---
title: "Sensitivity Analyses | Intro to Epidemiology"
description: "Sensitivity analyses in Intro to Epidemiology test how study results change when assumptions shift, showing how robust findings are to bias or missing data."
canonical: "https://fiveable.me/introduction-epidemiology/key-terms/sensitivity-analyses"
type: "key-term"
subject: "Intro to Epidemiology"
unit: "Unit 8"
---

# Sensitivity Analyses | Intro to Epidemiology

## Definition

Sensitivity analyses are checks in epidemiology that show how much study results change when you vary assumptions, missing data handling, or analysis choices. They tell you whether the finding is stable or fragile.

## What It Is

Sensitivity analyses are repeat checks that ask, “What happens if I change one key assumption in this epidemiologic study?” In Intro to Epidemiology, that usually means re-running an analysis with different choices about missing data, exposure definitions, outcome rules, or possible bias sources to see whether the conclusion stays the same.

The basic idea is simple: a study result is only as strong as the assumptions behind it. If your estimate changes only a little when you tweak those assumptions, the result is fairly robust. If the estimate flips direction or gets much larger or smaller, then the finding may depend too much on one modeling choice or one data decision.

This comes up a lot when researchers do not have perfect data. Maybe some participants are lost to follow-up, maybe a variable was measured a little differently across groups, or maybe the study has missing responses. A sensitivity analysis lets you test several reasonable scenarios instead of pretending there is only one correct answer.

In epidemiology, these analyses are not about proving a study right. They are about checking how vulnerable the result is to bias, uncertainty, or measurement error. That makes them closely tied to the bias unit in the course, because many sensitivity analyses are designed to ask, “Could this association be explained away if the bias were stronger or weaker than we assumed?”

A simple example is a study on a new screening program where some follow-up data are missing. One analysis might assume those missing participants had the same outcome rate as the rest of the sample. Another might assume they had a worse outcome rate. If the main conclusion stays similar across both versions, the findings are more trustworthy. If not, you have learned something useful about the limits of the study.

## Why It Matters

Sensitivity analyses matter because epidemiology is full of imperfect data and imperfect choices. You rarely get a clean population, perfect measurement, and complete follow-up all at once. This method shows how much confidence you should place in a reported association, risk estimate, or public health claim.

It also helps you read research more critically. A result can look impressive at first glance, but if it only appears under one narrow set of assumptions, it is less convincing. Sensitivity analyses reveal which assumptions are driving the story, which is a big deal when you are evaluating studies for bias, confounding, or missing information.

In public health, that matters for decisions. If a finding stays stable across several reasonable scenarios, it is easier to use in policy, screening, or intervention planning. If it falls apart under small changes, then the study may need better design, more data, or a more cautious conclusion.

For class work, this term also trains you to think like an epidemiologist. You are not just calculating an answer. You are asking how dependable that answer is when real-world problems, like loss to follow-up or measurement error, enter the picture.

## Connections

### Bias

Sensitivity analyses are often built around bias. You change an assumption about how much bias might be present, then see how much the study estimate moves. That makes the method a practical way to test whether a result is still believable after you account for possible distortion.

### Confounding

Confounding can make an exposure-outcome relationship look stronger or weaker than it really is. A sensitivity analysis can ask how much an unmeasured confounder would need to affect the data to change the conclusion. That helps you judge whether the observed association is sturdy or easy to overturn.

### [Loss to follow-up](/introduction-epidemiology/key-terms/loss-to-follow-up)

When participants disappear before the study ends, the final result may depend on what you assume happened to them. Sensitivity analyses let you compare different follow-up scenarios, such as best case and worst case outcomes. That is especially useful in cohort studies and intervention studies.

### [Measurement error](/introduction-epidemiology/key-terms/measurement-error)

If exposure or outcome data were recorded imperfectly, the main analysis may be off. Sensitivity analyses let you test alternative levels of measurement error to see whether the overall conclusion changes. This is a useful way to evaluate whether a weak result is just a data-quality problem.

## On the AP Exam

A quiz or case-based question may give you a study result and ask whether it still looks credible after the assumptions change. Your job is to explain what the sensitivity analysis is testing, not just name it. You might compare two scenarios for missing data, trace how a different bias assumption changes the risk estimate, or decide whether a conclusion is robust. If the numbers barely change, you would describe the finding as more stable. If a small change in assumptions makes the result collapse, you would point to that as evidence the study is sensitive to bias or uncertainty. On written responses, this term often shows up when you are asked to critique a paper, discuss limitations, or explain why researchers ran more than one analysis.

## Key Takeaways

- Sensitivity analyses check whether an epidemiologic result changes when you alter key assumptions or analysis choices.
- A stable result across several scenarios is more robust than one that changes a lot with a small tweak.
- This method is especially useful when data are incomplete, measurements are imperfect, or follow-up is missing.
- Sensitivity analyses do not erase bias, but they show how much a study depends on uncertain choices.
- If a conclusion only works under one assumption, you should treat it more cautiously.

## FAQs

### What is sensitivity analyses in Intro to Epidemiology?

Sensitivity analyses are repeated checks that test how a study’s results change when you vary assumptions, missing data rules, or possible bias. In Intro to Epidemiology, they help you judge whether a conclusion is solid or whether it depends on one narrow set of choices.

### How do sensitivity analyses relate to bias?

They are one of the main tools for seeing how bias might affect a study’s conclusion. You change the size or direction of a suspected bias and see whether the result stays the same. If it does, the finding is more convincing. If it changes a lot, the study may be more fragile than it first looked.

