What is the difference between a theory of change and a logframe?

A theory of change explains how and why change happens: the full causal pathway with assumptions, feedback loops and context. A logframe is a structured 4×4 summary (goal, outcomes, outputs, activities against indicators, means of verification and assumptions) used for planning and reporting. In practice you develop the theory of change first, then derive the logframe from it.

How do I build a theory of change for an NGO programme?

Start from the long-term change you want, work backwards through preconditions (outcomes), then map what the programme actually does (activities and outputs) to those outcomes. Make assumptions explicit at every link, identify which are evidence-backed versus untested, and revisit the map as evidence arrives. The Theory of Change Studio guides this sequence and exports the result.

What makes a good assumption in a theory of change?

A good assumption is specific, testable, and load-bearing: if it fails, the causal link fails. Vague assumptions like 'community participates' hide risk; a sharper version states who must do what, under what conditions, and what evidence exists that they will.

Back to MEL Course

Browse Membership

Free Studio

Theory of Change Studio

Build a structured Theory of Change step by step. Define your causal pathway from problem to impact, complete with assumptions and indicators. Step 7 also has a free canvas with sector templates and 203 behaviour change techniques, and the guide below carries worked examples, problem sets and sector guidance. Your work saves on this device, and the LogFrame Builder can import it.

i Start by clearly defining the problem your programme or intervention addresses. A well-defined problem grounds the entire Theory of Change. Be specific about who is affected, where, and why existing conditions persist.

Problem Statement

Describe the core problem or need your intervention addresses.

Target Population

Who are the primary beneficiaries or affected communities?

Geographic Context

Where does the intervention take place?

i Your long-term goal is the ultimate impact you aim to achieve. It should be aspirational but realistic, and clearly linked to the problem statement. Think 5-10 years ahead.

Long-term Goal / Impact Statement

What is the ultimate change you want to see? This sits at the top of your Theory of Change.

i Outcomes are the intermediate changes (in behaviour, knowledge, practice, or conditions) that must occur for your long-term goal to be achieved. Aim for 3-5 outcomes. Order them from most immediate to most distal.

i Outputs are the tangible products, services, or activities your programme delivers. Each output should logically contribute to at least one outcome. Link each output to its parent outcome.

i Assumptions are the conditions that must hold true for each causal link to work. Good assumptions are specific, testable, and linked to a particular step in the chain. Ask: "What else needs to be true for this output to lead to this outcome?"

! Common mistake: listing assumptions that are too vague (e.g., "political will exists"). Be specific: "District health officers allocate at least 20% of budget to community health workers during the project period."

i Indicators define how you will measure whether each outcome has been achieved. Use SMART criteria: Specific, Measurable, Achievable, Relevant, Time-bound. Each outcome should have at least one indicator.

i Good indicator: "Percentage of target households with a dietary diversity score of 5+ (out of 10 food groups) by end of Year 3." Bad indicator: "Food security improves."

Goal Outcomes Outputs Assumptions Indicators

Your Theory of Change diagram will appear here as you fill in the steps above.

Understanding Theory of Change

What is a Theory of Change?

A Theory of Change (ToC) is a comprehensive description and illustration of how and why a desired change is expected to happen in a particular context. It maps the causal logic from activities and outputs through intermediate outcomes to long-term impact.

Unlike a logframe, a ToC makes explicit the assumptions underlying each causal step, enabling more rigorous programme design and evaluation.

Why Build a ToC?

A well-articulated ToC helps teams align on programme logic, identify gaps in reasoning, design meaningful M&E frameworks, communicate with stakeholders and donors, and adapt programming when assumptions prove wrong.

Writing Good Outcomes

Outcomes should describe changes in people's lives, behaviours, or conditions. What the programme delivers is an output. Use active language: "Farmers adopt drought-resistant crop varieties" rather than "Drought-resistant seeds distributed."

Good: "Women in target communities report increased decision-making power over household finances." Bad: "Gender training conducted."

