---
title: "Underreporting — AP Stats Definition & Exam Guide"
description: "Underreporting is response bias where people report values lower than the truth, making estimates too low. Learn how AP Stats Topic 3.4 tests it on MCQs and FRQs."
canonical: "https://fiveable.me/ap-stats/key-terms/underreporting"
type: "key-term"
subject: "AP Statistics"
unit: "Unit 1"
---

# Underreporting — AP Stats Definition & Exam Guide

## Definition

Underreporting is a form of response bias in which subjects systematically report values lower than their true values (like calories eaten or hours of screen time), causing the sample statistic to underestimate the population parameter. It's tested in AP Stats Topic 3.4, Potential Problems with Sampling.

## What It Is

Underreporting happens when the people in your [sample](/ap-stats/unit-1/random-sampling-data-collection/study-guide/nQz8XwRMmIKKBS59qrew "fv-autolink") respond, but their answers are consistently lower than the truth. Ask someone how many calories they ate yesterday, how much they drink, or how many hours they spent on TikTok, and the average answer tends to come in below reality. Sometimes people genuinely forget. More often, they shade their answers toward what sounds acceptable.

The key word in the definition is *systematically*. Random measurement error bounces above and below the truth and tends to average out. Underreporting pushes every answer in the same [direction](/ap-stats/key-terms/direction "fv-autolink") (down), so the errors stack instead of canceling. That means your sample mean or sample proportion is [biased](/ap-stats/unit-3/biased-unbiased-point-estimates/study-guide/eZ5sR9XOkLB1o9KKpMHF "fv-autolink") low, and no amount of extra data fixes it. A bigger sample of people who all lowball their calorie intake just gives you a more precise wrong answer.

## Why It Matters

Underreporting lives in **[Unit 3](/ap-stats/unit-3 "fv-autolink"): Collecting Data, Topic 3.4 (Potential Problems with Sampling)**, under learning objective **3.4.A**: identify potential sources of [bias](/ap-stats/key-terms/bias "fv-autolink") in sampling methods. The essential knowledge there defines bias as occurring when certain responses are systematically favored over others, and underreporting is a textbook case where the lower responses are favored.

It matters beyond Unit 3 because bias poisons everything downstream. Every confidence interval and hypothesis test you build in Units 6 and 7 assumes the data measure what they claim to measure. If the data are underreported, your [interval](/ap-stats/unit-1/representing-quantitative-variable-with-graphs/study-guide/VWtyLVDvjzEgtbAi6v6j "fv-autolink") can be perfectly calculated and still miss the true parameter entirely. The AP exam loves making you spot that disconnect.

## Connections

### [Social desirability bias (Unit 3)](/ap-stats/key-terms/social-desirability-bias)

This is usually the *cause* of underreporting. People want to look good, so they report less of anything embarrassing (drinking, junk food, screen time) and more of anything admirable (flossing, exercise). Underreporting is the downward half of that pattern.

### Self-reporting bias and measurement bias (Unit 3)

Underreporting is one specific flavor of [self-reporting bias](/ap-stats/key-terms/self-reporting-bias "fv-autolink"), which is itself a form of measurement bias. The chain is simple. When the measurement instrument is a person describing themselves, the measurements drift toward flattering values.

### [Randomized response technique (Unit 3)](/ap-stats/key-terms/randomized-response-technique)

This is the classic fix. By adding a coin flip or other chance mechanism, respondents can answer sensitive questions honestly without anyone knowing their individual truth, which cuts down the incentive to underreport.

### Confidence intervals (Units 6-7)

Bias shifts the [center](/ap-stats/key-terms/center "fv-autolink") of your estimate, and a confidence interval is built around that center. If everyone underreports calories by 300, a 99% confidence interval for mean intake will be a precise interval around the wrong number. Confidence level controls random sampling error, not bias.

## On the AP Exam

Underreporting shows up on multiple-choice questions in two main ways. First, identify-the-bias stems: a survey scenario (a dentist asking patients face-to-face if they floss daily, participants self-reporting calorie intake) where you name the type of bias and, crucially, state the **direction**. If people underreport calories, the sample mean underestimates the true mean. Second, interpretation traps: a question gives you a valid-looking confidence interval built from self-reported data and asks whether it represents the true parameter. The answer hinges on recognizing that bias isn't fixed by the interval's math or a high confidence level.

On FRQs, this concept appears in study-design questions. The 2021 FRQ on a walking-and-cholesterol study is the kind of setup where self-reported behavior data invites bias commentary. When an FRQ asks you to identify a flaw or a source of bias, name the mechanism (people tend to report lower-than-true values for this behavior), then state the consequence (the estimate will be too low). Naming the bias without the direction usually leaves points on the table.

## underreporting vs Nonresponse bias

Both are sampling problems from Topic 3.4, but they break at different stages. Nonresponse bias happens when selected individuals don't answer at all, and the people who do respond differ from those who don't. Underreporting happens when people *do* respond but their answers are lower than the truth. Quick check: if the problem is missing people, think nonresponse. If the problem is dishonest or inaccurate answers, think response bias, and underreporting if those answers skew low.

