This web page is a concept developed by Chris Lysy of freshspectrum.com for educational purposes.  You can see the original CDC Framework by following this link.

The CDC Program Evaluation Framework, 2024

A practical approach to understanding what works, improving your programs, and making evidence-based decisions

Introduction

The CDC’s Program Evaluation Framework is designed to guide public health professionals in conducting program evaluation. It is a practical tool that summarizes and organizes essential elements of program evaluation. This guide will walk you through the framework, piece by piece.

CDC Evaluation Framework, 2024

The framework consists of:

  • Three Cross-Cutting Actions (the outer circle)
    • Advancing equity
    • Engage collaboratively
    • Learn from and use insights
  • Five Evaluation Standards (the center circle)
    • Relevance and utility
    • Rigor
    • Independence and objectivity
    • Transparency
    • Ethics
  • Six Evaluation Steps (the middle circle)
    • Assess context
    • Describe the program
    • Focus the evaluation questions and design
    • Gather credible evidence
    • Generate and support conclusions
    • Act on findings

Why We Evaluate

Whether you’re running a health program, managing a community initiative, or leading organizational change, you need to know if your work is making a difference. That’s where program evaluation comes in—not as an intimidating academic exercise, but as a practical tool for continuous improvement.

Think of program evaluation as a conversation between what you hope your program does and what it actually does. It’s systematic, yes—but it’s also deeply practical. While research aims to create generalizable knowledge, evaluation aims to help you make better decisions about your specific program, right now.

Good evaluation answers the questions that keep program managers up at night:

  • Are we reaching the right people?
    Understanding who benefits from your work
  • Are we doing what we said we’d do?
    Checking if implementation matches your plans
  • Is it working?
    Seeing if you’re achieving your intended outcomes
  • How can we do better?
    Finding concrete ways to improve
  • What unexpected effects are happening?
    Discovering both positive and negative unintended consequences

 

Evaluation is not the same as research, monitoring, or surveillance

Research

builds generalizable knowledge that applies broadly

Surveillance

tracks trends and patterns in populations over time

Monitoring

checks if you're hitting predetermined targets and goals

Evaluation

asks deeper questions about how and why your program works (or doesn't) in your specific context

Three Cross-cutting Actions

Before we dive into the step-by-step process, let’s talk about how you approach evaluation. Think of these three actions as the values that guide everything else. They’re not separate steps—they’re woven throughout the entire process.

Advance Equity

Ask yourself: Who benefits from our program? Whose voices are we hearing? What barriers exist? Build equity into every step by actively including marginalized communities and examining how your program addresses (or fails to address) health inequities.

Why this matters: Programs don’t exist in a vacuum. Historical and current inequities shape who accesses services, whose outcomes improve, and who gets heard in evaluation. If you’re not intentionally advancing equity, you might be perpetuating harm.

In practice: Include people from affected communities in your evaluation team. Use culturally responsive methods. Ask questions about equity at every step. Examine your own biases and how they might influence what you see.

Engage Collaboratively

Don’t do evaluation to people—do it with them. Bring in diverse voices from the start: program participants, staff, funders, community members, and yes, even skeptics. Multiple perspectives make your evaluation stronger and more credible.

Why this matters: When people feel ownership of the evaluation, they’re more likely to trust the findings and act on them. Plus, you’ll catch blind spots and assumptions you might have missed on your own.

In practice: Set up regular check-ins with your evaluation team. Create spaces where everyone feels comfortable sharing honest feedback. Value lived experience alongside professional expertise.

Learn and Use Insights​

Don’t wait until the end to share what you’re learning. Build a culture where insights flow regularly, mistakes are learning opportunities, and findings actually change how you work. The best evaluation happens when learning becomes continuous.

Why this matters: Evaluation reports that sit on shelves help no one. When you share insights early and often, you can course-correct before problems become crises. You also build evaluation capacity—people get better at asking good questions and using evidence.

