Hypothesis examples
Open the Repair a weak hypothesis room
This article opens a matching guided lab. Follow the on-screen prompts, make a prediction, and connect the model result back to this topic.
01 // Same format, different subjects
A hypothesis follows the same if-then-because structure no matter what subject it's written for — only the variables and reasoning change. Seeing several worked examples side by side makes the pattern easier to copy for a new experiment than reading the rule alone.
02 // Examples by subject
| Subject | Hypothesis |
|---|---|
| Biology | If a bean plant gets 12 hours of light a day instead of 6, then it will grow taller after two weeks, because more light gives it more energy for photosynthesis. |
| Chemistry | If more baking soda is added to the same amount of vinegar, then the balloon will inflate to a larger size, because more baking soda produces more carbon dioxide gas. |
| Physics | If a pendulum's string is made longer, then its period will increase, because a longer string takes more time to complete each swing. |
| Environmental science | If a stream sample is collected closer to a farm's runoff, then its nitrate level will be higher, because fertilizer runoff carries nitrates into nearby water. |
| Human biology | If a person's resting heart rate is measured after climbing stairs instead of sitting, then it will be higher, because physical activity increases the heart's workload. |
03 // What makes each one testable
- Every example names one specific thing being changed (the independent variable) and one specific thing being measured (the dependent variable) — never both at once, and never something vague like 'health' or 'growth' with no measurement attached.
- The 'because' clause isn't just filler — it should be a real scientific reason, based on something already known, not a restatement of the prediction itself.
- A hypothesis doesn't have to turn out true to be well written. All five examples above are testable and falsifiable whether or not the data ends up supporting them.
04 // Biology example: light and seedling growth
Question: how does daily light duration affect the dry mass of seedlings after four weeks? Hypothesis: if identical seedlings receive six hours rather than three hours of light each day, then their average dry mass will be greater because light supplies energy for photosynthesis. The response is specific and measurable, and the reason links the prediction to a biological process rather than saying simply that plants 'like light.'
The test needs matching species, starting size, soil, pot, water, and temperature. Use several seedlings per condition and record height and leaf condition as additional observations. A result that increases height but not dry mass may show why defining 'growth' matters. If one group is also warmer because it sits closer to a lamp, temperature becomes a possible confound.
05 // Physics example: pendulum length and period
Question: how does pendulum length affect the time for ten swings? Hypothesis: if the pendulum length increases while release angle remains small, then the average period will increase because a longer pendulum takes more time to complete each oscillation. Measure from pivot to the bob's center and time multiple swings to reduce reaction-time error.
Keep the bob, release angle, support, and counting method consistent. Repeat each length several times and calculate the average period. If the results do not show a clear change, inspect the length measurement, timing start point, and whether the string slipped. The prediction applies to the tested range and setup; it should not be generalized to every angle or every pendulum design without additional evidence.
06 // Chemistry example: temperature and dissolving time
Question: how does water temperature affect the time for a fixed mass of sugar to dissolve? Hypothesis: if water temperature rises from 20°C to 50°C, then the measured dissolving time will decrease because water molecules have greater average kinetic energy and interact with sugar more rapidly. Use equal water volumes and the same sugar mass, grain size, cup, and stirring rule.
Define the endpoint, such as the time when no crystals are visible at the bottom, and have the same observer apply it. Repeat each temperature and record actual readings rather than using labels like 'warm' and 'cold.' If stirring is part of the method, use a fixed rate or omit stirring in all trials. The prediction concerns rate under the chosen conditions, not how much sugar can eventually dissolve.
07 // Compare a weak prediction with a testable one
'If I add more fertilizer, then plants will do better because fertilizer is good' is difficult to test: it does not define the fertilizer amount or the meaning of 'better.' A revision might predict leaf count or biomass after a fixed duration and identify how water, soil, light, and plant type will be kept consistent. The reason should describe the nutrient role being tested, not assume that unlimited fertilizer always helps.
Before collecting data, ask whether the hypothesis names the variable, measurable response, direction if known, and rationale. Write down the original wording and date. After results arrive, compare evidence with that prediction and write a new hypothesis only for a later investigation. This keeps examples useful as models of scientific reasoning rather than answer templates to copy without understanding.
08 // Revise weak predictions into testable ones
A statement such as 'plants like sunlight' expresses a general expectation but does not specify a comparison or measurement. A stronger version names the tested light levels, plant type, growth measure, and time period, then gives a reason for the predicted direction. The exact detail depends on the materials, but a reader should be able to identify what data would count as evidence.
Avoid a hypothesis that predicts every possible result, such as 'the plant will grow taller, shorter, or stay the same.' That cannot be meaningfully challenged. If the mechanism is uncertain, write a clear directional or comparative prediction and acknowledge the uncertainty in the rationale. A testable hypothesis can be wrong and still guide a useful investigation.
09 // Questions people ask
Do all hypotheses have to use the exact words 'if, then, because'?
No — that's a template for learning the structure, not a required wording. What matters is that a hypothesis clearly states what's being changed, what's predicted to happen, and the reasoning behind that prediction, however it's phrased.
Can a hypothesis be wrong and still be a good hypothesis?
Yes. A hypothesis is judged by whether it's specific and testable, not by whether it turns out to be correct. A well-written hypothesis that the data doesn't support is still a successful, useful hypothesis.
What's a common mistake in writing a hypothesis?
The most common mistake is being too vague — writing something like 'fertilizer affects plant growth' instead of naming the specific independent variable, the specific measurable outcome, and the direction of the predicted change.