So far, you’ve learned what the PEO framework is and why it’s ideal for observational research using retrospective data. Now it’s time to put it into action. Writing a clear, focused research question is one of the most important steps in your study—it guides your analysis, ensures your dataset can answer your question, and strengthens your eventual manuscript or abstract.

If you’re new to PEO, check out our beginner’s guide to the PEO framework for a full breakdown of the elements. For a refresher on how PEO compares to PICO, see our PEO vs PICO guide.

In this post, we’ll walk you through how to write your first PEO question step-by-step, including practical tips and examples you can follow.

Step 1: Identify Your Population

The population is the group of people you are studying. A well-defined population ensures that your results are meaningful and comparable to other research.

Tips for defining your population:

• Use standard age ranges from the medical literature. For example, “adults 18 and older” is preferable to arbitrary ranges like “20 to 31.” This helps others interpret your results and compare across studies. Common age breakdowns could be children (age 0-18), adults (age 18+), middle age (age 40-64), or elderly (age 65+).
  Specify relevant conditions, settings, or demographics. For example: “residents in internal medicine programs,” “patients with type 2 diabetes,” or “community-dwelling older adults.”
• Be precise but not overly narrow. It is possible to have an extremely well-defined population but totally unusable. (e.g: Adults age 18-20 who make $40-50k per year and have a dog named Charlie). Extremely narrow populations reduce the sample size of your study and limit your interpretations. 
• Ensure your dataset actually contains the population you want to study.
  

Example populations:

• Adults ≥18 years in the U.S.
• Pregnant women receiving prenatal care in a regional hospital network
• Adolescents aged 12–17 who have seen their primary care doctor
  

Step 2: Select Your Exposure

The exposure is the naturally occurring factor you want to investigate. Remember, unlike PICO, you are not introducing an intervention—the exposure already exists in the real world.

Tips for choosing an exposure:

• Make it measurable and clearly defined in your dataset.
• Focus on exposures that have enough variation so you can compare groups.
• Avoid vague exposures like “stress” without specifying how it is measured.
  

Example exposures:

• Excessive workload, defined as >70 hours of work per week
• Smoking status (current vs. non-smokers)
• Low physical activity levels, defined as <150 minutes of activity per week 

Step 3: Define Your Outcome

The outcome is the measurable result you are studying. A clear outcome ensures your analysis answers your research question. 

Tips for defining outcomes:

• The ideal outcome measure is objective (no ambiguity or subjectivity) and clinically relevant. 
• The easiest outcomes to measure are binary outcomes or continuous outcomes. 
• Be specific—avoid overly broad outcomes like “poor health” without a measurable indicator.

Example outcomes:

• Burnout score on a validated scale
• Medication adherence based on prescription refill data
• Body mass index (BMI)
  

Step 4: Formulate Your PEO Question

Once you’ve defined the three elements, you can combine them into a clear research question:

Formula:

In [Population], is [Exposure] associated with [Outcome]?

Examples:

1. In internal medicine residents, is heavy workload (defined by >70 hours/week) associated with increased burnout levels?
2. Among adults who smoke, is smoking associated with lower adherence to antihypertensive medications?
3. In adolescents aged 12–17, is low physical activity associated with higher BMI?
   

Notice how these are observational questions—you’re examining naturally occurring exposures, not testing an intervention.

Step 5: Validate Your Question

Before moving forward, make sure your question is feasible:

• Check your dataset: Are the population, exposure, and outcome present and well-documented?
• Consider sample size: Is there enough data to detect meaningful differences between groups?
• Think about confounders: Are there variables you may need to account for in your analysis?
• Ask yourself “Is this plausible?”: Based on everything you’ve learned, would it make sense if your exposure was associated with your outcome?
  

Practical Tips for Beginners

• Avoid arbitrary populations—stick to standardized age ranges or categories recognized in the literature.
• Make exposures and outcomes specific and measurable.
• Remember that PEO doesn’t explicitly include a comparison group, but your analysis inherently compares participants with vs. without the exposure.
• Document your thought process—this makes it easier to explain your research in abstracts or publications.
  

Next Steps

Once your PEO question is formulated:

• Translate it into a literature search to see what’s already known
• Plan your research timeline
• Prepare for data analysis
  

For a detailed walkthrough on defining PEO and examples for beginners, see our PEO beginner guide.

To understand when PICO might be a better fit or see more comparison examples, check out our PEO vs PICO guide.

At Lumono, we help medical trainees turn retrospective data into publishable insights. Whether you’re crafting your first PEO question or ready to analyze your dataset, our tools and guidance support every step of your research journey.