Roanoke College Partnership Network

Key Connections & Insights

An analysis of the inter-organizational partnership network celebrated at Roanoke College's Spring 2024 event, revealing the structural patterns, hub organizations, and thematic clusters that define the region's collaborative ecosystem.

Network at a Glance

🔗

25 Partner Organizations

The network comprises 25 organizations invited to the Celebration event based on their partnerships with Roanoke College. Of these, 13 were "starred" — the primary celebrated partners — while 12 additional organizations were recognized as connected partners within the broader ecosystem.

🌐

49 Direct Connections

Across the 25 core partners, 49 unique connections were identified between organizations, meaning many partners don't just connect to Roanoke College individually — they are interconnected with each other. This creates a web-like structure rather than a simple hub-and-spoke.

🔭

42 Extended Partners

Nine of the celebrated partners reported a total of 42 additional organizational connections not already in the core network. These represent the second-order reach of the partnership ecosystem — organizations accessible through the existing partners.

📊

~16% Network Density

With 49 connections out of a possible 300 (25 × 24 ÷ 2), the network has a density of approximately 0.16. This is typical of collaborative networks: well-connected enough to facilitate resource flow, but not so dense that every organization connects to every other — leaving room for brokerage roles.

Hub Organizations

Five organizations emerge as structural hubs based on their degree centrality — the number of direct connections they maintain within the network. These hubs are the organizations through which information, resources, and collaborative opportunities most readily flow.

Carilion Clinic

13 connections — highest degree centrality

The single most connected organization in the network. Carilion connects across every thematic cluster: health services, higher education, research, and community organizations. It bridges the health/clinical world to the education/cultural world, giving it likely high betweenness centrality as well. Its connections span Fralin Biomedical Research Institute, Radford University, Virginia Tech, VWCC, Physicians to Children, CHIP, and Blue Ridge Literacy — research, education, direct services, and community literacy all in one ego network.

Healthcare Higher Ed Research Community Services

Virginia Western Community College

12 connections

Nearly as connected as Carilion, VWCC serves a distinctive bridging role as a community-accessible educational institution. It connects to both the major health/research corridor (Carilion, Fralin, Virginia Tech) and the community services corridor (Blue Ridge Literacy, Roanoke City Public Schools, Greater Roanoke Workforce Development Board). VWCC's position makes it a critical pathway between workforce development and the broader partnership ecosystem.

Higher Ed Workforce Healthcare

Radford University, Blue Ridge Literacy, & Roanoke City Parks

10 connections each

These three form a co-equal third tier of hubs, each serving different structural roles. Radford connects higher education institutions to each other and to community organizations. Blue Ridge Literacy bridges immigrant/refugee services (Casa Latina, CHIP) to education and cultural institutions (Taubmann Museum, VWCC). Roanoke City Parks is the primary connector for environmental and civic organizations, linking to Sister Cities, Public Schools, and the cultural sector.

Education Literacy & Immigration Parks & Culture

Thematic Clusters

The network naturally organizes into several thematic clusters — groups of organizations that are more densely connected to each other than to the rest of the network. These represent distinct but overlapping domains of partnership activity.

Health & Human Services Cluster

BREAST Roanoke, Carilion Clinic, CHIP Roanoke Valley, Physicians to Children, The Birth Nurse LLC, HCA LewisGale, and Virginia Department of Health form a tightly connected health services cluster. BREAST Roanoke serves as a specialized hub within this group, connecting maternal health organizations (The Birth Nurse, CHIP) to the major health systems (Carilion, HCA LewisGale). This cluster extends into 5 additional maternal/child health organizations through BREAST Roanoke's extended network.

Higher Education & Research Corridor

Radford University, Virginia Tech, Fralin Biomedical Research Institute, George Mason University, Mary Baldwin University, University of Lynchburg, Roanoke Higher Education Center, and VWCC form the education and research backbone. This cluster is notable for its geographic range — extending from the Roanoke Valley to Northern Virginia — and its connections to both clinical (Carilion) and community (Blue Ridge Literacy, Parks) organizations. The Roanoke Higher Education Center connects to 5 of these institutions as a physical convening point.