In practice: Share interim findings throughout the process. Create regular “learning moments” with your team. Plan how you’ll use findings before you collect data. Make it easy for people to access and understand what you’ve learned.

Five Evaluation Standards

Not all evaluations are created equal. Here are the benchmarks that distinguish rigorous, useful evaluation from checkbox exercises. These standards help you make tough decisions when trade-offs arise (for example, when speed conflicts with thoroughness).

Relevance and Utility

Answer questions that actually matter to decision-makers. Your findings should be actionable, timely, and presented in ways people can understand and use.

Ask yourself: Will this information arrive in time to inform decisions? Is it specific enough to guide action? Can people actually do something with what we’re learning?

Rigor

Use solid methods that people can trust. This doesn’t mean overly complex—it means appropriate to your questions, well-executed, and honest about limitations.

Rigor comes from thoughtful planning, qualified evaluators, appropriate methods, and clear interpretation. A simple evaluation done well beats a complex evaluation done poorly.

Independence and Objectivity

Keep the evaluation free from political pressure or conflicts of interest. Evaluators should be fair, acknowledge their own biases, and let the evidence speak.

This doesn’t mean evaluators must be completely external—internal evaluators can be objective too. What matters is protecting the evaluation from undue influence and being transparent about potential conflicts.

Transparency

Document your methods, decisions, and findings clearly. Others should be able to understand how you arrived at your conclusions and reproduce your work.

Document everything before you start: What questions will you ask? How will you collect data? Who will see the findings? This prevents cherry-picking results or tailoring methods to get desired outcomes.

Ethics

Protect participants’ privacy and dignity. Ensure fairness. Consider cultural contexts. Uphold the highest ethical standards throughout.

Ethics means more than informed consent forms. It means respecting cultural protocols, ensuring equitable treatment, protecting vulnerable populations, and considering how evaluation itself might cause harm.

Six Evaluation Steps

Now for the practical part. These six steps provide a roadmap from start to finish. While they’re presented in order, real evaluation is rarely linear—you’ll often revisit earlier steps as you learn more.

Think of this as a cycle, not a checklist.

Assess the Context

Before you do anything else, understand your landscape.

Jumping straight into evaluation without understanding the context is like performing surgery in the dark. You might get lucky, but you’ll probably make things worse. This step sets you up for success by clarifying what’s possible and what matters.

Is Your Program Ready for Evaluation?

Start by asking: Is this program ready to be evaluated? You might need an evaluability assessment—a quick check to see if your program has:

  • Clear goals that people agree on
  • Reasonable expectations about what’s achievable
  • Sufficient resources to support evaluation
  • Enough time for outcomes to potentially occur


If these aren’t in place, you might need to do some groundwork first. That’s okay! It’s better to discover this now than after you’ve invested heavily in evaluation that can’t produce useful answers.

 

Identify your interest holders (we’ve moved away from calling them “stakeholders” to be more inclusive—the term has violent connotations for some Indigenous communities and implies a power hierarchy).

These include:

  • People served by the program: Those who directly or indirectly benefit from your work
  • Staff and implementers: Everyone from leadership to frontline workers to partners
  • Funders and decision-makers: Those who control resources or make program decisions
  • Community members: People affected by the program, even if they don’t directly participate
  • Program skeptics: People who oppose or question your program (their concerns matter!)


Don’t assume everyone wants the same things from the evaluation. Different groups will have different priorities, questions, and definitions of success. That’s valuable information.

 

Understand the place—not just the physical location, but the community’s history, strengths, power dynamics, and the factors driving health inequities.

Ask about:

  • Community history: What’s happened here before? Have previous programs built trust or broken it?
  • Power and privilege: How is power distributed? Whose voices typically get heard? Whose don’t?
  • Strengths and assets: What does this community do well? What expertise exists here?
  • Health inequities: What disparities exist? What’s driving them? How does your program relate to these root causes?


