SDI Theory of Change V3
Sustained Dialogue Institute

Theory of Change and Evaluation Framework

Version 3 · March 2026 · Working Draft

People who disagree with each other, who hold different identities and different stakes in the same institutions, regularly fail to solve problems together. This is not because they lack information or good intentions. It is because the relationships between them are broken, shallow, or structured in ways that prevent collaboration.

SDI exists because this relational failure has a remedy. The remedy is not a workshop or a set of tips for talking across difference. It is a sustained process, developed from international peace process work, that transforms how groups of people relate to each other over time. When that transformation happens, something lasting changes in the people who go through it: they become more capable of acting in public life, negotiating with people who hold power, navigating institutions, and organizing change across difference.

Where the Process Comes From

Harold Saunders, SDI's founder, helped broker the Arab-Israeli peace process and later designed unofficial dialogues between American and Soviet citizens during the Cold War (the Dartmouth Conference) and between warring factions in Tajikistan's civil war. From that work, he developed a structured process for transforming how groups in conflict relate to one another.

Saunders found that transforming relationships follows a recognizable pattern and works on specific dimensions of how people relate. He developed a five-stage process and a theoretical framework, the Concept of Relationship, for understanding what the process changes and why. But his ambition was never limited to dialogue groups. In Politics Is about Relationship (2005), he argued that the relational work of citizens outside government is essential to democratic functioning.

The Concept of Relationship

Before describing what SDI does and what it produces, it is necessary to describe how SDI sees. The Concept of Relationship is the analytical framework that underlies all of SDI's work, from the first intake conversation with a campus partner through the most advanced institutional embedding. It is not something that happens only inside a dialogue group. It is a lens for reading any relational system.

Saunders identified five components that shape how groups relate to one another. These are not traits of individuals. They describe the forces operating in the space between groups: the relational field that peacebuilding, facilitation, and institutional change must engage.

Concept of Relationship: Five Components of the Relational Field
Identity
Group memberships, histories, and affiliations shaping how people see themselves and are seen by others in a conflict
Interests
What groups actually need from a situation, as distinct from the positions they publicly hold
Perceptions
Assumptions, narratives, and stereotypes groups hold about one another, often without direct contact
Power
Institutional, social, and positional forces determining who sets terms, who opts out, whose actions reshape the landscape
Patterns of Interaction
Established structures through which groups engage or avoid each other: formal channels, protest, silence, trust networks
These components interact simultaneously in every relational system SDI engages.
Based on Harold H. Saunders, Politics Is about Relationship (2005). The Concept of Relationship applies to international peace processes, campus conflicts, and institutional partnerships alike.
Why this matters for campuses

When a university administration revokes a student organization's recognized status during a conflict, that is a power move that restructures every other relationship on campus. When a student peace-builder is perceived as an administrative plant because of his religious identity, that is perceptions shaping the relational field before anyone enters a room. When groups default to protest or avoidance rather than direct engagement, those are patterns of interaction that Stage 1 must address. The Concept of Relationship is how SDI reads these dynamics and designs interventions that account for them.

The Five Stages

The five stages describe what human beings seem to do when they address destructive or unproductive relationships over time. They overlap, repeat, and deepen. All five components of the Concept of Relationship are active from Stage 1 forward. What changes across the stages is the depth and intentionality with which the relational field is engaged.

Five Stages of Relational Transformation
1Deciding to Engage
2Mapping the Relationship
3Probing Root Causes
4Scenario-Building
5Acting Together

Stage 1 in Practice

Stage 1 is often the most complex and time-consuming stage. It involves community mapping, identifying whose presence and absence at the table matters, assessing barriers for both moderators and participants, and building enough trust that people are willing to show up. Facilitators draw the relationships between groups, identify who holds power, whose voices are typically absent, and conduct pre-conversations to understand what would need to be true for people to gather.

In contested environments, Stage 1 may require months of shuttle diplomacy, multiple failed entry points, and repeated trust-building before a formal dialogue group can form.

From the field: Stage 1 under high-conflict conditions

In a case study following October 7, 2023, the Stage 1 process on one campus spanned an entire academic year. It involved partnership development with an interreligious council, survey initiatives that stalled, a student-led panel event, an encampment and institutional crackdown that restructured the power dynamics, and eventually a new round of student leadership that reopened possibilities the following semester. None of this was failure. It was Stage 1 doing what Stage 1 does in a high-conflict environment.

Topics are always locally chosen. Trained moderators determine whether a dialogue group is needed and what it addresses. For large campuses where students are not self-organizing, SDCN has found that using digital tools to surface community tensions or offering a specific topic can lower the barrier without undermining the locally driven principle.

Civic Agency: What the Process Produces

The outcome SDI's process produces has a name: civic agency. Civic agency is the capacity to act as a co-creator of democratic life, not just a participant in it. It encompasses the skills, motivation, and sense of power that enable a person to engage effectively in the public sphere.

