A hypothesis is a testable statement that predicts the relationship between two or more variables. Whether you’re conducting a scientific experiment, a psychology study, or a social science survey, writing a strong hypothesis is a crucial first step in the research process. This guide explains how to write a hypothesis, including what it is (and is not), the different types of hypotheses, how to recognize a good one, and a step-by-step method for creating your own.
Key Points: How to Write a Hypothesis
- A hypothesis is a testable, falsifiable prediction about a relationship between variables.
- The independent variable is what you change; the dependent variable is what you measure.
- A good hypothesis is specific, clear, and based on prior knowledge or observation.
- There are several types of hypotheses, including null, alternative, and statistical hypotheses.
- Writing a hypothesis follows a step-by-step process that begins with identifying a research question.
- Understanding common mistakes can help avoid vague, unfalsifiable, or biased hypotheses.
What Is a Hypothesis?
A hypothesis is a proposed explanation or prediction that can be tested through investigation and experimentation. It suggests how one variable (the independent variable) affects another (the dependent variable).
Examples
- Simple Hypothesis: If plants receive more sunlight (independent variable), then they will grow taller (dependent variable).
- Null Hypothesis: There is no difference in plant growth between those that receive sunlight and those that don’t.
- Alternative Hypothesis: Plants exposed to more sunlight grow significantly taller than those with less light.
What a Hypothesis Is Not
- ❌ A question: “Do plants grow faster with more sunlight?”
- ❌ A fact or widely accepted truth: “Water boils at 100°C at sea level.”
- ❌ A vague idea: “I think plants might do better outside.”
A hypothesis must be testable and measurable, making it suitable for scientific inquiry.
Why a Good Hypothesis Is Important
A well-constructed hypothesis:
- Gives your experiment a clear purpose and direction.
- Helps determine the structure of your experiment or study.
- Ensures your research is falsifiable, a key criterion in science.
- Forms the foundation of statistical testing and data analysis.
- Allows replication and validation by other researchers.
Types of Hypotheses (with Examples)
There’s more than one way to frame a hypothesis. The type you choose depends on your research design, the number of variables involved, and whether you’re conducting experimental, observational, or statistical research. Understanding the different types of hypotheses helps you communicate your predictions clearly and choose the right structure for your study. Here are the most common types, complete with examples and their appropriate use cases.
| Type | Description | Example |
|---|---|---|
| Simple Hypothesis | Predicts the relationship between one IV and one DV | If water is heated, then it boils faster. |
| Complex Hypothesis | Predicts relationships involving two or more IVs or DVs | If water is heated and stirred, then it boils faster and more uniformly. |
| Null Hypothesis (H₀) | Predicts no relationship or difference | There is no effect of caffeine on memory. |
| Alternative Hypothesis (H₁) | Opposes the null, predicts a relationship or difference | Caffeine improves memory performance. |
| Working Hypothesis | A provisional idea to guide initial research | A new fertilizer may increase crop yield. |
| Logical Hypothesis | Based on deductive reasoning or theory | If all metals expand when heated, iron should expand when heated. |
| Empirical Hypothesis | Supported by experimental data | Test results show that iron expands when heated. |
| Statistical Hypothesis | Framed for testing using statistics | H₀: μ₁ = μ₂; H₁: μ₁ ≠ μ₂ |
Hypothesis Examples by Subject Area
Helps readers see relevant, discipline-specific applications:
| Subject | Example |
|---|---|
| Biology | If salt concentration increases, then cell mass will decrease due to osmosis. |
| Chemistry | If acid concentration increases, then the rate of reaction will increase. |
| Physics | If the angle of a ramp increases, then a rolling object will accelerate faster. |
| Psychology | If people view positive images, then their mood scores will improve. |
| Environmental Science | If fertilizer runoff enters a stream, then algae growth will increase. |
Hypothesis vs Prediction vs Inference
Many students (and even some educators) confuse these related but distinct terms.
- Hypothesis: A testable statement about the relationship between variables.
- Prediction: A specific expected outcome based on the hypothesis.
- Inference: A logical conclusion based on observations and evidence.
Example:
- Observation: The plant in the sunniest spot grew the tallest.
- Inference: Sunlight promotes plant growth.
- Hypothesis: If a plant receives more sunlight, then it will grow taller.
- Prediction: The plant placed in full sun will grow 5 cm taller than the shaded one in 2 weeks.
Characteristics of a Good Hypothesis
✔ Testable – Can be evaluated through observation or experimentation
✔ Falsifiable – Can be proven false by evidence
✔ Clear and concise – Easy to understand and unambiguous
✔ Specific – Defines the variables and population
✔ Based on prior knowledge or theory – Grounded in observation, logic, or literature
✔ Measurable – Involves quantifiable outcomes
✔ Ethical – It’s not about what you can test, but what you should test!
✘ Characteristics of a Poor Hypothesis
- Vague: “Things will change under different conditions.”
- Not testable: “Aliens influence the stock market.”
- Biased or leading: “This brand is the best because it’s superior.”