This understanding will shape everything from who you involve to how you interpret findings.

 

Assess your evaluation capacity at both organizational and individual levels:

Organizational capacity:

  • What resources (funding, staff, time) can support evaluation?
  • Does the organization value and use evaluation findings?
  • Are there systems for sharing and acting on insights?
  • Who are the evaluation champions who can advocate for this work?


Individual capacity:

  • How familiar are team members with evaluation?
  • What past experiences shape people’s attitudes toward evaluation?
  • What skills and knowledge already exist?
  • What training or support might people need?


Understanding capacity helps you meet people where they are and build skills along the way.

 

Describe the Program

Get crystal clear about what your program actually does.

You can’t evaluate something if you can’t describe it clearly. This step forces you to articulate your program’s logic—how you think your activities will lead to your desired outcomes. Often, this is when programs discover gaps or assumptions that need addressing.

Creating Your Program Logic Model

A logic model is a visual roadmap showing how your activities lead to outcomes. Don’t overthink it. This is a planning tool, not a work of art. Your logic model should show:

Inputs: What resources do you have?

  • Funding and budget
  • Staff and volunteers
  • Partners and collaborators
  • Existing data (like surveillance systems)
  • Evidence base from research or previous programs


Activities: What do you actually do?

  • Services delivered
  • Programs implemented
  • Policies developed
  • Communications launched
  • Training provided


Outputs: What gets produced immediately?

  • Number of people served
  • Sessions delivered
  • Materials distributed
  • Trainings completed


Outcomes: What changes do you expect?

  • Short-term (weeks to months): Changes in awareness, knowledge, attitudes, or skills
  • Intermediate (months to a few years): Changes in behavior, practice, or policy
  • Long-term (years): Changes in health status, systems, or population-level indicators

 

Let’s make this concrete:

Inputs: Funding from state health department, CDC staff, media partners, evidence on effective cessation messages

Activities:

  • Develop TV and social media ads showing real health consequences
  • Coordinate with healthcare providers on cessation resources
  • Launch telephone quitline
  • Train community health workers


Short-term outcomes:

  • Increased awareness of smoking’s health risks
  • More positive attitudes toward quitting
  • Increased knowledge of cessation resources


Intermediate outcomes:

  • More smokers attempt to quit
  • Increased use of quitline and cessation services
  • More healthcare providers discuss cessation with patients


Long-term outcomes:

  • Reduced smoking rates
  • Fewer smoking-related illnesses and deaths
  • Decreased healthcare costs

 

Your logic model should reflect your theory of change—the “why” and “how” of your program logic. Why do you think these activities will lead to these outcomes? What assumptions are you making?

For example: “We assume that showing real health consequences will motivate behavior change” or “We believe that increasing access to cessation resources will lead more people to quit successfully.”

Naming these assumptions is crucial. They’re testable through evaluation, and if they’re wrong, you’ll want to know early.

 

Your program description should also include:

Contextual factors: External things that might affect success

  • Policy environment
  • Economic conditions
  • Social determinants of health
  • Other programs operating in the same space


Stage of development: Where is your program in its lifecycle?

  • Planning: Activities are being designed but not yet implemented
  • Implementation: Activities are launching and being refined
  • Maintenance: Program is mature and stable


Your program’s stage matters enormously for what type of evaluation makes sense.

 

Your logic model isn’t set in stone. As you learn more—from evaluation or just from implementation experience—update it. Programs evolve. Your description should too.

Focus the Evaluation Questions and Design

Decide what you most need to learn.

You could evaluate a thousand things about your program. You can’t evaluate everything, and you shouldn’t try. This step is about making hard choices: What do we most need to know? What questions will actually inform decisions?

Crafting Focused Evaluation Questions

Work with interest holders to develop evaluation questions that are:

  • Clearly stated: Everyone understands what’s being asked
  • Answerable: You can actually collect data to respond to them
  • Aligned with your program’s stage: Don’t ask about long-term outcomes if your program launched six months ago
  • Feasible: You have the resources, time, and access to answer them


Actionable: The answers will inform actual decisions, not just satisfy curiosity.