The SD process produces civic agents: people who carry the capacity to transform relationships and navigate public life into their careers, communities, and civic roles for years after the dialogue ends.

Most dialogue and bridging programs claim attitudinal change: participants feel warmer toward outgroups, more willing to engage, more intellectually humble. SDI can claim those outcomes too. But attitudinal shifts are proximal indicators. SDI's distinctive claim is that the process produces people who carry the capacity for relational transformation long after the dialogue ends.

The Civic Agency Measure (Wegemer, Wray-Lake, Hope et al., 2025) provides a validated 16-item instrument for measuring this outcome across four dimensions:

DimensionWhat It MeansHow It Shows Up
CompetenceSkills and knowledge to participate in civic lifeA moderator holds space for conflict. Faculty use conflict curricula to prepare students for high-stakes disagreements.
DriveMotivation and responsibility to actWhat gets people through the door. Under pressure when students overcommit, faculty do not sign up, or international students feel unsafe.
Individual PowerThe sense that one can influence outcomesStudents who did SD run for student government. A Fellow designs a faculty mediation system.
Collective PowerThe belief that people working together create changeA dialogue group undertakes joint action. A campus runs SD independently. Most dependent on institutional conditions.

Research Evidence

SDI's evidence base includes longitudinal research, a controlled experiment, pedagogical analysis, case documentation from contested campus environments, and a validated measurement instrument.

DomainFindingStudy
Civic BehaviorAlumni reported a "restringing" effect: pervasive changes across civic, social, and workplace domains, affecting career choices, civic participation, and family aspirations years after dialogue endedDiaz & Perrault (2010)
Interethnic TrustSD decreased mistrust and increased trust between ethnic groups while increasing participants' awareness of their own group identity, without producing aggression (controlled experiment, 716 students)Svensson & Brouneus (2013)
Pedagogical DesignRecruitment, purpose framing, time structure, and relationship-building directly shape outcomes. Three-hour session blocks necessary; shorter sessions forced trade-offs between relationship-building, skill-building, and dialogueFlint, McLelland, & Kouaho (2025)
Stage 1 ComplexityHigh-conflict campus environments require months of trust-building, multiple entry points, and sensitivity to institutional power dynamics before formal dialogue can proceedSDI Case Study (2025)
Civic AgencyCompetence, drive, individual power, and collective power now measurable with validated 16-item instrument (N > 1,700)Wegemer et al. (2025)

Four Levels of Campus Work

These levels describe where SDI intervenes, not what it produces. What it produces is civic agency. The levels describe the contexts in which civic agency develops and the institutional conditions that support or hinder it. The Concept of Relationship is the analytical framework applied at every level.

Four Levels of Work Mapped to Partnership Tracks
Level
What Happens
What SDCN Provides
Partnership Track
Level 1
Preparing the Ground
Creating conditions for people to show up. Orientation dialogues, student-led panels, digital tools to surface tensions, post-election processing.
Phone-based facilitation buddy, orientation toolkits, workshop designs, monthly info sessions.
Foundational
Primary focus for campuses where gathering itself is hard.
Level 2
Experiencing Dialogue
Full five-stage process with formed groups. Small groups of six working through sustained engagement over weeks and months.
Dialogue group design support, facilitation training access, topic and structure guidance.
Foundational
First dialogue experiences alongside Level 1 work.
Level 3
Building Local Capacity
Creating local people who can do Levels 1 and 2 without SDCN. Moderator training, faculty development, Leaders Summit pipeline.
Facilitation training, Leaders Summit, moderator coaching, CultureShift Fellow nominations, Help Desk pilot.
Emerging
Campus has activity and is building practitioners who sustain it.
Level 4
Embedding Capacity
Making dialogue sustainable in institutional structures. Curriculum integration, CultureShift Fellows de-siloing, training embedded in professional programs.
PIVOT curriculum, institutional coaching, advanced evaluation, sustainability planning, Fellow support.
Committed
Campus is embedding practice into how the institution operates.
A new campus may be entirely at Level 1. A mature partner operates at all four simultaneously. The partnership track describes strategic focus, not a judgment of quality.
What is hard right now

Format, not just motivation. Research on dialogue pedagogy finds that sessions shorter than three hours force trade-offs between relationship-building, skill-building, and dialogue. Tension surfaces just as abbreviated sessions end, with no time to work through it. When students disengage from training, the answer is often not better recruitment but better format.

Conditions, not just commitment. The political climate makes gathering feel risky. Institutional actions during campus conflicts can restructure the power landscape and close pathways to dialogue for extended periods. International students may feel unsafe participating.