- Unethical or dangerous
How to Write a Hypothesis (Step-by-Step)
1. Ask a Research Question
Start with a question you want to answer.
Example: Does music affect concentration?
2. Do Background Research
Review scientific literature or past studies to understand what is already known.
3. Identify Variables
- Independent Variable (IV): What you change (e.g., type of music)
- Dependent Variable (DV): What you measure (e.g., concentration test score)
4. Formulate Your Hypothesis
Use an if…then or statement format.
If students listen to classical music, then their test scores will improve.
5. Write the Null Hypothesis
This is the default position for statistical testing.
There is no effect of classical music on test performance.
6. Check It Against the Criteria
Ensure your hypothesis is clear, testable, and measurable. Now you’re read to test the hypothesis with an experiment!
How to Write a Hypothesis Free Resources
Download and print free hypothesis resources:
- Steps of a Hypothesis Graphic: PNG | PDF
- Identify the Variables Worksheet (PDF): Worksheet | Answer Key
Rewriting Poorly Written Hypotheses
Even experienced students sometimes struggle to craft well-structured, testable hypotheses. Often, the issue lies in vague language, lack of clarity about variables, or making untestable claims. Practice identifying common problems in poorly written hypotheses and see how to rewrite them into strong, scientific statements.
Examples of Poorly Written Hypotheses (and Why They Fail)
- “I think plants like sunlight.”
- ❌ Why it’s poor: It’s subjective (“I think”), vague (“like”), and not testable.
- ✅ Rewritten: If a plant receives more sunlight, then it will grow taller than a plant with less sunlight.
- “Music is good for concentration.”
- ❌ Why it’s poor: “Good” is subjective and undefined; it doesn’t identify variables.
- ✅ Rewritten: If students listen to instrumental music while studying, then their test scores will be higher than students who study in silence.
- “The weather affects people.”
- ❌ Why it’s poor: It’s too general and doesn’t specify which aspect of the weather or which effects.
- ✅ Rewritten: If the temperature exceeds 90°F, then individuals will report higher levels of irritability on a mood survey.
- “More screen time is bad.”
- ❌ Why it’s poor: “Bad” is subjective and doesn’t define a measurable outcome.
- ✅ Rewritten: If teenagers spend more than 4 hours per day on screens, then their sleep duration will be shorter compared to those with less screen time.
- “Energy drinks are dangerous.”
- ❌ Why it’s poor: “Dangerous” is too broad and lacks a measurable effect.
- ✅ Rewritten: If participants consume one energy drink, then their heart rate will increase more than participants who drink water.
Tips for Rewriting Hypotheses
- Be Specific: Name the independent and dependent variables.
- Be Measurable: Use quantifiable terms (e.g., height, time, score).
- Avoid Subjectivity: Stay away from vague words like “better,” “good,” or “worse.”
- Avoid First Person: Keep hypotheses objective and impersonal.
- Make It Testable: You should be able to design an experiment to support or refute it.
Common Misconceptions
- Myth: A hypothesis must be correct to be useful.
Truth: A disproven hypothesis still advances knowledge. - Myth: A hypothesis must be proven.
Truth: Hypotheses are tested, not proven. Evidence supports or refutes them. - Myth: Only scientists use hypotheses.
Truth: Hypotheses are used in business, education, psychology, and everyday life. - Myth: You must use “if…then” wording.
Truth: It’s common, but not required.
FAQs About How to Write a Hypothesis
Q: What’s the difference between a hypothesis and a theory?
A: A hypothesis is an initial testable idea. A theory is a well-substantiated explanation backed by a large body of evidence.
Q: Can a hypothesis be a question?
A: No. A hypothesis answers a question, but it must be a statement.
Q: Do all experiments need a hypothesis?
A: Most do, but some exploratory research may not begin with a formal hypothesis.
Q: What is the role of the null hypothesis in statistics?
A: It provides a baseline that statistical tests can reject or fail to reject based on the data.
Q: How many hypotheses can I test at once?
A: You can test multiple hypotheses, but each should be clear and independently testable.
References
- Hempel, C. G. (1952). Fundamentals of Concept Formation in Empirical Science. Chicago: University of Chicago Press. ISBN: 978-0226575971.
- Hilborn, Ray; Mangel, Marc (1997). The Ecological Detective: Confronting Models with Data. Princeton University Press. ISBN 978-0-691-03497-3.
- Hinkelmann, Klaus; Kempthorne, Oscar (2008). Design and Analysis of Experiments. Volume I: Introduction to Experimental Design (2nd ed.). Wiley. ISBN 978-0-471-72756-9.
- Krauss, Alexander (2024). “Redefining the scientific method: As the use of sophisticated scientific methods that extend our mind”. PNAS Nexus. 3 (4): 112. doi:10.1093/pnasnexus/pgae112
- Shields, Patricia M.; Tajalli, Hassan (2006). “Intermediate Theory: The Missing Link in Successful Student Scholarship”. Journal of Public Affairs Education. 12 (3): 313–334. doi:10.1080/15236803.2006.12001438