Crafting Testable Assumptions

Assumptions should be specific enough to monitor. Instead of "communities participate," try "At least 60% of invited households send a representative to monthly farmer field school sessions." This lets you track whether your theory holds during implementation.

SMART Indicators

Each indicator should be Specific (precisely defined), Measurable (quantifiable or verifiable), Achievable (realistically collectible), Relevant (connected to the outcome), and Time-bound (with a target date or frequency).

Example: "% of target households consuming 3+ meals per day, measured quarterly via household survey, target: 75% by Month 36."

Common ToC Mistakes

Watch out for these pitfalls: confusing outputs with outcomes, missing causal links (logical gaps), unstated assumptions, indicators that measure activity not change, overly linear thinking (ignoring feedback loops), and building the ToC after the programme is designed rather than before.

Rules of Thumb

1 Start with the long-term goal and work backwards: what must change for this to happen?
2 Aim for 3-5 outcomes. Fewer than 3 usually means oversimplification; more than 5 becomes hard to manage and evaluate.
3 Every outcome should have at least one indicator and one assumption. If you cannot measure it, you cannot know whether your theory holds.
4 Assumptions are not risks. An assumption is something you believe to be true; a risk is something you fear may go wrong. Both matter, but they serve different roles in a ToC.
5 Revisit and revise your ToC regularly. It is a living document, not a one-time exercise.

Continue Learning

MEL Foundations Course Sampling Design Studio

Theory of Change Guide

Foundations, worked examples, behaviour change techniques, problem sets and sector guidance. Open any section while you build. For a second pair of eyes, book a ToC coaching call, join a ToC Design Dojo, or request a workshop for your team. The full list of techniques, with South Asian examples, is in the BCT Repository.

Foundations

Core concepts for building rigorous Theories of Change

What is a Theory of Change?

A Theory of Change (ToC) is a comprehensive description and illustration of how and why a desired change is expected to happen in a particular context. It maps the causal logic from inputs through activities, outputs, outcomes, to long-term impact, and makes explicit the assumptions and evidence underlying each causal link.

Unlike a simple logical framework, a ToC explains why each step leads to the next, grounds causal claims in evidence, and surfaces the assumptions that must hold for the program to succeed.

Results chain

The standard causal pathway:

Input → Activity → Output → Outcome → Impact

Each link requires a causal mechanism (why does this lead to that?) and evidence (what tells us this mechanism works?).

Key components

Long-term change (impact statement)
Preconditions and pathways of change
Causal linkages with evidence
Assumptions at each level
Indicators for each outcome
Interventions mapped to the pathway
Context and stakeholder analysis

ToC vs. logframe

Dimension	Theory of Change	Logical Framework
Focus	Why and how change happens	What will be delivered and measured
Structure	Non-linear, multiple pathways	Linear hierarchy (4 levels)
Assumptions	Central, tested explicitly	Listed but often untested
Evidence	Supports each causal link	Indicators verify achievement
Adaptability	Living document, updated regularly	Usually fixed at design stage
Best for	Complex, systemic programs	Straightforward, linear programs

Common pitfalls in ToC design

Confusion of outputs with outcomes: Counting workshops conducted (output) vs. actual skill adoption (outcome)
Missing assumptions: Failing to articulate what must hold true for each causal link
Overly linear thinking: Ignoring feedback loops, interaction effects, and unintended consequences
No evidence grounding: Causal claims based on hope rather than research evidence
Ignoring context: Transplanting a ToC from one setting without adaptation to local realities

Behavior change techniques for ToC

Integrate evidence-based BCTs from the 203-technique taxonomy into your causal pathways. Drag BCTs onto your nodes in the free canvas →

Why BCTs matter for Theory of Change

Behavior change is often the critical "black box" between program activities and outcomes. Specifying which BCTs you will deploy, and the evidence for their effectiveness, strengthens your causal logic and makes your ToC testable. The BCT Taxonomy v1 (Michie et al., 2013) provides a standardized vocabulary of 93 techniques, extended to 203 in the ImpactMojo repository with South Asian contextualizations.