 

Choose the right type of evaluation based on where your program is:

Formative Evaluation: Is our approach feasible before we launch?

  • When to use: Planning stage or when modifying an existing program
  • Questions it answers: Is this approach acceptable to our audience? Are our materials clear and culturally appropriate? What barriers might we face?


Process/Implementation Evaluation: Are we implementing as planned?

  • When to use: Early implementation or anytime you need to understand operations
  • Questions it answers: Are we reaching our intended audience? Is the program being delivered with quality and fidelity? What facilitators and barriers exist?


Outcome Evaluation: Are we achieving our intended outcomes?

  • When to use: After sufficient time for outcomes to occur (usually for more mature programs)
  • Questions it answers: Has knowledge, behavior, or policy changed? Are we seeing movement on our intended outcomes?


Impact Evaluation: Did our program cause the observed outcomes?

  • When to use: Mature programs with stable implementation and ability to compare to a control/comparison group
  • Questions it answers: What would have happened without this program? Can we attribute the changes we see to our specific intervention?


Economic Evaluation: Are we getting good value for our investment?

  • When to use: When cost-effectiveness matters for funding or replication decisions

Questions it answers: What’s the cost per outcome achieved? How do benefits compare to costs?

 

Process questions:

  • To what extent is the program reaching its intended priority populations?
  • What facilitators and barriers affect implementation?
  • How does implementation vary across different sites or communities?


Outcome questions:

  • To what extent did the program increase participants’ knowledge of health risks?
  • Did policy change following the advocacy campaign?
  • For whom is the program most and least effective, and why?


Impact questions:

  • What changes in health outcomes can be attributed to the program?
  • How much did the program contribute to observed changes in the community?

 

Your evaluation design is the overall structure—it determines how you’ll answer your questions with scientific rigor.

Three main design types:

Experimental designs: Randomly assign people or communities to receive the program or not

  • Strengths: Strongest evidence of causation, can confidently attribute results to your program
  • Limitations: Often not feasible, ethical, or appropriate; expensive and complex; may not capture real-world conditions
  • Example: Randomized controlled trial (RCT)


Quasi-experimental designs: Compare groups that weren’t randomly assigned

  • Strengths: More feasible than experiments while still providing evidence of program effects
  • Limitations: Can’t rule out all alternative explanations; requires careful analysis to control for differences
  • Example: Comparing communities with and without the program; time series showing trends before and after implementation


Observational designs: Describe what’s happening without a comparison group

  • Strengths: Feasible for most programs; great for understanding process and context
  • Limitations: Can’t establish causation; limited ability to attribute outcomes to the program


Example: Case studies, pre/post comparisons without a control group, surveys of participants

 

The “best” design isn’t the most rigorous—it’s the one that:

  • Answers your questions
  • Is feasible in your context
  • Can actually be implemented well
  • Produces findings your interest holders will trust and use


A simple evaluation done well beats a complex evaluation done poorly.

 

Gather Credible Evidence

Determine what data you’ll collect and how.

This is where evaluation gets real. You’re making concrete decisions about what counts as evidence, how you’ll collect it, and from whom. The key is being intentional—every choice should connect back to your evaluation questions.

Start with Expectations

Before you collect any data, collaborate with interest holders to establish expectations. What would success look like? What level of change would be meaningful?

For example:

  • “We expect at least 60% of participants to report increased knowledge”
  • “We hope to see a 20% increase in people accessing cessation services”
  • “We’ll consider it successful if policy changes in at least half of targeted jurisdictions”


Setting expectations in advance prevents moving the goalposts later and provides a clear benchmark for interpretation.