Rational choices. Students are not flaky. When dialogue does not seem like it will produce results, or when institutional actions have undermined trust, students choose other forms of engagement. SDCN has learned that when students disengage, the first question is whether the conditions are right, not whether the students are committed enough.

From Intake to Measurement

The intake conversation applies the Concept of Relationship to a specific campus. It identifies a partnership track, which determines the levels of focus, which determines what gets measured. This is a single pipeline, not separate processes.

Intake-to-Measurement Pipeline

Intake Conversation

MI-based questions surface readiness, barriers, and the relational field

Who holds power? What perceptions shape willingness? What patterns of interaction exist?

Partnership Track

Foundational, Emerging, or Committed

Based on where the campus is and what is realistic right now

Levels of Focus

Which of the four levels the campus will work at this year

Foundational: Levels 1-2. Emerging: Level 3. Committed: Level 4.

Priority Measures

2-3 measures proportional to depth

Track determines defaults. Campus adds 1-2 optional measures matched to emphasis.
Every intake conversation produces a one-page partnership summary naming the track, interventions, and measures. SDCN handles complex evaluation centrally.

How SDI Measures Whether Civic Agency Is Developing

SDI's evaluation framework has two parts: validated survey instruments and operational tracking for what surveys cannot capture.

Civic Agency Measure (Anchor)

A 16-item scale measuring competence, drive, individual power, and collective power (Wegemer et al., 2025). Administered pre/post and at 6 to 12 month follow-up. Published under CC BY-NC 4.0.

SCIM Bridging Questionnaire (Proximal Indicators)

SCIM BQ items capture attitudinal shifts that indicate civic agency is developing. They are the "something is shifting" measures. The Civic Agency Measure is the "here is what it adds up to" measure.

MeasureItemsRoleWhat It Tracks
Intellectual Humility2CoreOpenness to being wrong
Respect/Understanding2CoreRelating to people you disagree with
Self-Efficacy: Bridge-Builder2CoreFeeling equipped; closest to civic agency competence
Intergroup Empathy2SupplementalCross-group understanding
Belonging2SupplementalCampus connection
Personal Agency2SupplementalAdjacent to individual power
Collaboration Norms2SupplementalWorking across difference
Pluralist Norms2SupplementalInstitutional embedding

What Surveys Cannot Capture

What Needs TrackingWhy It MattersMethod
Stage progressionGroups at early stages are unlikely to produce full civic agencyModerator logs and stage rubric
Action documentationJoint action at later stages is civic agency in practiceAction logs and case documentation
Whether trained people facilitateTraining without practice means competence did not developLongitudinal follow-up
Institutional adoptionWhether dialogue is embedded in structuresCase documentation and independence rubric
Willingness and felt safetyWhether people feel safe enough to show upCustom pre/post items
Retention and dosageWho stays and how session structure relates to outcomesModerator operational data

Assumptions and External Factors

External: Decreased funding for DEI-associated work. Increased polarization. A cultural moment where gathering itself is harder. Post-COVID disengagement persists. Institutional actions during campus conflicts restructure the relational field. Well-funded competitors have entered the space.

Organizational: Two full-time staff and a small contractor team. Grant funding will likely decrease. Campus partners often have accountability structures where the person managing the SDCN relationship does not supervise the person implementing.

Design implications: This framework centers a measurable outcome (civic agency) and helps each partner choose a focused set of goals. The Concept of Relationship provides the analytical lens at every level, from intake through institutional embedding.

References

Boyte, H. (2008). "Building Civic Agency: The Public-Work Approach." openDemocracy.

Diaz, A. & Perrault, R. (2010). "Sustained Dialogue and Civic Life: Post-College Impacts." Michigan Journal of Community Service Learning.

Flint, M., McLelland, L., & Kouaho, J.E. (2025). "'It Takes Time': Examining the Pedagogical Implications of Political Dialogue Facilitation." College Teaching. DOI: 10.1080/87567555.2025.2450720.

Saunders, H. (2005). Politics Is about Relationship: A Blueprint for the Citizens' Century.

Saunders, H. (2011). Sustained Dialogue in Conflicts: Transformation and Change.

Sustained Dialogue Institute. (2025). "Case Study: Attempts at Implementing Intergroup Dialogue in a Campus Setting after October 7, 2023." Prepared for the AAA-ICDR Foundation.

Svensson, I. & Brouneus, K. (2013). "Dialogue and interethnic trust: A randomized field trial of 'sustained dialogue' in Ethiopia." Journal of Peace Research.

Watts, R. & Flanagan, C. (2007). "Pushing the Envelope on Youth Civic Engagement." Journal of Community Psychology, 35(6), 779-792.

Wegemer, C.M., Wray-Lake, L., Hope, E.C., et al. (2025). "A Multidimensional Conceptualization and Measure of Youth Civic Agency." Journal of Community Psychology.