Below are key BCT categories most relevant to ToC design. Browse all 203 techniques in the technique panel of the free canvas.

Techniques for setting goals, action planning, and problem-solving. Critical for ToC activity-to-output links.

Technique	Example in ToC	Evidence
Goal setting (behavior)	SHG members set weekly savings targets	Strong
Action planning	Farmers create step-by-step plans for adopting new seeds	Strong
Problem solving	CHWs identify barriers to ANC attendance with mothers	Strong
Review behavior goals	Monthly SHG review of savings and loan targets	Moderate
Behavioral contract	Signed commitment to latrine construction by village leaders	Moderate

Using social dynamics for behavior change, which matters a great deal in South Asian community-based programs.

Technique	Example in ToC	Evidence
Social support (practical)	Peer support groups for new mothers practicing exclusive breastfeeding	Strong
Demonstration of behavior	Lead farmer demonstrates improved cultivation techniques	Strong
Social comparison	Village scorecard comparing ODF progress across hamlets	Moderate
Information from credible source	ASHA worker sharing health information during home visits	Strong
Restructuring social environment	Forming mothers' groups to shift norms around child nutrition	Moderate

Essential for the output-to-outcome link: how do participants know they're making progress?

Technique	Example in ToC	Evidence
Self-monitoring of behavior	Growth monitoring cards tracked by mothers	Strong
Feedback on behavior	CHW gives feedback on handwashing practice during home visits	Strong
Self-monitoring of outcome	Farmers track crop yield improvements season over season	Moderate
Biofeedback	Water quality testing kits showing contamination levels	Moderate

Reinforcement mechanisms that sustain changed behavior over time.

Technique	Example in ToC	Evidence
Social reward	Public recognition of ODF villages at gram sabha	Strong
Material reward (behavior)	Token incentives for completing ANC visits	Moderate
Non-specific reward	Certificate of completion for SHG financial literacy training	Moderate
Self-reward	Encouraging mothers to celebrate immunization milestones	Moderate

Changing the context to make desired behaviors easier, through structural interventions in the ToC.

Technique	Example in ToC	Evidence
Adding objects to environment	Installing handwashing stations at school entrances	Strong
Prompts/cues	Reminder stickers on water storage containers for treatment	Strong
Restructuring physical environment	Improved cookstoves placed in existing kitchen areas	Moderate
Avoidance/reducing exposure to cues	Removing open defecation spots via CLTS triggering	Moderate

Browse all 203 BCTs in the free canvas →

Worked examples

Full Theory of Change case studies with BCT annotations and indicator mapping

Case study 1: Self-help groups for women's empowerment, Bihar

A livelihoods program forming SHGs in rural Bihar to improve household income and women's decision-making power.

Impact statement

Women in target communities exercise greater economic agency and household decision-making power, contributing to poverty reduction (SDG 1) and gender equality (SDG 5).

Causal pathway

Level	Statement	BCTs Applied	Indicator
Input	Trained community mobilizers and seed capital for SHG formation	—	# mobilizers trained; seed capital disbursed
Activity	Weekly SHG meetings with savings, credit, and financial literacy	Goal setting Action planning Social support (practical)	# meetings held; attendance rate
Output	200 SHGs formed with 2,400 members; ₹12L cumulative savings	Self-monitoring of behavior Social comparison	# SHGs formed; total savings; repayment rate
Outcome	Increased household income; women make financial decisions	Behavioral contract Social reward	% income increase; women's decision-making index
Impact	Reduced poverty; improved gender equity in target villages	—	Poverty headcount ratio; GDI change

Key assumptions

Women have time and social permission to attend weekly meetings
Local money lenders don't actively undermine SHG credit operations
Bank linkage (SHG-Bank Linkage Programme) is accessible and functional
Male family members are supportive or at minimum non-obstructive

BCT design note

The Social support (practical) BCT is the backbone of SHG models: peer savings pressure, collective credit decisions, and social accountability. Combined with Goal setting (weekly savings targets) and Self-monitoring (passbook records), these three BCTs create a powerful feedback loop. Evidence from NABARD evaluations shows this combination yields 23% higher repayment rates than credit-only programs.

