What Is a Null Hypothesis?
A null hypothesis is a statement that there is no effect, no difference, or no relationship between the variables in a study. It is the default position researchers start from — the assumption that nothing interesting is happening — until the evidence proves otherwise.
Think of it this way. Suppose you want to test whether a new study technique improves exam scores. Your null hypothesis would say: the new study technique makes no difference to exam scores. You then collect data, run a statistical test, and see whether the evidence is strong enough to reject that statement. If it is, you conclude the technique probably does work. If it is not, you keep the null hypothesis — you do not prove it true, you simply fail to reject it.
The null hypothesis is usually written as H0 (H-naught). It is the backbone of hypothesis testing, the statistical framework behind almost every quantitative study, from psychology experiments to medical trials.
Why Do Researchers Need a Null Hypothesis?
It might seem strange to start a study by stating the opposite of what you hope to find. But the null hypothesis exists for a very practical reason: science advances by trying to prove ideas wrong, not by trying to prove them right.
A researcher who believes a new therapy reduces anxiety cannot simply collect a few success stories and declare victory. That would be cherry-picking. Instead, the scientific method demands a fair test: assume the therapy does nothing, then ask whether the data are surprising enough under that assumption to abandon it. If patients improve far more than chance alone would explain, the null hypothesis is rejected and the therapy earns a serious hearing.
The null hypothesis also gives statistics something concrete to calculate. A statistical test asks: if the null hypothesis were true, how likely would these results be? That likelihood is the p-value. A small p-value (conventionally below 0.05) means the results would be very unlikely if nothing were going on, so the null hypothesis is rejected. A larger p-value means the results are perfectly ordinary under the null, so it stays.
Null Hypothesis vs Alternative Hypothesis
Every hypothesis test has two competing statements. The null hypothesis (H0) says there is no effect. The alternative hypothesis (H1 or Ha) says there is an effect — it is usually the claim the researcher actually hopes to support.
| Null Hypothesis (H0) | Alternative Hypothesis (H1) | |
|---|---|---|
| What it claims | No effect, no difference, no relationship | An effect, a difference, or a relationship exists |
| Researcher’s attitude | Assumed true until evidence says otherwise | What the researcher usually hopes to support |
| Example | There is no difference in stress levels between students who meditate and students who do not. | Students who meditate have lower stress levels than students who do not. |
| Possible outcomes | Rejected, or not rejected (never “proven”) | Supported when the null is rejected |
| Symbol | H0 | H1 or Ha |
Notice the asymmetry. A study can reject the null hypothesis, but it can never prove it. Failing to reject the null simply means the evidence was not strong enough — perhaps the effect is genuinely absent, or perhaps the sample was too small to detect it. This is why careful researchers write “we found no significant difference” rather than “there is no difference”.
How to State a Null Hypothesis: 5 Rules
Writing a good null hypothesis is a skill, and examiners notice when it is done well. Follow these five rules.
1. State it as a claim about the population, not the sample
Your hypothesis is about the wider group you want to understand, not the handful of people in your study. Write “There is no difference in mean anxiety scores between the two groups in the population” rather than “The two groups in my sample scored the same.”
2. Make it testable and specific
A null hypothesis must be something data can actually challenge. “Social media is bad for students” is a vague opinion, not a null hypothesis. “There is no relationship between daily social media use (in hours) and exam scores among undergraduates” is specific, measurable, and testable.
3. Name the variables clearly
Every null hypothesis should make its variables obvious. If your study is about sleep and memory, say so: “There is no difference in memory test scores between participants who sleep 8 hours and participants who sleep 5 hours.” A reader should never have to guess what you measured.
4. Keep it in the “no effect” form
The null always claims no difference, no relationship, or no change. It is not a prediction of the opposite effect — “Group A will score lower than Group B” is an alternative hypothesis, not a null. The null for that study is “There is no difference between Group A and Group B.”
5. Match it to your statistical test
Different tests answer different null hypotheses. A t-test asks whether two group means differ. A correlation test asks whether two variables are related. A chi-square test asks whether two categorical variables are associated. Before you write your null hypothesis, know which test you will run, and phrase the null in that test’s language.
Null Hypothesis Examples
Seeing real examples is the fastest way to learn the pattern. Here are null hypotheses from different fields, each paired with its alternative.
Example 1: Psychology experiment
Research question: Does mindfulness training reduce test anxiety?
Null hypothesis (H0): There is no difference in test anxiety scores between students who complete mindfulness training and students who do not.
Alternative hypothesis (H1): Students who complete mindfulness training have lower test anxiety scores than students who do not.
