How to Write a Hypothesis

It doesn’t matter if you’re working on a third grade science fair project or completing a Ph.D. thesis. Any science student must learn to write a good hypothesis. That’s a crucial part of the scientific method, which forms the backbone of academic experimentation. In some ways, drafting a hypothesis is the most creative part of a scientific experiment. With a hypothesis, the experimenter mixes existing knowledge with his or her own experience to create a testable question to answer through scientific experimentation. Although the scientist begins with an educated guess about what the results of the experiment might be, he or she doesn’t know the outcome. Writing in Science Magazine about his experience judging science fair projects, Adam Ruben, Ph.D., explains, “If there’s one lesson I hope the kids take from the science fair, it’s…A science experiment doesn’t have a ‘right answer.'” 

A hypothesis is a testable statement or question that offers a framework to understand the natural world. The hypothesis statement fuels the research design by providing the researcher with a thesis around which the experiment or research project revolves. As Encyclopedia Britannica expounds, “The two primary features of a scientific hypothesis are falsifiability and testability, which are reflected in an ‘If…then’ statement summarizing the idea and in the ability to be supported or refuted through observation and experimentation.” An “if…then” statement refers to the phrasing of a traditional hypothesis, which offers a theory on what may happen if a certain set of procedures are followed. 

For example:

If men take vitamin E supplements before age thirteen, then they will be less likely to develop dementia in old age. 

That hypothesis could be tested by following a population of men from birth to old age and recording their use of vitamin E supplements and dementia diagnoses.

Your writing, at its best

Compose bold, clear, mistake-free, writing with Grammarly's AI-powered writing assistant

Types of Hypotheses

Hypothesis testing may involve data collection in many forms, through experimentation, statistical analysis, or some combination of both. Any testable hypothesis must have at least one independent variable (cause) and one dependent variable (effect). In the example hypothesis above, the independent variable would be the use of vitamin E supplement taken before age thirteen, and the dependent variable would be a positive dementia diagnosis. In order to be “testable,” experiments or scientific research projects require more than one possible outcome.

  • Simple Hypothesis—This type of hypothesis involves a simple statement with only one dependent and one independent variable. 
  • Complex Hypothesis—Complex hypotheses describe the relationship between multiple dependent variables and multiple independent variables. 
  • Logical Hypothesis—A logical hypothesis offers a possible explanation for a relationship between variables that is too difficult to test experimentally. A logical hypothesis should still be testable, even if the test cannot yet be performed. 
  • Statistical Hypothesis—This type of hypothesis involves quantitative research that substitutes a sample population for an entire population that would be too large to study.
  • Empirical Hypothesis—This type of hypothesis forms the basis for an empirical experiment. 
  • Null Hypothesis—A null hypothesis posits that there is no relationship between two variables. 
  • Alternative Hypothesis—An alternative hypothesis seeks to disprove a null hypothesis by suggesting a new theory to explain the relationship between the variables.

Drafting a Hypothesis

The first step, before you ever write a hypothesis, is to conduct a literature review, where you read academic writing on the topic. If you want to conduct an experiment that deals with soil health and plant growth, you’ll first need to understand the studies that have already been published on the topic of soil health. Identify any gaps in the existing literature. Identify a question that remains unanswered, and try to find a creative way to answer the question that accounts for, or doesn’t conflict with, the existing data. 

At that point, you’ll be able to write a few research questions. Usually, this is a good time to identify possible dependent and independent variables. For example, you might ask yourself, “Does the presence of a particular mineral in the soil (X) influence the growth of the plant (Y)?”

Next, you’ll want to turn your research question into either a research hypothesis or an experimental hypothesis by adding your proposed answer (or prediction) of a possible outcome that may result when the relationship between the variables is tested. In an experimental hypothesis, you would use empirical study to prove or disprove the hypothesis. For a research hypothesis, you would rely on statistical information about a well-defined population to test your prediction. 

According to the Canada-Wide Science Fair, scientific understanding accounts for 50% of a science fair score. Along with other criteria that prove scientific understanding, judges for this science fair look for:

  • A hypothesis or project design that is clear and well stated based on reading, study, and/or observation. The depth of study is a factor here.
  • An experimental procedure that is effective in testing the hypothesis, or an innovative design that is an effective solution to the problem posed, or a study designed to produce significant new insights.

To summarize, whether you’re working on a science fair project or a Ph.D. thesis, a good hypothesis shows a thorough understanding of the current scientific knowledge. An experimenter or researcher should not repeat experiments or research projects that already have a known outcome. If a scientist hopes to prove an alternative hypothesis, the experimental or research design should be altered from previous studies to allow for new insights. In addition, the research or experiment must be thoughtfully designed to answer the question posed by the hypothesis. If the data collection procedure does not seem as though it will prove or disprove the hypothesis at hand, the scientist should revise the procedure before beginning the experiment or research.