A hypothesis is a testable prediction that proposes an explanation for something you’ve observed. Think of it as an educated guess, but with an important requirement: it has to be something you can actually test and potentially prove wrong. In science, a hypothesis acts as the bridge between a question and an experiment, giving researchers a specific claim to investigate.
How a Hypothesis Works in the Scientific Method
The scientific method follows a sequence: observe something interesting, ask a question about it, form a hypothesis, then design an experiment to test that hypothesis. The hypothesis comes third in this chain, after you’ve already noticed a pattern or puzzle and done some background research. It’s the moment you commit to a specific, testable explanation before collecting new data.
This timing matters. A hypothesis formed before the experiment keeps the process honest. If you only explain your results after seeing them, you’re just describing what happened, not predicting it. A good hypothesis forces you to put your idea on the line, then let the evidence decide whether it holds up. When hypotheses are tested correctly, they reveal not just whether something is true, but how likely it is to happen again.
What Makes a Hypothesis Valid
Not every guess counts as a scientific hypothesis. Researchers use what’s sometimes called the “5E rule” to evaluate whether a hypothesis is strong enough to build an experiment around:
- Explicit: It states a clear, specific prediction rather than a vague idea.
- Evidence-based: It’s grounded in existing knowledge, not just a random hunch.
- Ex-ante: It’s formulated before the experiment, not reverse-engineered from results.
- Explanatory: It offers a reason why something happens, not just that it happens.
- Empirically testable: It can be confirmed or disproven through observation or experiment.
That last criterion, testability, is the non-negotiable one. The philosopher Karl Popper argued that falsifiability is what separates science from non-science. A hypothesis must make predictions that could be disproven by experimental evidence. If no possible result could prove your hypothesis wrong, it isn’t really a scientific hypothesis. “Luck influences exam scores” isn’t testable. “Students who sleep 8 hours before an exam score higher than those who sleep 4 hours” is.
How to Write One
The most common format is the “if… then…” structure. You state a condition and predict an outcome. For example: “If you apply Treatment A for male pattern baldness, then you will see a 50% increase in hair growth within three months.” The “if” portion identifies what you’re changing (the independent variable), and the “then” portion identifies what you expect to happen as a result (the dependent variable).
Another workable structure is “when X, then Y.” For instance: “When cattle are fed a steady diet of Antibiotic X, then traces of Antibiotic X remain in beef produced for consumers.” Both formats accomplish the same thing: they pin down specific, measurable variables so the hypothesis can actually be tested. If you can’t identify what you’d observe or measure to know whether your prediction came true, the hypothesis needs to be more specific.
Null and Alternative Hypotheses
When researchers move from forming a hypothesis to statistically testing it, they split it into two opposing statements. The null hypothesis states that there is no meaningful difference or relationship between the things being studied. It represents the default assumption: nothing interesting is going on. The alternative hypothesis states that there is a real difference or relationship, and it’s what the researcher is trying to find evidence for.
Say a researcher wants to know whether a new teaching method improves test scores. The null hypothesis would be: “There is no difference in scores between students taught with the new method and students taught with the old method.” The alternative hypothesis would be: “Students taught with the new method score higher.” The entire experiment is then designed to determine which of these two statements the data supports.
Statistical tests produce a number called a p-value, which tells you how likely your results would be if the null hypothesis were true. By convention, if the p-value falls below 0.05 (a 5% probability), researchers reject the null hypothesis and accept the alternative. This threshold isn’t magic, but it’s the standard cutoff used across most health and science research. In practical terms, it means there’s less than a 1-in-20 chance the results happened by random variation alone.
Real-World Examples in Health Research
Hypotheses drive every clinical trial. In the famous Physicians’ Health Study, the hypothesis was that taking aspirin would reduce mortality compared to a placebo. Researchers didn’t just give people aspirin and see what happened. They predicted a specific outcome (lower death rates), designed a controlled experiment, and then measured whether the data supported that prediction. A secondary hypothesis in the same study predicted that aspirin would also reduce fatal and nonfatal heart attacks.
Each of these hypotheses met the criteria: they were specific, testable, formed before the trial began, and grounded in earlier evidence suggesting aspirin had cardiovascular benefits. The structure of the trial, including the use of a placebo group, existed specifically to test whether the hypotheses held up.
Hypothesis vs. Theory vs. Law
These three terms are often used interchangeably in casual conversation, but they mean very different things in science, and they’re not steps on a ladder where a hypothesis “graduates” into a theory.
A hypothesis is a single testable prediction that hasn’t yet been confirmed. It’s narrow and specific. A scientific theory is a broad, self-consistent framework that has been verified experimentally across many situations. The theory of evolution, for example, is a vast collection of proven principles, observations, and laws that together explain how species change over time. It didn’t start as one hypothesis that got promoted. It’s built from thousands of confirmed hypotheses, experiments, and observations unified into a coherent system.
A scientific law is different from both. It’s a single, proven statement describing something the universe consistently does. Newton’s law of universal gravitation, for instance, is one equation describing how objects attract each other. It tells you what happens, not why. A theory explains the “why” by weaving together many laws, concepts, and facts into a complete picture.
So a hypothesis is your starting point: one specific, testable idea. If it survives repeated testing, it contributes to the larger body of evidence. Enough confirmed evidence, organized into a coherent explanation, eventually forms a theory.