You have two broad categories of data:

Quantitative (numeric data):

  • Surveys with scaled responses
  • Administrative data (like service utilization records)
  • Health outcomes from medical records
  • Surveillance data
  • Cost data


Strengths: Can summarize large amounts of data, compare across groups, track trends over time, test statistical significance

Limitations: May miss nuance and context, can’t explain the “why” behind numbers, may not capture what matters most to participants.

Qualitative (narrative data):

  • In-depth interviews
  • Focus groups
  • Observations of program delivery
  • Document review
  • Open-ended survey responses

Strengths: Captures context and nuance, explains mechanisms and processes, centers participants’ perspectives, flexible and adaptive

Limitations: Time-intensive to collect and analyze, can’t easily quantify or generalize, smaller sample sizes

The power move: Use both (mixed methods)

Combining quantitative and qualitative data gives you the best of both worlds. Numbers tell you what and how much. Stories tell you why and how. Together, they create a complete picture.

 

Indicators are the specific, measurable things you’ll track. They bridge the gap between abstract concepts and concrete data.

For each part of your logic model, identify indicators:

Input indicators:

  • Amount of funding allocated
  • Number of staff with required qualifications
  • Number of partner organizations engaged


Activity/output indicators:

  • Number of people trained
  • Number of services delivered
  • Percentage of planned activities completed


Outcome indicators:

  • Percentage of participants reporting increased knowledge
  • Number of people who adopted the new behavior
  • Policy changes enacted

 

Evaluation question: To what extent did the smoking cessation campaign increase knowledge of health risks among smokers?

Concept: Knowledge of health risks

Indicator: Level of confirmed awareness of campaign messages about specific health conditions caused by smoking

Measures:

  • Percentage who correctly identify three or more smoking-related health conditions (pre and post campaign)
  • Percentage who recall specific campaign messages
  • Difference in knowledge scores between campaign-exposed and non-exposed groups


Data source: Phone surveys with random sample of adults in target communities

Expectation: At least 50% of those exposed to campaign will demonstrate increased awareness compared to baseline

 

Consider both primary data (new data you collect) and secondary data (existing data from other sources):

Primary data sources:

  • Surveys of participants, staff, or community members
  • Interviews with key informants
  • Focus groups with target populations
  • Direct observations of program implementation
  • Program records and tracking systems


Secondary data sources:

  • Surveillance systems (like disease registries)
  • Administrative data (like hospital discharge records)
  • Census and demographic data
  • Previously published research or evaluations

 

Pro tip: Before collecting new data, check if existing data can answer your questions. Secondary data is often cheaper and faster—but make sure it’s trustworthy and actually measures what you need.

 

How you collect data matters as much as what you collect. Culturally responsive data collection means:

  • Matching methods to context: Don’t use written surveys in oral-tradition communities
  • Using appropriate languages: Provide instruments in languages people speak, with professional translation
  • Respecting cultural protocols: Some communities have specific requirements for research and data collection
  • Involving community members: Train community members as data collectors when appropriate
  • Protecting privacy: Understand cultural norms around sensitive topics


Honoring data sovereignty: Respect communities’ rights to own and govern their data, especially for Indigenous communities

 

Data quality means your data are:

  • Reliable: Consistent and replicable
  • Valid: Actually measuring what you think they’re measuring
  • Accurate: Free from errors
  • Complete: No major gaps
  • Timely: Collected at the right time


Data quantity means having enough data to answer your questions confidently, without overburdening respondents or collectors.

The sweet spot: Collect the minimum data needed to answer your questions well. More isn’t always better—it’s often just more expensive and more burdensome.

 

Every person who provides data deserves:

  • Informed consent: Clear information about what participation involves and their rights
  • Privacy protection: Safeguards for personal information
  • Confidentiality: Assurance that individual responses won’t be shared
  • Freedom to decline: Ability to refuse or withdraw without consequences


Respect: Acknowledgment of their time and contribution.

Generate and Support Conclusions

Turn your data into actionable insights.

