Navigating Complexity and Trust: Using Outcome Harvesting and AI-Assisted Analysis for Adaptive Evaluation
技能培训班 | Online
About the Event
This workshop is designed for participants who are new to Outcome Harvesting, as well as those who have some familiarity with evaluation in complex settings and want to explore how AI can support analysis and reflection in a responsible way. No prior technical knowledge of AI is required. The workshop presents Outcome Harvesting as a practical methodology for evaluating non-linear and emergent change. In addition, it introduces a focused discussion on the careful use of AI-assisted tools to support the analysis of harvested outcomes, for example through clustering, synthesis, and causal mapping. The event does not present AI as a replacement for evaluative judgment, but as a bounded support tool that may enhance reflection and sense-making in complex settings.
Competency Areas (Top 3):
● Complexity-aware evaluation design and methodology
● Adaptive management, reflection, and evidence use
● Responsible AI use in evaluation
What participants will learn:
Participants will learn when and how to use Outcome Harvesting in complex and adaptive interventions. They will gain a practical understanding of the main steps of Outcome Harvesting, including how to formulate outcome statements, identify evidence, substantiate findings, and use harvested outcomes for reflection, learning, and adaptive management. Participants will also reflect on how AI-supported tools may assist analysis and synthesis in Outcome Harvesting processes, while preserving methodological rigor, context sensitivity, and human judgment.
Background and Rationale
Glocal Evaluation Week 2026 asks how the evaluation profession should respond to AI while safeguarding trust, methodological rigor, and human judgment. It also highlights three useful lenses for this discussion: evidence needs, workflow enhancements, and capability development.
This workshop responds to that agenda through a focused exploration of Outcome Harvesting (OH) as a method for evaluating complex, adaptive, and non-linear interventions. Traditional results-based approaches often struggle in contexts where outcomes emerge over time, where multiple actors shape change, and where contribution matters more than simple attribution. OH offers a practical alternative by identifying observable changes first and then working backwards to understand how an intervention contributed to them.
The workshop also explores where AI may be useful in OH processes, especially in organizing qualitative evidence, clustering outcome statements, supporting pattern recognition, and aiding synthesis across large volumes of harvested data. At the same time, it addresses the risks Glocal raises so clearly: where apparent efficiencies may compromise rigor, where context may be lost, and why evaluative judgment and stakeholder validation remain essential.
This makes Outcome Harvesting a strong fit for Glocal 2026 because it sits precisely at the intersection of complexity, learning, evidence use, and careful human interpretation.
Objectives
By the end of the workshop, participants will be able to:
● Understand the core principles, steps, and applications of Outcome Harvesting.
● Recognize when Outcome Harvesting is more appropriate than conventional results-based approaches in complex or adaptive interventions.
● Design and implement a basic Outcome Harvesting process to identify, formulate, and substantiate meaningful change.
● Explore selected use cases where AI can support Outcome Harvesting analysis, synthesis, and reflection.
● Identify ethical and methodological risks of using AI in qualitative and complexity-aware evaluation.
● Apply practical safeguards to ensure AI supports, rather than undermines, trust, rigor, and stakeholder voice.
Session structure:
Session 1: Evaluation, Complexity, and Trust in the Age of AI Introduces the Glocal 2026 theme, complexity in evaluation, and the limits of conventional linear approaches 45 min
Session 2: Introduction to Outcome Harvesting Core concepts, principles, and the six iterative steps of OH 60 min
Session 3: Designing and Implementing Outcome Harvesting Framing harvest questions, formulating outcome statements, substantiation, and use 60 min
Session 4: AI-Assisted Analysis for Outcome Harvesting Practical uses of AI in clustering, pattern recognition, causal mapping, and synthesis, with discussion of risks and safeguards 60 min
Session 5: Applied Group Exercise Small-group work to develop sample outcome statements, review evidence, and reflect on where AI could or could not add value 75 min
Session 6: Reflection and Action Planning Using OH findings for adaptive management, learning, and responsible evidence use 30 min
Learning methods
● Interactive presentations with practical examples
● Facilitated group discussion
● Small-group exercises
● Case-based application of Outcome Harvesting
● Guided reflection on AI use, risks, and safeguards
● Template-based action planning
Materials needed
● Presentation slides on Outcome Harvesting
● OH templates for outcome statements and substantiation
● Case study handout
● Reflection worksheet on AI-supported analysis
● End-of-session feedback form
Competency Areas (Top 3):
● Complexity-aware evaluation design and methodology
● Adaptive management, reflection, and evidence use
● Responsible AI use in evaluation
What participants will learn:
Participants will learn when and how to use Outcome Harvesting in complex and adaptive interventions. They will gain a practical understanding of the main steps of Outcome Harvesting, including how to formulate outcome statements, identify evidence, substantiate findings, and use harvested outcomes for reflection, learning, and adaptive management. Participants will also reflect on how AI-supported tools may assist analysis and synthesis in Outcome Harvesting processes, while preserving methodological rigor, context sensitivity, and human judgment.
