What is the null hypothesis?

What is the null hypothesis?

• A. There is always a detectable effect in experiments.
• B. The experimental treatment is guaranteed to work.
• C. The null hypothesis can never be rejected.
• D. There is no effect or difference; it is tested statistically.

Answer: (D)

In hypothesis testing, you start with a default assumption that there is no effect or difference between groups or treatments. This default is the null hypothesis. The data are analyzed with statistics to see how likely the observed result would be if there really were no effect. If that likelihood (the p-value) is very small, you have evidence against the null and you reject it in favor of an alternative; if not, you don’t reject the null. The statement that best matches this idea is that there is no effect or difference; it is tested statistically. Other options misrepresent how this works: effects aren’t guaranteed or always detectable, and the null can be rejected with sufficient evidence.

In hypothesis testing, you start with a default assumption that there is no effect or difference between groups or treatments. This default is the null hypothesis. The data are analyzed with statistics to see how likely the observed result would be if there really were no effect. If that likelihood (the p-value) is very small, you have evidence against the null and you reject it in favor of an alternative; if not, you don’t reject the null. The statement that best matches this idea is that there is no effect or difference; it is tested statistically.

Other options misrepresent how this works: effects aren’t guaranteed or always detectable, and the null can be rejected with sufficient evidence.