You’ve collected all this data. Now what? This step is about analysis—but more importantly, it’s about interpretation. Raw numbers and quotes don’t speak for themselves. Your job is to make meaning from them in collaboration with your interest holders.

Planning Your Analysis

Your analysis plan should be decided before you collect data (not after you see the results). This prevents cherry-picking and maintains objectivity.

For quantitative data:

  • Clean your data (check for errors, missing values, outliers)
  • Calculate descriptive statistics (percentages, means, frequencies)
  • Conduct appropriate statistical tests (if comparing groups or testing relationships)
  • Create visualizations (charts, graphs, tables)


For qualitative data:

  • Transcribe interviews or document observations
  • Read through all data to get familiar with it
  • Code the data (identify themes and patterns)
  • Organize codes into categories
  • Look for connections and relationships across themes
  • Select illustrative quotes


For both: Follow established, rigorous procedures. If you don’t have expertise in-house, bring in someone who does.

 

Here’s where evaluation gets interesting. Analysis produces results. Interpretation produces meaning.

To interpret findings:

  1. Compare to your expectations: Remember those benchmarks you set in Step 4? How do your results stack up? Did you hit, exceed, or fall short of expectations?

     

  2. Look for patterns: What themes emerge? What’s surprising? What confirms what you already suspected?

     

  3. Consider context: How do findings make sense (or not make sense) given what you know about the program, community, and environment?

     

  4. Engage interest holders: Bring your evaluation team together to discuss what the findings mean. Different perspectives will surface different interpretations—and that’s valuable.

     

  5. Consult existing evidence: How do your findings compare to what research or other evaluations have found?

     

Stay humble: Be clear about what you can and cannot conclude from your data. Acknowledge limitations. Resist overreaching.

 

Finding: Post-campaign surveys show that 55% of smokers exposed to the campaign correctly identified three or more smoking-related health conditions, compared to 30% pre-campaign.

Poor interpretation: “The campaign was a huge success and proves that media campaigns are the best way to change health behavior.”

Good interpretation: “The campaign appears to have increased awareness of smoking’s health risks among those exposed, exceeding our 50% target. However, we don’t yet know if increased awareness leads to quit attempts or successful cessation. We also note that the campaign didn’t equally reach all demographic groups—rural residents and Spanish speakers had lower exposure. For the campaign to achieve its ultimate goal of reducing smoking rates, we may need to strengthen distribution channels and language accessibility while also providing robust cessation support services.”

See the difference? Good interpretation:

  • References the benchmark
  • Notes successes
  • Identifies equity issues
  • Acknowledges what’s still unknown
  • Connects to broader program goals
  • Suggests actionable next steps

 

Recommendations are your suggestions for action based on the findings. Good recommendations are:

  • Rooted in evidence: Clearly connected to what you found
  • Specific and actionable: People know exactly what to do
  • Feasible: Realistic given resources and context
  • Prioritized: Not everything can be top priority
  • Assigned: Clear about who should take action
  • Timebound: Include a suggested timeline when appropriate


Example recommendations:

Less useful: “Improve the program.”

More useful: “To increase reach among Spanish speakers, the program should: (1) Partner with Spanish-language media outlets to place ads (responsibility: Communications Team, timeline: next campaign cycle), (2) Ensure the quitline offers services in Spanish during extended hours (responsibility: Quitline Coordinator, timeline: within 3 months), and (3) Train community health workers in Spanish-speaking communities to promote cessation resources (responsibility: Training Manager, timeline: within 6 months).”

 

Every evaluation has limitations. Naming them isn’t weakness—it’s scientific integrity. Common limitations include:

  • Sample size or response rate issues
  • Inability to determine causation
  • Short follow-up period
  • Missing data from key groups
  • Self-reported data (which can be biased)
  • Inability to access certain populations
  • Contextual changes during the evaluation

Don’t hide limitations in footnotes. Address them honestly and explain how they affect what you can conclude.