Arts, Culture & Civic Engagement Cluster

Taubmann Museum, Roanoke Valley Sister Cities, Roanoke City Parks, and Roanoke City Public Schools form a cultural and civic cluster. The Taubmann Museum alone connects to 7 additional cultural organizations (SWVA Ballet, Local Colors, Harrison Museum of African American History and Culture, Roanoke Children's Choir, among others), making it the richest source of extended cultural connections. This cluster bridges to education through VWCC and Radford, and to health through Carilion.

Community Services & Social Support Cluster

CareForward, Blue Ridge Literacy, Casa Latina, and the Greater Roanoke Workforce Development Board represent the community services dimension. CareForward is notably the only starred partner with zero connections to other partners within the core network — its partnerships (Renovation Alliance, Roanoke City and County DSS, Sunset Ministries) exist entirely within its extended network. This structural isolation makes CareForward a potential target for deeper network integration.

Structural Insights

Bridge Organizations

Carilion Clinic and VWCC serve as the primary bridge organizations, connecting clusters that would otherwise be more isolated. If either were removed, the health/human services cluster would become significantly more disconnected from the education/cultural clusters. This makes these two organizations structurally critical to the network's overall cohesion.

Isolated & Peripheral Nodes

Several organizations occupy peripheral positions with only 1–2 connections: Casa Latina (1), The Birth Nurse LLC (1), Virginia Department of Health (1), Physicians to Children (1), Greater Roanoke Workforce Development Board (1), and Roanoke Valley Sister Cities (2). Additionally, CareForward and Reserva Araponga have zero connections to other core network partners. These peripheral positions don't indicate lesser importance — they may indicate untapped potential for integration.

The Extended Network Effect

The 42 extended partners dramatically expand the network's reach: national health workforce organizations (through ASPPH), environmental stewardship (through Roanoke City Parks), cultural arts (through Taubmann Museum), and food security (through Radford University's connections to LEAP and Feeding SW Virginia). This penumbra network represents the partnership ecosystem's latent capacity — organizations that could be activated for new collaborative initiatives.

Degree Centrality Rankings

Organization	Type	Connections
Carilion Clinic	★ Celebrated	13
Virginia Western Community College	★ Celebrated	12
Radford University	★ Celebrated	10
Blue Ridge Literacy	★ Celebrated	10
Roanoke City Parks	★ Celebrated	10
Taubmann Museum	★ Celebrated	9
Virginia Tech	Network Partner	5
BREAST Roanoke	★ Celebrated	5
Roanoke Higher Education Center	Network Partner	5
Roanoke City Public Schools	★ Celebrated	4
Fralin Biomedical Research Institute	★ Celebrated	4
Assoc. of Schools & Programs of Public Health	★ Celebrated	3
CHIP Roanoke Valley	Network Partner	3
George Mason University	Network Partner	2
HCA LewisGale	Network Partner	2
Roanoke Valley Sister Cities	★ Celebrated	2
Casa Latina	Network Partner	1
The Birth Nurse, LLC	Network Partner	1
Virginia Dept. of Health	Network Partner	1
Physicians to Children	Network Partner	1
Greater Roanoke Workforce Dev. Board	Network Partner	1
CareForward	★ Celebrated	0
Reserva Araponga	★ Celebrated	0
Mary Baldwin University	Network Partner	0
University of Lynchburg	Network Partner	0

About This Sociogram

A guide to understanding the network analysis methodology, data structures, and visualization techniques used to map the Roanoke College partnership ecosystem — and how this approach fits within the broader tradition of Social Network Analysis (SNA) in qualitative and mixed-methods research.

What Is a Sociogram?

The interactive network visualization on the first tab is a sociogram — a graphical representation of social relationships and connections within a defined group. The term was coined by Jacob Moreno in the 1930s as part of his development of sociometry, the quantitative study of social relationships. In its simplest form, a sociogram depicts individuals or organizations as points (nodes) and their relationships as lines (edges or ties) connecting them.

This particular sociogram depicts an inter-organizational network: the unit of analysis is the organization rather than the individual, and the ties represent acknowledged partnership connections between organizations. This places the visualization within the tradition of organizational sociology and institutional network analysis, distinct from the interpersonal networks Moreno originally studied but using the same foundational visual grammar.

Key Distinction

A sociogram is the visual representation of a network. Social Network Analysis (SNA) is the broader methodological and analytical framework that encompasses the data collection, mathematical analysis, and interpretation of network structures. This visualization is a sociogram produced through SNA methods.

