice that the group in green (lower left) is connected to the rest of the group through VM, but only by two people (PA and BS). We could ask the questions, “Why is that?” and also “Is it important that these people be connected to people in the other two groups?”

5.  We notice people who are little connected. For example, JM appears to know only VM. Questions might be, “Is JM new to the organization?” “Does JM have a distinct and unique job that doesn’t require her to know or work with other people in the organization?” “Do we need to connect JM into the network?”

The primary result of a successful ONA is that it leads to both (1) good questions, and (2) action. The questions should lead to actions. In our example, perhaps there needs to be a group event at which people will have an opportunity to talk about their work so that others can know what they do. (That is, in fact, what this group did following their ONA.)

Each survey can result in literally hundreds of maps. There are five or six possible maps for each question. For example, we could show the map of whether people even know each other all (a value that is greater than 1), whether they know the person even slightly (a value greater than or equal to 1), and so on. The figure on the right here is the map of all relationships of “somewhat agree” that they understand this person’s knowledge and skills.” You can see that by adjusting the value represented by the menu choice that the network is very different.

We can create separate maps for subgroups, for example, if we want to look at only one organization, one level of hierarchy, or any other attribute that we collect in the demographic information.

Network Metrics

In addition to maps, the data from the survey also provides metrics. There are numerous mathematical algorithms that can be applied to the data. These algorithms count the number of ties, numbers and directions of ties between people, and relationships that span many “degrees of separation.” In an ONA, we look particularly at two sets of metrics. These metrics provide the baseline that we can use to determine, in future, what the changes have occurred in the network.

Connectivity

Network metrics indicate the overall connectivity of a network are density and distance.

Density is the percentage of connections that exist out of a possible 100% (everyone is connected to everyone else). In our examples above:

·  The density of the network in the first diagram is 11% . This is the percentage of the people who know each other “very well.”

·  When we looked at the map showing relationships of “somewhat know” the density is 51%.

In these examples, it is easy to see from the diagrams the difference in density. In diagrams where there are large numbers of people, it is not easy to see from the map so having the metrics is very important.

Distance is the average number of degrees of separation between any two people in the network. That is, how many steps does it take to get from one person to another. In our network of people who know each other very well, the distance is 2.2; to reach “somewhat well”, the distance is 1.5. You can see from the map that most people need to go through VM and possibly one other person to find out the knowledge and skills of someone else. It is not always easy to see this in a map with many people, so it’s important to have the metric.

Centrality

Centrality metrics indicate the key people in a network. A simple metric is to count the number of lines going into each person. This is a Degree Centrality metric. Sometimes you can “count” this number by looking at a map, but if the network is very large you cannot easily count.

More sophisticated calculations are not discernable from the map. The resulting metrics allow us to see which people are better positioned to:

·  Move knowledge and information around in a network. These are people who are on the most paths across the network from one person to another. This is the Betweenness Centrality. In our example, VM has the highest betweenness centrality, followed by SB and PA.