Act on Findings

Translate insights into action.

This is it—the step that determines whether your evaluation effort was worth it. Findings that sit in a report gathering dust help no one. Your job is to facilitate the use of what you’ve learned.

Planning for Use from the Start

Here’s a secret: If you wait until you have final findings to think about use, you’ve waited too long. Planning for use happens throughout the evaluation.

Ask early:

  • Who will use these findings?
  • For what specific decisions or actions?
  • When do they need the information?
  • In what format will it be most useful?
  • What might prevent them from acting on findings?
 

Timing matters: Information is only useful if it arrives when decisions are being made. A perfect evaluation delivered too late is useless.

Format matters: Not everyone wants a 50-page report. Consider:

  • Executive summaries (2-3 pages max) with key findings and recommendations
  • Slide decks for presentations
  • Infographics for visual learners
  • Briefs tailored to specific audiences (funders, policymakers, community members)
  • Data dashboards for ongoing monitoring
  • Story-based formats that humanize the data


Multiple formats for multiple audiences: Your funders might want an executive summary. Your community partners might prefer a presentation. Your communications team might need talking points. Create what people will actually use.

Good communication is:

Clear: Use plain language. Avoid jargon. If you must use technical terms, define them.

Balanced: Share both successes and challenges. Cherry-picking only positive findings destroys credibility.

Contextualized: Help people understand why findings matter and what they mean in the bigger picture.

Visual: Use charts, graphs, and images to make data accessible. But keep visualizations simple and honest—don’t distort data to make a point.

Culturally responsive: Consider your audience’s communication preferences, languages, and cultural norms. Avoid deficit-based language that blames communities.

 

As an evaluator, you’re not just a data analyst—you’re a facilitator of learning and change. To move from findings to action:

  1. Create space for dialogue: Don’t just present findings—discuss them. What surprises people? What resonates? What raises questions?
  2. Collaborate on recommendations: Work with interest holders to identify feasible actions. They know the context better than you do.
  3. Support decision-making: Help groups think through implications and options. What would it take to implement each recommendation? What are the trade-offs?
  4. Address barriers: What might prevent action? Politics? Resources? Capacity? Can those barriers be reduced?
  5. Celebrate learning: Frame evaluation as learning, not judgment. Create psychological safety for people to acknowledge what isn’t working so they can fix it.
  6. Follow up: Check in after findings are shared. Are they being used? What support do people need? What additional questions have emerged?

 

Misuse happens when evaluation findings are:

  • Taken out of context
  • Applied to questions they weren’t designed to answer
  • Cherry-picked to support a predetermined position
  • Used to punish rather than improve


To prevent misuse:

  • Be crystal clear about what the evaluation can and cannot tell you
  • Document limitations prominently
  • Provide the full context for any finding
  • Share findings with all interest holders simultaneously
  • Correct misinterpretations quickly and publicly
 

The best outcome of any evaluation isn’t a report—it’s a stronger culture of learning, inquiry, and evidence-based decision-making.

When evaluation is done well and used consistently, organizations develop:

  • Evaluative thinking: Staff routinely ask “Is this working? For whom? Under what conditions?”
  • Comfort with evidence: Data inform decisions as a matter of course
  • Learning from failure: Mistakes become opportunities for improvement, not sources of shame
  • Continuous improvement: Programs evolve based on ongoing learning


That’s when evaluation truly transforms how organizations work.

 

Putting It All Together: What Good Evaluation Looks Like

Let’s recap with a real-world lens. Good program evaluation:

Starts with humility: You don’t know everything. That’s why you’re evaluating.

Centers people: Those affected by the program help shape the evaluation and interpretation.

Tells the truth: Even when findings are uncomfortable or unexpected.

Balances rigor with feasibility: Uses the strongest methods the context allows.

Produces actionable insights: Findings lead to concrete improvements.