## Key Takeaways

- Underreporting is a form of response bias where respondents give values systematically lower than the truth, so the sample statistic underestimates the population parameter.
- It usually hits sensitive or embarrassing topics like calorie intake, alcohol use, or screen time, often driven by social desirability bias.
- Increasing the sample size does not reduce underreporting, because bias is a systematic error, not random sampling variability.
- A confidence interval built from underreported data can be computed correctly and still completely miss the true parameter.
- On the exam, always state the direction of the bias, for example that the estimate of mean daily calories will be lower than the true mean.
- Anonymous surveys and the randomized response technique are the standard design fixes for underreporting on sensitive questions.

## FAQs

### What is underreporting in AP Stats?

Underreporting is a systematic response bias where people report values lower than their true values, like saying they ate 1,800 calories when they really ate 2,400. It causes sample statistics to underestimate the true population parameter, and it's covered in Topic 3.4 of Unit 3.

### Does a bigger sample size fix underreporting?

No. Sample size reduces random sampling variability, but underreporting is a systematic error that pushes every response in the same direction. A larger sample of people lowballing their answers just gives a more precise estimate of the wrong value.

### How is underreporting different from nonresponse bias?

With nonresponse bias, selected people fail to answer at all, and responders differ from non-responders. With underreporting, people answer but give values below the truth. One is a missing-data problem, the other is a wrong-data problem.

### Is underreporting the same thing as response bias?

Underreporting is one specific type of response bias. Response bias covers any systematic distortion in how people answer, including overreporting good behaviors (like 75% of dental patients claiming they floss daily). Underreporting is specifically the downward distortion.

### How do researchers reduce underreporting in surveys?

Use anonymous or confidential surveys, neutral question wording, and for very sensitive questions, the randomized response technique, which uses a chance mechanism so no individual answer can be traced back to a person. All of these reduce the pressure to shade answers downward.

## Related Study Guides

- [1.12 Potential Problems with Sampling](/ap-stats/unit-1/potential-problems-with-sampling/study-guide/nndgaR2dJGCIQs2UsTYa)

## Structured Data

```json
{"@context":"https://schema.org","@graph":[{"@type":"LearningResource","@id":"https://fiveable.me/ap-stats/key-terms/underreporting#resource","name":"Underreporting — AP Stats Definition & Exam Guide","url":"https://fiveable.me/ap-stats/key-terms/underreporting","learningResourceType":"Concept explainer","educationalLevel":"AP® / High School","about":{"@id":"https://fiveable.me/ap-stats/key-terms/underreporting#term"},"audience":{"@type":"EducationalAudience","educationalRole":"student"},"dateModified":"2026-06-11T05:22:53.868Z","isPartOf":{"@type":"Collection","name":"AP Statistics Key Terms","url":"https://fiveable.me/ap-stats/key-terms"},"publisher":{"@type":"Organization","name":"Fiveable","url":"https://fiveable.me"}},{"@type":"DefinedTerm","@id":"https://fiveable.me/ap-stats/key-terms/underreporting#term","name":"underreporting","description":"Underreporting is a form of response bias in which subjects systematically report values lower than their true values (like calories eaten or hours of screen time), causing the sample statistic to underestimate the population parameter. It's tested in AP Stats Topic 3.4, Potential Problems with Sampling.","url":"https://fiveable.me/ap-stats/key-terms/underreporting","inDefinedTermSet":{"@type":"DefinedTermSet","name":"AP Statistics Key Terms","url":"https://fiveable.me/ap-stats/key-terms"}},{"@type":"FAQPage","mainEntity":[{"@type":"Question","name":"What is underreporting in AP Stats?","acceptedAnswer":{"@type":"Answer","text":"Underreporting is a systematic response bias where people report values lower than their true values, like saying they ate 1,800 calories when they really ate 2,400. It causes sample statistics to underestimate the true population parameter, and it's covered in Topic 3.4 of Unit 3."}},{"@type":"Question","name":"Does a bigger sample size fix underreporting?","acceptedAnswer":{"@type":"Answer","text":"No. Sample size reduces random sampling variability, but underreporting is a systematic error that pushes every response in the same direction. A larger sample of people lowballing their answers just gives a more precise estimate of the wrong value."}},{"@type":"Question","name":"How is underreporting different from nonresponse bias?","acceptedAnswer":{"@type":"Answer","text":"With nonresponse bias, selected people fail to answer at all, and responders differ from non-responders. With underreporting, people answer but give values below the truth. One is a missing-data problem, the other is a wrong-data problem."}},{"@type":"Question","name":"Is underreporting the same thing as response bias?","acceptedAnswer":{"@type":"Answer","text":"Underreporting is one specific type of response bias. Response bias covers any systematic distortion in how people answer, including overreporting good behaviors (like 75% of dental patients claiming they floss daily). Underreporting is specifically the downward distortion."}},{"@type":"Question","name":"How do researchers reduce underreporting in surveys?","acceptedAnswer":{"@type":"Answer","text":"Use anonymous or confidential surveys, neutral question wording, and for very sensitive questions, the randomized response technique, which uses a chance mechanism so no individual answer can be traced back to a person. All of these reduce the pressure to shade answers downward."}}]},{"@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"AP Statistics","item":"https://fiveable.me/ap-stats"},{"@type":"ListItem","position":2,"name":"Key Terms","item":"https://fiveable.me/ap-stats/key-terms"},{"@type":"ListItem","position":3,"name":"Unit 1","item":"https://fiveable.me/ap-stats/unit-1"},{"@type":"ListItem","position":4,"name":"underreporting"}]}]}
```