Data Structure: The Sociomatrix

The underlying data for this sociogram comes from two complementary structures captured in the original spreadsheet, both standard in Social Network Analysis:

Node Attribute Table (Sheet 1)

The first sheet lists all 25 partner organizations, serving as the node list — the roster of actors in the network. Each organization is a node, and attributes (such as whether it is a "starred" celebrated partner) are node-level properties. In SNA terminology, this is the vertex attribute table.

Adjacency Matrix / Sociomatrix (Sheet 2)

The second sheet is a sociomatrix (also called an adjacency matrix) — a square matrix where organizations appear on both axes and cell values indicate the presence or absence of a connection. In this case, green-filled cells indicate a tie between the row organization and the column organization. This is the foundational data structure for any network analysis, and it can be directly imported into SNA software such as Gephi, NodeXL, UCINET, or Pajek. The sociomatrix here is symmetric (undirected): if Organization A is connected to Organization B, the reverse is also true.

Extended Network Data (Column AA)

An additional column captures second-order connections — organizations that are partners of the starred partners but are not themselves part of the core 25-node network. In network science, these represent the penumbra or ego network extensions, and they illustrate the broader reach accessible through the existing network without requiring direct ties.

Visualization Method: Force-Directed Layout

The interactive sociogram uses a force-directed graph layout algorithm (specifically D3.js's force simulation) to position nodes. This is one of the most widely used layout algorithms in network visualization and works by simulating physical forces:

How the Layout Works

Repulsion: All nodes repel each other (like charged particles), preventing overlap and spreading the graph out.

Attraction: Connected nodes are pulled toward each other (like springs along the edges), so linked organizations appear closer together.

Equilibrium: The simulation runs iteratively until the forces reach a stable balance, producing a layout where proximity roughly corresponds to network distance — organizations with more shared connections tend to cluster together visually.

The result is that visual clusters in the sociogram correspond to structural clusters in the data. Organizations that appear close together share many connections, while distant organizations are more structurally separated. However, because the algorithm involves random initial placement, the exact layout varies each time the page loads — the structure is preserved but the orientation is not fixed.

Visual Encoding

The sociogram uses several visual properties to encode network information, each corresponding to a standard SNA concept:

Node Size → Degree Centrality

Larger nodes have more connections (higher degree centrality). Degree centrality is the simplest and most intuitive centrality measure in SNA: it counts the number of direct ties each node maintains. In this network, Carilion Clinic's node is the largest because it has 13 connections — the highest degree centrality.

Node Color → Organizational Role

Teal nodes represent the 13 "starred" celebrated partners — the primary organizations recognized at the event. Gold nodes represent the 12 additional network partners invited based on their connections to the starred partners. Red nodes (visible when the extended network is toggled on) represent the 42 second-order partners reported by starred organizations.

Edges (Lines) → Partnership Ties

Each line represents an acknowledged partnership connection between two organizations. The ties are undirected (no arrow), meaning the connection is mutual. In SNA terminology, these are binary, symmetric ties — either a connection exists or it doesn't, with no weighting for strength or directionality.

Hover Highlighting → Ego Network

When you hover over a node, the visualization highlights that node's ego network — the focal node and all nodes directly connected to it, with all other nodes dimmed. Ego networks are a fundamental concept in SNA, representing the local neighborhood structure of any given actor.

Key SNA Metrics Represented

Metric	Definition	In This Network
Degree Centrality	Number of direct connections a node has	Ranges from 0 to 13 (Carilion Clinic)
Network Density	Ratio of actual ties to possible ties	49 / 300 = 0.163 (~16%)
Network Size	Number of nodes in the network	25 core + 42 extended = 67 total
Components	Disconnected subgroups within the network	One main component (21 nodes) + 4 isolates
Isolates	Nodes with zero connections to other core members	CareForward, Reserva Araponga, Mary Baldwin, U. of Lynchburg

Whole Network vs. Ego Network Analysis

This visualization represents a whole network analysis (Provan & Kenis, 2008) — the analytic focus is on the structure of the entire partnership ecosystem, including all ties among all organizations. This is distinct from an ego network analysis, which would focus on a single organization (e.g., Roanoke College) and examine only its direct partners. The whole network approach reveals emergent structural properties — clusters, bridges, isolates — that are invisible from any single organization's perspective.

