The Tool Desk
Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Check the original source, who was surveyed, when and how they were surveyed, and what the question actually asked. A poll is a dated estimate about a defined population—not an election result or a promise about what will happen on Election Day. Election statistics need their own provenance, reference date, and uncertainty notes.
Start with the original source, pollster, and sponsor
Do not judge a poll from a cropped graphic or repost if the original release or methodology note is available. Find the organization that conducted the poll and the organization that paid for it. Sponsorship is relevant context for evaluating incentives, but it does not by itself show that a result is false. AAPOR’s Journalist’s Guide to Understanding Polls & Surveys recommends checking both.
A useful release should provide a way to assess the poll’s population, recruitment and sample construction, survey mode, sample size, full question wording and answer options, weighting, and methods. AAPOR’s Disclosure Standards and Best Practices for Survey Research describe transparency information to look for. If key details are missing, treat the claim as harder to verify—not automatically false.
Check who was surveyed and when
Identify the population the poll represents: adults, registered voters, likely voters, or another group. Check the geography and the field dates—the dates interviews took place. A poll of adults does not automatically support a claim about likely voters in a particular state. And a poll taken before a major event cannot tell you how people responded to it.
Do these 3 things before closing this tab:
1Scan for outdated or missing drivers - takes under a minute2Repair Windows errors before they cause bigger problems3Fix the driver behind crashes, sound loss and screen glitches#1 Best Overall
- Used Book in Good Condition
Election polls are snapshots. AAPOR says they are not intended to predict an outcome and cautions against wording results as if they were actual election results or saying that a candidate is winning a race. More precise wording is: “The poll estimated support among [population] in [geography] during [field dates].” Views and turnout can change after the survey closes; late decisions may also affect the eventual result. See AAPOR’s Polling Accuracy for context on how polling estimates relate to election outcomes.
Read the question, choices, and surrounding context
Compare the headline with the exact question asked. Look for loaded assumptions, unequal descriptions of alternatives, missing answer choices, or a headline that describes a different issue than the question measured. The available choices and the order or context of questions can affect answers, too.
When a claim says opinion has changed, compare the question wording and preceding context across the surveys. AAPOR recommends keeping wording, framing, and methodology as similar as possible when measuring change. If a question must change, split-ballot testing—asking different versions to separate groups in the same survey—can help assess the effect of the change. A difference in answers is not necessarily a real shift in opinion.
Find out how respondents entered the sample
Ask how people were recruited, what survey mode was used, and whether the poll uses a probability or non-probability sample. In a probability sample, selection probabilities are known or can be calculated under the design. In a non-probability sample, people may be recruited through other methods; the pollster needs a defensible approach for relating those respondents to the population it aims to describe.
Recommended Free Tools
A self-selected poll on a website does not become representative just because many people took part. AAPOR says a conventional margin of sampling error should not be reported for non-probability samples. If a poll presents one, check whether its design supports that measure and what uncertainty approach was used.
Inspect weighting and likely-voter assumptions
Weighting changes how much each respondent contributes to an estimate, often to align the sample with population benchmarks. It can address some imbalances, but it does not automatically make an unrepresentative sample representative. Check which benchmarks were used and whether they match the population named in the claim.
For an election poll, also ask how likely voters were identified or modeled. Whether a person will vote is uncertain, and assumptions about turnout among different groups can move an estimate. A release should explain enough about its likely-voter approach to let readers understand what population the topline is meant to represent.
Interpret the margin of error and other uncertainty
A margin of sampling error describes sampling-related uncertainty under a relevant sample design; it is not an all-purpose accuracy guarantee. It does not, by itself, account for problems such as nonresponse, incomplete population coverage, question wording, survey mode, data processing, weighting choices, or likely-voter modeling. A large sample alone does not resolve those issues.
Quick wins for a faster PC:
Scan for outdated or missing drivers - takes under a minuteDriver Scan →Repair Windows errors before they cause bigger problemsFix Now →AAPOR’s Election Polling Resources says a candidate usually needs a lead of 1.5–2 times the margin of sampling error for that lead to be statistically significant. That is a rule of thumb in AAPOR’s current journalist guide, not a universal guarantee; the design and the specific comparison matter. Avoid describing a narrow lead as decisive without support from the poll’s analysis.
Be especially careful with subgroup results
Estimates for subgroups—such as voters in a particular age range—are based on fewer respondents than full-sample estimates and generally have greater sampling error. Check that the release gives the subgroup count and relevant uncertainty. A small subgroup lead may be much less informative than the full-sample topline suggests.
Compare polls without mistaking differences for movement
Before comparing two polls, line up their target populations, geographies, field dates, questions, recruitment methods, modes, weights, and likely-voter assumptions. A changed question, a different mode, or different turnout assumptions may produce different results even if opinion has not changed. Conversely, similar toplines do not prove that two polls used similar methods.
- Target: Are both polls about the same population, place, and voter-status definition?
- Timing: Were interviews conducted at similar times and in relation to the same major events?
- Question: Do wording, answer options, order, and preceding context match?
- Sample: Are recruitment and survey mode comparable?
- Adjustment: Do weighting benchmarks and likely-voter assumptions differ?
- Uncertainty: Are the estimates and comparison precise enough to support a claim of change?
If the methods or questions differ, describe what each poll found rather than asserting a trend. AAPOR’s survey best practices explain why consistent wording and methods matter when assessing change.
Verify election statistics at their original source
A poll is a survey estimate. Official election results come from election administration and reporting processes; other election-related statistics may come from separate programs with their own coverage and methods. For a chart or claim about official counts or statistical data, locate the original provider, confirm what the data cover and the reference date, and read its methods and uncertainty notes.
The U.S. Census Bureau’s Statistical Quality Standard E2: Reporting Results calls for source and date information, identification of sampling and non-sampling error, and appropriate uncertainty measures for relevant inferences and comparisons. It states that results that are not statistically significant must not be discussed in a way that implies they are significant. If a reported difference does not pass the applicable statistical test, do not present it as an established change.
Quick Recap
Red flags that call for a closer check
- A headline says a candidate “is winning” based only on a poll.
- A poll omits its sponsor, field dates, population, question wording, or methods.
- A self-selected, opt-in poll gives a conventional margin of error without explaining a defensible uncertainty method.
- A claimed trend compares different questions, populations, modes, or turnout assumptions.
- A small subgroup result appears without a respondent count.
- A statistic lacks an original source, reference date, or uncertainty information.
- A claimed change is treated as real even though the statistical comparison does not support it.
- A purported poll uses political telemarketing to push a message; AAPOR distinguishes such activity from legitimate polling and message testing in its election polling resources.
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
