A binomial test checks whether the proportion of successes in a binary outcome matches a claimed value. In Intro to Statistics, you use it for yes/no or success/failure data when you want to test a specific proportion.
A binomial test is a hypothesis test for one categorical variable with two outcomes, usually called success and failure. In Intro to Statistics, you use it when you want to check whether the true success probability matches a claimed value, like whether a coin is fair or whether a treatment works for a certain percent of patients.
The setup starts with a fixed number of trials, each trial has only two outcomes, the trials are independent, and the success probability stays the same from trial to trial. If those conditions fit, you can model the number of successes with the binomial distribution and compare your observed result to what the null hypothesis predicts.
That null hypothesis usually says the population proportion equals some specific value, written as p = p0. For example, if a manufacturer claims 10% of items are defective, a binomial test checks whether the sample has too many or too few defects to be explained by chance alone. The test asks, "If the claim were true, how surprising is this sample?"
The p-value comes from binomial probabilities. For a one-tailed test, you add the probabilities of results at least as extreme in one direction. For a two-tailed test, you look for extreme outcomes in both directions, which is why the exact p-value can take a little care to compute.
A small sample is where this test is especially useful, since you may not want to rely on normal approximations. If the sample is larger and the expected counts are comfortably big, some classes move to an approximation instead, but the exact binomial test stays tied directly to the binomial model.
A common mistake is to treat any yes/no data as automatically binomial. You still need a fixed number of trials, independent observations, and the same success probability each time. If those pieces are missing, the test no longer fits the situation well.
The binomial test shows you how Intro to Statistics turns a claim about a proportion into a decision. Instead of guessing whether an observed count looks "high" or "low," you use probability to measure how unusual the sample would be if the null hypothesis were true.
That makes it a useful bridge between binomial distributions and hypothesis testing. Early in the course, you learn how to calculate probabilities for exact numbers of successes. The binomial test takes that same machinery and uses it to evaluate evidence in a data set, which is a big step toward formal statistical inference.
It also shows up any time the data are naturally binary. Survey responses like yes/no, quality control checks like defective/not defective, and medical outcomes like response/no response all fit the structure. Once you can spot the binary outcome and the fixed trial count, you know when the binomial test is a good tool.
This term also sharpens your interpretation skills. A binomial test does not prove a claim true or false in an absolute sense. It tells you whether the sample result is surprising enough, under the null model, to reject that claim at your chosen significance level.
Keep studying Intro to Statistics Unit 4
Visual cheatsheet
view galleryBinomial Distribution
The binomial test is built on the binomial distribution. You use the distribution to calculate the chance of getting a certain number of successes when the probability of success stays fixed across independent trials. The test borrows that probability model and turns it into a hypothesis test about a population proportion.
Null Hypothesis
In a binomial test, the null hypothesis usually states a specific success probability, such as p = 0.50 or p = 0.10. Your observed sample is then compared to what you would expect if that null claim were true. If the sample is too unusual, you reject the null hypothesis.
P-value
The p-value is the output you use to judge the binomial test. It measures how likely it would be to see your sample result, or something more extreme, if the null hypothesis were correct. A small p-value means your data are not very compatible with the claimed proportion.
Normal Approximation
When the sample is larger, classes may compare the exact binomial test to the normal approximation. The approximation is faster, but it is not always accurate for small samples or extreme probabilities. The exact binomial test stays closer to the true binomial probabilities.
A quiz or problem set question will usually give you a sample size, a claimed proportion, and the number of successes, then ask you to test the claim. Your job is to identify the null hypothesis, decide whether the test is one-tailed or two-tailed, and find the p-value from the binomial model or calculator output.
You may also need to explain the conclusion in context, not just say "reject" or "fail to reject." For example, if a company claims 20% of customers prefer a product, you would write whether the sample gives enough evidence that the true preference rate differs from 20%. The strongest answers connect the math back to the real setting and avoid saying the null is "proven false."
These two are often mixed up because both deal with binomial outcomes and sample proportions. The binomial test is exact and uses the binomial distribution directly, which is best for smaller samples or edge cases. The normal approximation replaces the binomial with a normal curve when conditions are strong enough and the sample is large.
A binomial test checks whether a sample proportion is consistent with a specific claimed success probability.
It works only when the data fit a binomial setting: fixed trials, two outcomes, independence, and the same success probability each time.
The p-value comes from binomial probabilities, so the result is tied directly to the binomial distribution.
Use the null hypothesis as the baseline claim, then judge whether the observed number of successes is unusually far from that claim.
In Intro to Statistics, this test often appears with yes/no data, quality control counts, or simple treatment response data.
A binomial test is a hypothesis test for a single binary outcome, like success/failure or yes/no. It checks whether the observed number of successes matches a claimed proportion under the null hypothesis. You use the binomial distribution to see how surprising the sample would be if that claim were true.
Use a binomial test when the sample is small or when you want an exact probability based on the binomial distribution. A z test for proportions is often an approximation that works better with larger samples. If your class wants exact inference for a one-proportion question, binomial test is usually the safer choice.
Check for four things: a fixed number of trials, only two outcomes, independent trials, and the same success probability on each trial. If one of those conditions fails, the binomial test may not fit. A common mistake is forgetting that the probability has to stay constant across trials.
The p-value is the probability of getting your observed result, or something more extreme, if the null hypothesis is true. A small p-value means the sample is unlikely under the claimed proportion. That gives you evidence against the null, but it does not prove a new proportion with certainty.