The fact that this network was generated from a Celebration event place mat — where partners themselves identified their connections — makes this a form of participant-generated network mapping. This is methodologically significant because it captures emic (insider) perceptions of network structure rather than imposing external definitions of what constitutes a tie. The data reflects how organizations perceive their connections, which may differ from what formal agreements or co-funding relationships would show.

Analytical Lineage & Related Methods

This sociogram draws on several overlapping traditions in qualitative and mixed-methods research:

Social Network Analysis (SNA)

The primary methodological framework. SNA encompasses the full range of techniques for collecting, analyzing, and visualizing relational data. Key foundational works include Freeman (1979) on centrality measures, Wasserman & Faust (1994) on social network methods, and Borgatti, Everett & Johnson (2018) on analyzing social networks.

Collaborative / Partnership Network Mapping

In the context of multi-organizational partnerships, network mapping is used to understand collaborative structures, identify coordination gaps, and plan for network development. Provan & Kenis (2008) distinguish between shared governance networks, lead organization networks, and network administrative organizations — each with distinct structural signatures visible in sociograms.

Community-Based Participatory Research (CBPR)

The place mat data collection method — where partners self-reported their connections during a community event — aligns with CBPR principles of participant engagement in data generation. The network structure reflects community members' own understanding of their relational landscape rather than researcher-imposed categories.

Qualitative Network Visualization

While traditional SNA emphasizes quantitative metrics, there is growing use of network visualization as a qualitative tool — using the visual structure of sociograms to identify patterns, generate hypotheses, and facilitate discussions with stakeholders. This sociogram serves that purpose: it makes visible the structure of a partnership ecosystem that would be difficult to apprehend from lists or tables alone.

Tools & Compatibility

The data underlying this sociogram is stored in a standard sociomatrix format that can be directly imported into dedicated SNA software for more advanced analysis:

Software Compatibility

Gephi — Open-source network analysis and visualization platform. The sociomatrix can be imported as a CSV adjacency matrix or converted to GEXF/GraphML format for full metric computation including betweenness centrality, modularity-based community detection, and publication-ready visualizations.

NodeXL — Excel-based SNA tool. The existing spreadsheet format is nearly directly importable with minor restructuring into NodeXL's edge list format.

UCINET — The standard academic SNA software package, particularly strong for computing network-level statistics, positional analysis, and blockmodeling.

R (igraph / statnet packages) — For programmatic analysis, the adjacency matrix can be read directly into R for computation of any network metric, statistical modeling (e.g., ERGM), and automated community detection.

Limitations & Considerations

Self-reported ties. The connection data was generated by partner organizations during a single event, reflecting perceived connections at one point in time. Organizations absent from the event could not report their connections, and social desirability may influence which ties are reported.

Binary ties only. The sociomatrix records only whether a connection exists or not — it does not capture tie strength (how deep or active the partnership is), tie type (funding, referral, co-programming, etc.), or directionality (who initiated, who benefits more). A weighted or multiplex network analysis would reveal richer structural dynamics.

Snapshot, not longitudinal. This represents the network at one point in time. Partnership networks are dynamic — new ties form, existing ties strengthen or dissolve. Repeated measurement would enable analysis of network evolution.

Roanoke College as implicit center. All 25 organizations were invited based on their connection to Roanoke College, meaning the college is an implicit hub not depicted as a node. In a complete representation, Roanoke College would connect to all 25 organizations and would have the highest degree centrality by definition.

Suggested References

Borgatti, S. P., Everett, M. G., & Johnson, J. C. (2018). Analyzing Social Networks (2nd ed.). SAGE.
Freeman, L. C. (1979). Centrality in social networks: Conceptual clarification. Social Networks, 1(3), 215–239.
Moreno, J. L. (1934). Who Shall Survive? A New Approach to the Problem of Human Interrelations. Nervous and Mental Disease Publishing.
Provan, K. G., & Kenis, P. (2008). Modes of network governance. Journal of Public Administration Research and Theory, 18(2), 229–252.
Wasserman, S., & Faust, K. (1994). Social Network Analysis: Methods and Applications. Cambridge University Press.
Borgatti, S. P., & Halgin, D. S. (2011). On network theory. Organization Science, 22(5), 1168–1181.
Luke, D. A., & Harris, J. K. (2007). Network analysis in public health. Annual Review of Public Health, 28, 69–93.