Case study 2: Digital education for foundational literacy, Jharkhand

A technology-enabled education program using tablets and adaptive learning software to improve reading outcomes in government primary schools.

Impact statement

Children in Grades 1-3 in target schools achieve grade-level foundational literacy and numeracy (FLN), contributing to equitable quality education (SDG 4).

Causal pathway

Level	Statement	BCTs Applied	Indicator
Input	Tablets with adaptive learning software; teacher training on blended pedagogy	—	# tablets procured; # teachers trained
Activity	Daily 45-min tablet sessions with adaptive content; weekly teacher review of dashboards	Graded tasks Feedback on behavior Prompts/cues	Avg. minutes/student/day; teacher dashboard logins
Output	5,000 students completing 80%+ of learning modules	Self-monitoring of outcome Non-specific reward	Module completion rate; quiz pass rates
Outcome	Students read at grade level; improved numeracy scores	Demonstration of behavior	ASER reading level; NAS numeracy scores
Impact	Grade-level FLN achievement in target schools	—	% students at grade-level FLN (endline vs. baseline)

Key assumptions

Electricity and charging infrastructure available in schools
Teachers have sufficient digital literacy and motivation
Children attend school regularly enough for dosage effect
Adaptive software content is culturally and linguistically appropriate

BCT design note

The Graded tasks BCT (adaptive difficulty) is the core mechanism of edtech interventions. It keeps learners in their zone of proximal development. Combined with Feedback on behavior (immediate quiz feedback) and Self-monitoring of outcome (progress dashboards), these BCTs create a self-reinforcing learning loop. RCT evidence from Mindspark (Muralidharan et al., 2019) shows 0.37σ improvement in math when this BCT combination is deployed.

Indicator development

How to build measurable, meaningful indicators for each level of your ToC

SMART indicator framework

Criterion	Question	Common Pitfall
Specific	Does it clearly describe what is being measured?	Vague indicators like "improved awareness"
Measurable	Can it be quantified or verified objectively?	Subjective assessments without validated scales
Achievable	Is the target realistic given resources and timeline?	Aspirational targets with no basis in evidence
Relevant	Does it actually measure the stated outcome?	Proxy indicators too distant from actual change
Time-bound	By when should the target be achieved?	No timeline specified for target achievement

Indicator mapping by ToC level

ToC Level	Indicator Type	Data Source	Frequency
Input	Resource tracking (budget spent, staff deployed)	Financial MIS, HR records	Monthly
Activity	Process monitoring (sessions conducted, reach)	Activity registers, digital tracking	Weekly/Monthly
Output	Immediate deliverables (# trained, # built)	Program MIS, registers	Quarterly
Outcome	Behavior/practice change (adoption rates, knowledge scores)	Surveys, assessments, observations	Semi-annual
Impact	Long-term population-level change	NFHS, Census, panel surveys	Endline/2-3 years

Exercise: Write your indicators

For each outcome in your ToC, complete this template:

Outcome: [Your outcome statement]

Indicator: [What exactly will you measure?]

Baseline: [Current value] → Target: [Expected value by when]

Data source: [Survey / MIS / admin data / observation]

Collection frequency: [Monthly / Quarterly / Annual]

Disaggregation: [Gender / age / caste / geography / disability]

Problem sets

Practice building and critiquing Theories of Change. Try them in the steps above, or in the free canvas.

Problem 1: Maternal health, Uttarakhand

A mountain-district health program aims to increase institutional deliveries from 35% to 70%. Design a ToC identifying at least 3 BCTs that address supply-side and demand-side barriers. Consider terrain, distance to facilities, and cultural practices around childbirth.