Example 2: Education study
Research question: Is there a relationship between class attendance and grades?
Null hypothesis (H0): There is no relationship between class attendance and final grades among first-year students.
Alternative hypothesis (H1): There is a relationship between class attendance and final grades among first-year students.
Example 3: Health research
Research question: Does a new drug lower blood pressure more than the standard drug?
Null hypothesis (H0): There is no difference in blood pressure reduction between patients taking the new drug and patients taking the standard drug.
Alternative hypothesis (H1): Patients taking the new drug show greater blood pressure reduction than patients taking the standard drug.
Example 4: Business research
Research question: Do customers prefer the new packaging?
Null hypothesis (H0): There is no difference in customer preference between the new packaging and the old packaging.
Alternative hypothesis (H1): Customers prefer the new packaging over the old packaging.
Notice the pattern in every example: the null hypothesis always claims “no difference” or “no relationship”, while the alternative claims the effect the researcher is investigating.
Common Mistakes When Writing a Null Hypothesis
Mistake 1: Writing the null as a hope, not a “no effect” claim
Beginners often write their null hypothesis as what they want to happen. If you hope a tutoring program raises scores, your null is not “the program raises scores” — it is “the program makes no difference to scores”. The null is always the sceptical position.
Mistake 2: Saying the null hypothesis is “proven” or “accepted”
A non-significant result never proves the null hypothesis. The correct language is “we failed to reject the null hypothesis”. Statistics can only tell you the evidence was not strong enough — absence of evidence is not evidence of absence.
Mistake 3: Making the null hypothesis untestable
“Students are happier when they study what they love” cannot be tested until “happier” and “what they love” are defined and measured. A null hypothesis needs named variables and a clear measurement, or the statistical test has nothing to work with.
Mistake 4: Confusing the null with the research question
A research question asks something (“Does sleep affect memory?”). A null hypothesis answers it in advance with the sceptical position (“There is no difference in memory scores between well-rested and sleep-deprived participants”). Both belong in a thesis, but they are different sentences doing different jobs.
Mistake 5: Writing a null that does not match the test
If your null hypothesis is about a relationship between two continuous variables but you run a t-test, something is wrong — a t-test compares group means, it does not test correlations. Always check that your null, your test, and your variables line up before you collect data.
What Happens After You State the Null Hypothesis?
Once the null and alternative hypotheses are written, the rest of hypothesis testing follows a standard sequence. First, choose a significance level — usually 0.05 — before looking at the data. Then collect the data and run the appropriate statistical test. The test produces a p-value: the probability of seeing results at least this extreme if the null hypothesis were true. If the p-value is below the significance level, reject the null hypothesis and conclude the evidence supports the alternative. If it is above, fail to reject the null and report the result honestly.
Two errors are possible along the way. A Type I error means rejecting a null hypothesis that is actually true — a false alarm. A Type II error means failing to reject a null hypothesis that is actually false — a missed signal. Good research design, an adequate sample size, and honest reporting are the best defences against both.
Key Takeaways
- A null hypothesis states that there is no effect, no difference, or no relationship between the variables being studied.
- It is the default, sceptical position that researchers try to disprove with data — never something a study can prove.
- Every null hypothesis has a partner: the alternative hypothesis, which claims the effect the researcher is investigating.
- A good null hypothesis is specific, testable, names its variables, and matches the statistical test that will be used.
- When results are not significant, the correct conclusion is “we failed to reject the null hypothesis” — not “the null is true”.
Frequently Asked Questions
What is a null hypothesis in simple words?
A null hypothesis is a statement that says “nothing is happening” — no difference between groups, no relationship between variables, no effect of a treatment. Researchers assume it is true at the start, then use data to test whether it should be rejected.
What is the difference between a null hypothesis and an alternative hypothesis?
The null hypothesis (H0) claims there is no effect or difference. The alternative hypothesis (H1) claims there is an effect or difference. The statistical test tries to reject the null in favour of the alternative; if the evidence is too weak, the null is kept.
Can a null hypothesis ever be proven?
No. A study can reject the null hypothesis or fail to reject it, but it can never prove it true. A non-significant result only means the evidence was not strong enough — it does not prove that no effect exists.
How do you write a null hypothesis?
Write a clear, testable sentence that claims no difference or no relationship between your named variables, stated about the population rather than your sample. For example: “There is no difference in exam scores between students who attend revision classes and students who do not.”
What is an example of a null hypothesis?
If you are testing whether a new teaching method improves grades, the null hypothesis is: “There is no difference in grades between students taught with the new method and students taught with the traditional method.” The alternative is that the new method produces different — usually higher — grades.