Research and statistical analysis help us understand how the world works, how effective medications and treatments are, what influences our health, the best approach for business practices, and much more. It’s a vital aspect of many careers, including medical research, financial analysis, market research, and behavioral health.

An important part of scientific research is the null hypothesis. Understanding how null hypotheses work can support your analytical, research, problem-solving, and critical thinking. Below, we discuss what a null hypothesis is and how they’re tested, followed by a null hypothesis example.

**What is a null hypothesis?**

A null hypothesis is used in research and statistical analysis. Also referred to as the default hypothesis, it’s a type of assumption indicating there’s no significant difference between samples being analyzed. In simpler terms, a null hypothesis proposes that there’s no meaningful relationship or difference between two things. The things, or items being assessed, can be tangible or abstract.

The goal is for researchers to work to disprove, reject, or nullify the null hypothesis. To do so, they come up with what’s referred to as an alternative hypothesis.

**Alternative hypothesis vs. null hypothesis**

Where a null hypothesis indicates there is no significant difference between two items or a set of figures, an alternative hypothesis suggests there is a difference between the two items or figures. The alternative hypothesis then contradicts the null hypothesis.

Researchers, analysts, and statisticians create an alternative hypothesis, and in some instances, more than one, to disprove the null hypothesis. The alternative hypothesis is one that the researchers, analysts, or statisticians believe to be true.

**Why is the term “null” used?**

“Null” in the context of null hypothesis implies it’s a commonly accepted concept that individuals work to nullify. It doesn’t imply that the statement is null in and of itself but that the goal is to nullify the hypothesis statement. Yes, it can be confusing!

**Why can’t I just prove the alternative hypothesis?**

Depending on your field, you might be able to get away with not testing the null hypothesis, but it’s not recommended. In scientific research, you must test the null hypothesis. Not doing so would be considered poor practice.

In scientific research, several processes are used to prove or disprove theories, ensuring that new hypotheses have zero flaws. If you choose not to test the null hypothesis and go straight to testing the alternative hypothesis, your work, at a minimum, likely won’t be taken seriously, and you’ll be setting yourself up for failure before you begin.

**How is a null hypothesis tested?**

Researchers and analysts assume a null hypothesis is accurate and true until there is enough statistically significant evidence to suggest otherwise. To test hypotheses, researchers:

- Identify the question.
- Identify the null and the alternative hypotheses.
- Develop a plan for research and data collection.
- Test the data and research.
- Interpret the test results.

In research, p-values are used as evidence that goes against a null hypothesis. The smaller the p-value, the stronger the statistical data and research that disproves the null hypothesis. Significant tests are used in research to showcase confidence in a null hypothesis and to investigate if the data is due to chance.

Based on the results, researchers will either:

**Fail to reject the null hypothesis:**This occurs when the p-value is greater than the level of significance, indicating that the results of the experiment are not statistically significant. Errors in data, insufficient evidence, or other reasons can lead to experimenters failing to reject the null hypothesis. A failure to reject the null hypothesis doesn’t indicate that the testing failed to find answers but that further testing is needed to determine if there’s a relationship between the variables being tested.**Reject the null hypothesis:**This occurs when the p-value is less than or equal to the level of significance, indicating that the results support the alternative hypothesis. Researchers can reject the null hypothesis since the data is statistically significant.

**Null Hypothesis Example**

Here is a null hypothesis example to help you develop your own null hypothesis.

**Step 1:** Develop your question.

- Does a reduction in calories impact a person’s weight?

**Step 2:** Develop your two hypotheses.

**Null hypothesis:**A reduction in calories has no effect on a person’s weight**Alternative hypothesis:**A reduction in calories does affect a person’s weight**According to the null hypothesis:**The number of pounds lost does not differ between people that don’t reduce their calorie intake and those that do reduce their calorie intake.

**Step 3:** Develop a plan for data collection.

- We will use data collected through survey form from two random sample groups, Group A and Group B.

**Step 4: **Test the Null Hypothesis.

- Collect data from a sample of 500 people, Group A, that reduce their calorie intake by 500 calories a day for a month.
- Collect data from a sample of 500 people, Group B, that do not reduce their calorie intake by 500 calories a day for a month, all other variables remaining the same.
- Identify the number of people that lost weight, stayed the same, and gained weight from Group A.
- Identify the number of people that lost weight, stayed the same, and gained weight from Group B.
- Analyze the results to assess whether to reject your null hypothesis.

**Calculating and interpreting test results — Some final thoughts**

It’s important to keep in mind that, when conducting research, there are many variables to consider that can influence results. Dependent and independent variables are commonly discussed, though other variables can also impact the results of research, including moderator, control, confounding, and extraneous variables that need to be accounted for. For example, items that could influence the results of the weight example above include culture, ethnic background, the percentage of male vs. female participants, participants’ honesty, participants’ consistency, and so on, which would fall into one of the variable categories for consideration when interpreting results.

Different research methods can be used based on what’s being tested and the goal of testing. One key goal of scientific research is to confirm validity and reliability. Reliability refers to how frequently you get the same result with a test, and validity refers to whether the test assesses what it’s meant to assess.

Now you have some insight into what a null hypothesis is and why it’s used. Try coming up with a few on your own to stretch your research and critical thinking skills.