Solved example available

Problem 2: Farmer producer organizations, Maharashtra

An FPO program wants to improve smallholder farmers' price realization by 25%. Build a ToC that maps from FPO formation through collective bargaining to market linkages. Include assumptions about market access and social capital.

Solved example available

Problem 3: Adolescent SRH, Rajasthan

Design a ToC for an adolescent sexual and reproductive health program in a conservative setting. Identify BCTs for behavior change, address social norm barriers, and specify how you'd measure outcomes while maintaining ethical safeguards.

Unsolved: try it yourself

Problem 4: Urban WASH, informal settlements, Bengaluru

Build a ToC for improving sanitation in urban informal settlements. Consider tenure insecurity, shared facilities, municipal coordination, and community-level collective action.

Unsolved: try it yourself

Problem 5: Digital financial inclusion, North-East India

A fintech-NGO partnership aims to increase digital payment adoption among rural women. Design a ToC that addresses digital literacy, trust barriers, and infrastructure constraints. Identify which BCTs can overcome initial resistance to technology adoption.

Unsolved: try it yourself

Problem 6: Climate-resilient agriculture, Odisha

Develop a ToC for promoting climate-smart agricultural practices among marginal farmers in cyclone-prone coastal Odisha. Map the pathway from training through adoption to resilience outcomes. Include environmental and gender cross-cutting considerations.

Unsolved: try it yourself

Problem 7: School dropout prevention, Telangana

A state government program aims to reduce secondary school dropouts by 40%, especially among SC/ST girls. Build a ToC spanning community engagement, school-level interventions, and social protection. Specify 4+ BCTs and identify the weakest assumption in your chain.

Solved example available

Problem 8: Mental health in conflict-affected areas, J&K

Design a ToC for a community mental health program in a conflict-affected region. Address stigma, limited clinical workforce, cultural idioms of distress, and the Do No Harm framework. Specify BCTs for help-seeking behavior change.

Unsolved: try it yourself

Build your solution in the free canvas →

Sector guidance

Sector-specific tips, frameworks and BCTs for Theory of Change design

Gender & women's economic empowerment

Apply a gender lens throughout the causal chain, not just as a cross-cutting afterthought.

Conduct gender analysis before ToC design
Map gendered barriers at each pathway level
Include women's agency as an intermediate outcome
Use GESI-sensitive indicators (decision-making, mobility, asset ownership)
Apply BCTs: Social support Restructuring social environment

Health & nutrition

Health ToCs often need both supply-side (service delivery) and demand-side (behavior change) pathways.

Map the health system pathway: facility → CHW → household
Specify BCTs for each behavior change objective
Include service quality as an intermediate outcome
Use WHO/UNICEF standard indicators where possible
Apply BCTs: Information from credible source Prompts/cues

Education & FLN

Education ToCs must distinguish between access, enrollment, attendance, and learning outcomes.

Don't conflate enrollment with learning
Map teacher behavior change as a separate pathway
Include household demand-side factors
Use ASER/NAS-aligned measurement frameworks
Apply BCTs: Graded tasks Feedback on behavior

Livelihoods & market systems

Market systems ToCs require understanding of supply chains, value addition, and systemic change.

Map the market system (input supply → production → aggregation → market)
Distinguish individual vs. systemic change outcomes
Include sustainability beyond project period
Address power dynamics in value chains
Apply BCTs: Goal setting Action planning

WASH & environmental health

WASH ToCs combine infrastructure (hardware) with behavior change (software) pathways.

Map hardware (infrastructure) and software (behavior) separately
Include O&M sustainability as a key outcome
Use CLTS community-level indicators alongside household
Address equity in access (women, disabled, elderly)
Apply BCTs: Adding objects to environment Social comparison

Climate & resilience

Climate ToCs must account for uncertainty, adaptive pathways, and systems-level change.