### What is a simple example of a sensitivity analysis?

Imagine a study with missing follow-up data. One analysis assumes the missing participants had outcomes similar to the rest of the sample, while another assumes worse outcomes. If both analyses give nearly the same answer, the finding is robust. If the conclusion changes, missing data is clearly affecting the result.

### Is sensitivity analysis the same as controlling for confounding?

No. Controlling for confounding is part of the main analysis, while a sensitivity analysis checks how much an unmeasured or uncertain factor could change the conclusion. They work together, but they are not the same move. Sensitivity analysis is more about stress-testing the result.

## Related Study Guides

- [8.1 Types of bias in epidemiologic studies](/introduction-epidemiology/unit-8/types-bias-epidemiologic-studies/study-guide/hncr5f4OS7iZ8EDy)

## About This Document

Canonical Fiveable pages are available as Markdown at the same path plus `.md`.

- [llms.txt](https://fiveable.me/llms.txt): index of Fiveable's sections and URL patterns
- [llms-full.txt](https://fiveable.me/llms-full.txt): complete subject and unit listing
- [MCP server](https://fiveable.me/mcp): call Fiveable as tools instead of fetching pages (`https://fiveable.me/api/mcp`)
- [MCP server for AP teachers](https://fiveable.me/mcp/teachers): a teacher's classes, assignments and AP-rubric grading (`https://fiveable.me/api/mcp/teacher`)

## Structured Data

```json
{"@context":"https://schema.org","@graph":[{"@type":"LearningResource","@id":"https://fiveable.me/introduction-epidemiology/key-terms/sensitivity-analyses#resource","name":"Sensitivity Analyses | Intro to Epidemiology","url":"https://fiveable.me/introduction-epidemiology/key-terms/sensitivity-analyses","learningResourceType":"Concept explainer","educationalLevel":"AP® / High School","about":{"@id":"https://fiveable.me/introduction-epidemiology/key-terms/sensitivity-analyses#term"},"audience":{"@type":"EducationalAudience","educationalRole":"student"},"dateModified":"2026-07-03T02:22:45.907Z","isPartOf":{"@type":"Collection","name":"Intro to Epidemiology Key Terms","url":"https://fiveable.me/introduction-epidemiology/key-terms"},"author":{"@type":"Organization","@id":"https://fiveable.me/editorial#content-team","name":"Fiveable Content Team","url":"https://fiveable.me/editorial","parentOrganization":{"@id":"https://fiveable.me/#organization"},"member":[{"@type":"Person","name":"Amanda DoAmaral","sameAs":["https://www.linkedin.com/in/adoamaral"]},{"@type":"Person","name":"Becky Bahr","sameAs":["https://www.linkedin.com/in/becky-bahr-b52743132"]}]},"publisher":{"@type":"Organization","name":"Fiveable","url":"https://fiveable.me"}},{"@type":"DefinedTerm","@id":"https://fiveable.me/introduction-epidemiology/key-terms/sensitivity-analyses#term","name":"sensitivity analyses","description":"Sensitivity analyses are checks in epidemiology that show how much study results change when you vary assumptions, missing data handling, or analysis choices. They tell you whether the finding is stable or fragile.","url":"https://fiveable.me/introduction-epidemiology/key-terms/sensitivity-analyses","inDefinedTermSet":{"@type":"DefinedTermSet","name":"Intro to Epidemiology Key Terms","url":"https://fiveable.me/introduction-epidemiology/key-terms"}},{"@type":"FAQPage","mainEntity":[{"@type":"Question","name":"What is sensitivity analyses in Intro to Epidemiology?","acceptedAnswer":{"@type":"Answer","text":"Sensitivity analyses are repeated checks that test how a study’s results change when you vary assumptions, missing data rules, or possible bias. In Intro to Epidemiology, they help you judge whether a conclusion is solid or whether it depends on one narrow set of choices."}},{"@type":"Question","name":"How do sensitivity analyses relate to bias?","acceptedAnswer":{"@type":"Answer","text":"They are one of the main tools for seeing how bias might affect a study’s conclusion. You change the size or direction of a suspected bias and see whether the result stays the same. If it does, the finding is more convincing. If it changes a lot, the study may be more fragile than it first looked."}},{"@type":"Question","name":"What is a simple example of a sensitivity analysis?","acceptedAnswer":{"@type":"Answer","text":"Imagine a study with missing follow-up data. One analysis assumes the missing participants had outcomes similar to the rest of the sample, while another assumes worse outcomes. If both analyses give nearly the same answer, the finding is robust. If the conclusion changes, missing data is clearly affecting the result."}},{"@type":"Question","name":"Is sensitivity analysis the same as controlling for confounding?","acceptedAnswer":{"@type":"Answer","text":"No. Controlling for confounding is part of the main analysis, while a sensitivity analysis checks how much an unmeasured or uncertain factor could change the conclusion. They work together, but they are not the same move. Sensitivity analysis is more about stress-testing the result."}}]},{"@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Intro to Epidemiology","item":"https://fiveable.me/introduction-epidemiology"},{"@type":"ListItem","position":2,"name":"Key Terms","item":"https://fiveable.me/introduction-epidemiology/key-terms"},{"@type":"ListItem","position":3,"name":"Unit 8","item":"https://fiveable.me/introduction-epidemiology/unit-8"},{"@type":"ListItem","position":4,"name":"sensitivity analyses"}]}]}
```