Appendix: Partnership Intake and Strategy Framework

How SDCN assesses where a campus is, decides what to focus on, and selects the right measures. The intake conversation is Stage 1 work: it applies the Concept of Relationship to a specific campus.

A. How the Intake Conversation Works

The intake is a conversation, not an application. SDCN uses a motivational interviewing approach: listen for readiness, barriers, and capacity. Do not sort prematurely. The same process works for new campuses and for annual reassessment of existing partners.

Core Intake Questions

"On a scale of 1 to 10, how important is it to your campus right now to go deeper than one-time programming?"

Follow-up: "What makes you say that number?" This elicits the partner's own reasoning. It tells SDCN more than the number itself.

"What would need to be true on your campus for a multi-meeting dialogue process to succeed?"

Surfaces structural conditions (staffing, supervision) and climate conditions (political safety, willingness to gather). The answer reveals which Concept of Relationship components are most active.

"What concerns come up when you think about committing to sustained, ongoing work?"

Concerns are data. "We do not have someone who can follow through" = supervision gap. "Students will not come" = Level 1 barrier. "The administration shut down a student org and trust is broken" = power dynamics reshaping the field.

"Walk me through what happened the last time your campus tried something like this."

Stories surface barriers faster than hypotheticals.

B. Three Partnership Tracks

The tracks describe a campus's partnership stage with SDCN. They are not a judgment of quality. A Foundational campus that successfully runs one orientation dialogue in a hostile environment has accomplished something significant.

Foundational · Levels 1 and 2

You are probably here if: Gathering itself is hard. No trained moderators yet. Student engagement is unpredictable.

Focus: Getting people into rooms and giving them an experience of dialogue that makes them want more.

Default measures: Participation counts, willingness and felt safety (custom items), SCIM BQ core (Intellectual Humility, Respect/Understanding).

Emerging · Level 3

You are probably here if: Some dialogue activity exists. Ready to build local people who can sustain it.

Focus: Developing local practitioners who can run dialogue and respond to conflict without SDCN present every time.

Default measures: SCIM BQ core + Self-Efficacy: Bridge-Builder, moderator competence tracking, group continuity, stage progression.

Committed · Level 4

You are probably here if: Existing capacity, thinking about sustainability. Dialogue has happened across multiple semesters.

Focus: Embedding dialogue into institutional structures so it survives graduation, turnover, and funding shifts.

Default measures: Civic Agency Measure (full 16-item), institutional adoption indicators, action documentation, SCIM supplemental (Pluralist Norms, Democratic Norms).

C. Selecting Priority Measures

The goal is focus, not comprehensiveness. The track narrows the menu. Beyond defaults, each campus adds one or two optional measures matched to emphasis:

If the campus emphasizes...Consider adding...
Cross-group understandingIntergroup Empathy, Affective Polarization
Community belongingBelonging (2 items)
Leadership developmentPersonal Agency (2 items)
Faculty or staff engagementValue Listening, Self-Efficacy: Bridging Community

Campus partners spend most of their energy running good programming. SDCN handles complex evaluation centrally: longitudinal follow-up, cross-campus comparisons, qualitative case documentation.

D. The Partnership Summary

Every intake conversation produces a one-page summary. This is the anchor document for the relationship.

SDCN Partnership Summary | 2026-2027

Campus
Midlands State University
Primary Contact
Jordan Lee, Associate Director of Student Life
Track
Emerging
Context
Midlands ran one staff-led dialogue group last year (six participants, one semester) with pre/post surveys showing movement on intellectual humility and respect. Jordan has been through facilitation training. Two students attended the Leaders Summit but did not follow through on moderating (one took a student government role, the other had a heavy course load). Supportive VP of Student Affairs but no dedicated budget line. Jordan does not supervise the GA who coordinates logistics.
Primary Levels of Work
Level 2 (sustaining dialogue groups) and Level 3 (building moderator capacity)
Priority Interventions
  • Send 2-3 new student leaders to the Leaders Summit with a pre-commitment plan for fall facilitation
  • Run one peer-led dialogue group in fall using the facilitation buddy tool
Measures
  • SCIM BQ core: Intellectual Humility, Respect/Understanding, Self-Efficacy: Bridge-Builder (pre/post)
  • Group continuity: did fall group sustain, did moderators facilitate at least three sessions
  • Stage progression: moderator log indicating whether groups moved beyond Stage 2
Key Conditions to Monitor
  • Jordan does not supervise the GA. SDCN will check in directly with Jordan monthly.
  • Last year's Summit attendees did not follow through. This year includes a structured re-entry plan with specific fall commitments before the Summit.
SDCN Will Provide
  • Monthly check-in calls with Jordan
  • Leaders Summit registration and pre-Summit planning
  • Facilitation buddy tool customized for fall topic
  • Mid-semester coaching call for student moderators
Next Check-in
October 2026