Include climate scenarios in assumptions
Map both mitigation and adaptation pathways
Use adaptive management as a design principle
Include ecosystem health as an outcome
Apply BCTs: Problem solving Information about consequences

Governance & social accountability

Governance ToCs must account for the political economy and the distribution of power.

Conduct political economy analysis before design
Map both citizen engagement and state responsiveness pathways
Include accountability mechanisms as outcomes, not just activities
Address sustainability of governance reforms
Apply BCTs: Social support (practical) Restructuring social environment

SRHR & adolescent health

Sensitive programs requiring careful ethical design and social norm navigation.

Apply Do No Harm throughout the ToC
Map social norms as both barriers and levers
Include privacy and confidentiality in service delivery pathway
Distinguish knowledge, attitudes, and practices as separate outcomes
Apply BCTs: Social support Information from credible source

Cross-cutting considerations

Lenses to apply across your entire Theory of Change

Gender equality & social inclusion (GESI)

Apply at every ToC level, not just as a tag-on.

Gender analysis of barriers and enablers at each level
Intersectional lens (gender × caste × class × disability)
GESI-specific outcomes and indicators
Meaningful participation in design and monitoring
Women's agency as a pathway, not just a target

Climate & environmental sustainability

Screen your ToC for climate risks and environmental co-benefits.

Climate risk screening for each assumption
Environmental co-benefits and risks mapped
Adaptation pathways for climate-sensitive outcomes
Carbon footprint of program activities
Green transition as a cross-cutting outcome

Do no harm & ethics

Ensure your program doesn't inadvertently cause harm.

Conflict sensitivity analysis at design stage
Power dynamics and unintended consequences mapped
Safeguarding policies for vulnerable populations
Informed consent protocols for data collection
Grievance redress mechanisms in place

Digital & innovation

Use technology thoughtfully. It is no silver bullet.

Digital readiness assessment before tech-based interventions
Data protection and privacy safeguards
Digital divide considerations (gender, age, location)
Human-centered design for technology adoption
Sustainability of digital infrastructure post-project

Measurement design

Designing rigorous evaluation strategies that test your ToC

Evaluation methods mapped to ToC

Method	Best For	Limitation	ToC Application
RCT	Causal attribution of impact	Expensive; ethical constraints; external validity	Testing the overall impact pathway
Difference-in-Differences	Quasi-experimental comparison	Parallel trends assumption	Evaluating outcome-level change
Most Significant Change	Understanding qualitative pathways	Subjective; not generalizable	Unpacking HOW change happened
Contribution Analysis	Complex, multi-actor programs	Weaker causal claims	Verifying contribution, not attribution
Outcome Harvesting	Emergent, non-linear change	Resource-intensive	Identifying unexpected outcomes and pathways
Process Tracing	Tracing causal mechanisms	Case-specific; limited generalizability	Testing specific causal links in the ToC

Data collection toolkit

Quantitative

Household surveys
Facility assessments
Administrative data (MIS)
Digital trace data

Qualitative

Key Informant Interviews
Focus Group Discussions
Case studies
Participatory methods (PRA)

Mixed

Sequential explanatory
Concurrent triangulation
Embedded design
Participatory statistics

Adaptive Theory of Change

Your ToC is a living document. This section shows how to keep it current

When to update your ToC

Monitoring data shows an assumption isn't holding
Midterm review reveals unexpected pathways or barriers
Context shifts: policy change, COVID, climate event, conflict
New evidence: published research changes your causal logic
Stakeholder feedback: communities identify missing pathways
Annually: as a minimum, conduct annual ToC review

Adaptive management cycle

Plan (ToC) → Do (Implement) → Study (Monitor & Learn) → Act (Adapt ToC) → Repeat

Each cycle should revisit and potentially revise: causal assumptions, BCT selection, indicator targets, and implementation strategy. Redraw your ToC in the free canvas each time it changes, and download a PNG of each version.