The term null hypothesis may cause a bit of confusion for even the brainiest researchers among us. In this post, we’ll cover what null hypotheses are, when you might use them, and how to word these statements in acceptable ways for scientific writing.
To Have an Effect, or Not to Have an Effect?
Research is all about determining if there’s a cause-and-effect relationship between variables. For example, does listening to music have an effect on a person’s running speed?
Researchers manipulate and measure independent variables (e.g., listening to music) and dependent variables (e.g., running speed) to understand the relationships between them. Once the data is gathered, we need to know if the results are statistically significant – in other words, is this result unlikely to have happened solely by chance?
There’s a process for determining whether a relationship between variables is statistically significant. Inferential statistics use a sample to make reasonable guesses about the larger population. In this process, predictions are phrased into two main hypotheses: null and alternative.
A null hypothesis (Ho) makes the prediction that there is no effect on the population.
An alternative hypothesis (H1) makes the claim that there is an effect on the population.
Rejecting the Null Hypothesis or Failing to Reject the Null Hypothesis
Hypothesis testing starts with the assumption that the null hypothesis is true – that there is no effect on the population. It’s similar to the legal assumption of a person’s innocence until proven guilty. The null hypothesis is assumed to be true until the sample provides enough evidence to reject the claim.
After applying a statistical test, if the sample data provides enough evidence against the null hypothesis, then we can reject it. Otherwise, we fail to reject (or we retain) the null hypothesis.
It’s important to use the terms “reject” or “fail to reject” for the null hypothesis, even if it feels unnatural. You might be inclined to use the words “prove” or “accept,” but that’s not the way to go.
Find this useful?
Subscribe to our newsletter and get writing tips from our editors straight to your inbox.
More Null Hypothesis Examples
Null hypotheses will often include negative phrases such as “no effect,” “no change,” or “no difference.” They might also include negative phrases, such as “does not increase” or “does not affect.” Here are some additional examples:
- Does book club participation affect political party affiliation?
2. Does drama/theater experience decrease fear of public speaking?
In Short
Null and alternative hypotheses go hand-in-hand. They’re competing claims used by researchers in statistical tests, and they help determine how statistically significant findings are.
● The null hypothesis claims there’s no effect on the population.
● The alternative hypothesis claims there’s an effect on the population.
● The assumption that there’s no effect (the null hypothesis) comes first, until evidence allows researchers to reject this claim.
● We either reject or fail to reject the null hypothesis.
That’s it! Now that you’re all studied up on null hypotheses, you can approach research writing with confidence. If only the statistics were this easy! If you’d like to submit a free sample of your scientific or technical writing to our editors, we can provide you with valuable editing feedback.