Background and Rationale
Glocal Evaluation Week 2026 asks how the evaluation profession should respond to AI while safeguarding trust, methodological rigor, and human judgment. It also highlights three useful lenses for this discussion: evidence needs, workflow enhancements, and capability development.
This workshop responds to that agenda through a focused exploration of Outcome Harvesting (OH) as a method for evaluating complex, adaptive, and non-linear interventions. Traditional results-based approaches often struggle in contexts where outcomes emerge over time, where multiple actors shape change, and where contribution matters more than simple attribution. OH offers a practical alternative by identifying observable changes first and then working backwards to understand how an intervention contributed to them.
The workshop also explores where AI may be useful in OH processes, especially in organizing qualitative evidence, clustering outcome statements, supporting pattern recognition, and aiding synthesis across large volumes of harvested data. At the same time, it addresses the risks Glocal raises so clearly: where apparent efficiencies may compromise rigor, where context may be lost, and why evaluative judgment and stakeholder validation remain essential.
This makes Outcome Harvesting a strong fit for Glocal 2026 because it sits precisely at the intersection of complexity, learning, evidence use, and careful human interpretation.
Objectives
By the end of the workshop, participants will be able to:
● Understand the core principles, steps, and applications of Outcome Harvesting.
● Recognize when Outcome Harvesting is more appropriate than conventional results-based approaches in complex or adaptive interventions.
● Design and implement a basic Outcome Harvesting process to identify, formulate, and substantiate meaningful change.
● Explore selected use cases where AI can support Outcome Harvesting analysis, synthesis, and reflection.
● Identify ethical and methodological risks of using AI in qualitative and complexity-aware evaluation.
● Apply practical safeguards to ensure AI supports, rather than undermines, trust, rigor, and stakeholder voice.
Session structure:
Session 1: Evaluation, Complexity, and Trust in the Age of AI Introduces the Glocal 2026 theme, complexity in evaluation, and the limits of conventional linear approaches 45 min
Session 2: Introduction to Outcome Harvesting Core concepts, principles, and the six iterative steps of OH 60 min
Session 3: Designing and Implementing Outcome Harvesting Framing harvest questions, formulating outcome statements, substantiation, and use 60 min
Session 4: AI-Assisted Analysis for Outcome Harvesting Practical uses of AI in clustering, pattern recognition, causal mapping, and synthesis, with discussion of risks and safeguards 60 min
Session 5: Applied Group Exercise Small-group work to develop sample outcome statements, review evidence, and reflect on where AI could or could not add value 75 min
Session 6: Reflection and Action Planning Using OH findings for adaptive management, learning, and responsible evidence use 30 min
Learning methods
● Interactive presentations with practical examples
● Facilitated group discussion
● Small-group exercises
● Case-based application of Outcome Harvesting
● Guided reflection on AI use, risks, and safeguards
● Template-based action planning
Materials needed
● Presentation slides on Outcome Harvesting
● OH templates for outcome statements and substantiation
● Case study handout
● Reflection worksheet on AI-supported analysis
● End-of-session feedback form
Speakers
| 名称 | 标题 | Biography |
|---|---|---|
| George Theuri | Mr. | George Theuri is an M&E and development practitioner with over a decade of experience working across Africa. He is currently working as an evaluation analyst for the CGIAR Evaluation Function. He has supported INGOs, government agencies, and grassroots organisations in integrating evidence-based adaptive management and learning processes into their programmes. George has led outcome harvesting for flagship entrepreneurship and youth development programmes and facilitated participatory evaluations using OH across sectors such as gender, livelihoods, peacebuilding, and health. |
| Cristina Repede | Ms. | Cristina Repede is a M&E and public policy consultant with over 13 years of international experience. She has applied OH in various evaluations, designed OM/OH materials for training, and contributed to advancing evaluation tools through her work on a handbook on impact assessment and evaluation methods under a Horizon 2020 research project. During her tenure at the European Commission, she supported Balkan states national authorities in building M&E systems, with a focus on strengthening institutional capacity for evidence-based policymaking. Cristina holds a Master’s in M&E from Saarland University and brings a reflective, systems-oriented lens to capacity building. |