Advances equity: Examines and addresses who benefits, who doesn’t, and why.

Builds capacity: People learn evaluation skills and value evidence.

Drives continuous improvement: Programs get better over time based on what’s learned.

 

Common Evaluation Misconceptions (Debunked)

Myth #1: “Evaluation is too expensive and time-consuming.”

Truth: Evaluation costs and time should match your program’s scale and needs. A simple evaluation can be quick and affordable while still delivering valuable insights. You can’t afford not to evaluate—flying blind is more expensive in the long run.

Myth #2: “Evaluation requires fancy statistics and complex methods.”

Truth: The best evaluation uses methods appropriate to the question, not the fanciest methods available. Many valuable evaluations use straightforward approaches. What matters is rigor in execution, not complexity.

Myth #3: “Evaluation is punitive—it’s about catching people doing things wrong.”

Truth: Evaluation should be about learning and improvement, not punishment. When evaluation is used punitively, people stop being honest and the evaluation loses value. Create psychological safety and frame evaluation as a tool for success.

Myth #4: “We can’t evaluate because our outcomes take decades to achieve.”

Truth: Evaluate what you can measure now. Process evaluation examines whether you’re implementing well. Short and intermediate outcome evaluation tracks early progress. You don’t need to wait for long-term outcomes to learn valuable things.

Myth #5: “Evaluation findings never get used anyway.”

Truth: Findings get used when evaluation is designed with use in mind from the start. Involve decision-makers throughout. Time findings to inform decisions. Present findings in accessible ways. Plan for use explicitly. When done right, evaluation absolutely influences action.

Your Next Steps: Getting Started

Ready to evaluate your program? Here’s how to begin:

1. Assess your readiness 

  • Is your program clear about what it’s trying to achieve?
  • Do you have basic resources to support evaluation?
  • Who needs to be involved?

     

2. Get your program description clear

  • Create a simple logic model
  • Get consensus on the program logic
  • Identify your stage of development

     

3. Start with one focused question

  • What do you most need to know right now?
  • What decision would this inform?
  • Is it answerable with your resources?

     

4. Keep it simple

  • Use the simplest appropriate methods
  • Don’t try to answer everything at once
  • Build complexity as you build capacity

     

5. Get help if you need it

  • Evaluation is a skill—it’s okay to bring in expertise
  • Partner with universities or evaluation consultancies
  • Use free resources and tools available online

     

6. Remember: Done is better than perfect

  • An imperfect evaluation that gets used beats a perfect evaluation that never happens
  • Start small and build from there
  • Learn by doing

     

Final Thoughts: Evaluation as an Investment in Your Program's Future

Program evaluation isn’t a luxury or an afterthought—it’s an essential tool for creating change that matters. When you commit to evaluation, you’re saying:

We care enough about our work to know if it's working

We value the people we serve enough to ensure we're helping, not harming

We're willing to learn and adapt

We believe in evidence-based decision-making

We're accountable to the communities we serve

The frameworks, steps, and standards in this guide give you a roadmap. But remember: evaluation is as much art as science. It requires judgment, creativity, cultural responsiveness, and a genuine commitment to learning.

Your program deserves good evaluation. The people you serve deserve programs that work. And the insights you gain from evaluation will make you more effective, more accountable, and more capable of creating the change you want to see.

So start where you are. Use what you have. Learn as you go. And remember that every evaluation—even an imperfect one—is a step toward better programs and better outcomes.

Additional Resources

For more detailed guidance:


For building evaluation skills:

  • American Evaluation Association offers training and professional development
  • Many universities offer certificate programs in program evaluation
  • Online courses cover everything from basic concepts to advanced methods


For finding evaluators:

  • American Evaluation Association’s consultant directory
  • Local university evaluation centers
  • State and local health department evaluation staff
 

This web page is a concept developed by Chris Lysy of freshspectrum.com for educational purposes.  You can see the original CDC Framework by following this link.