8.4 Distributions Required for a Hypothesis Test
LEARNING OBJECTIVES
- Identify the distribution required to conduct a hypothesis test.
Earlier in the course, we discussed sampling distributions: the sampling distribution of the sample mean and the sampling distribution of the sample proportion. These distributions play a role in hypothesis testing.
If the hypothesis test is on a population mean, we use the distribution of the sample means in the hypothesis test. As we learned previously, the distribution of the sample means follows a normal distribution if the population the sample is taken from is normal or if the sample size is large enough ([latex]n \geq 30[/latex]). For a hypothesis test on a population mean we use a normal distribution when the population standard deviation is known or a [latex]t[/latex]-distribution when the population standard deviation is unknown.
If the hypothesis test is on a population proportion, we use the distribution of the sample proportions in the hypothesis test. As we learned previously, the distribution of the sample proportions follows a normal distribution if [latex]n \times p \geq 5[/latex] and [latex]n \times (1-p) \geq 5[/latex] or a binomial distribution if one of [latex]n \times p \lt 5[/latex] or [latex]n \times (1-p) \lt 5[/latex]. For a hypothesis test on a population proportion we use either a normal distribution or a binomial distribution, depending on which of the above conditions is met.
Assumptions
When we perform a hypothesis test of a single population mean[latex]\mu[/latex] and the population standard deviation is known, we take a simple random sample from the population. We use a normal distribution, assuming the population is normal or the sample size is large enough ([latex]n \geq 30[/latex]). The [latex]z[/latex]-score we need is the [latex]z[/latex]-score from the distribution of the sample means: [latex]\displaystyle{z=\frac{\overline{x}-\mu}{\frac{\sigma}{\sqrt{n}}}}[/latex].
When we perform a hypothesis test of a single population mean[latex]\mu[/latex] and the population standard deviation is unknown, we take a simple random sample from the population. We use a [latex]t[/latex]-distribution, assuming the population is normal or the sample size is large enough ([latex]n \geq 30[/latex]). We use the sample standard deviation to approximate the population standard deviation. The [latex]t[/latex]-score we need is: [latex]\displaystyle{t=\frac{\overline{x}-\mu}{\frac{s}{\sqrt{n}}}}[/latex].
When we perform a hypothesis test of a single population proportion [latex]p[/latex], we take a simple random sample from the population. We use a normal distribution when [latex]n \times p \geq 5[/latex] and [latex]n \times (1-p) \geq 5[/latex]. In this case, the [latex]z[/latex]-score we need is the [latex]z[/latex]-score from the distribution of the sample proportions: [latex]\displaystyle{z=\sqrt{\frac{p \times (1-p)}{n}}}[/latex]. Otherwise, we use a binomial distribution when one of [latex]n \times p \lt 5[/latex] or [latex]n \times (1-p) \lt 5[/latex].
Concept Review
In order for a hypothesis test’s results to be generalized to a population, certain requirements must be satisfied.
Testing a population mean:
- Population standard deviation is known: use a normal distribution, assuming the population is normal or [latex]n \geq 30[/latex].
- Population standard deviation is unknown: use a [latex]t[/latex]-distribution, assuming the population is normal or [latex]n \geq 30[/latex].
Testing a population proportion:
- Use a normal distribution when [latex]n \times p \geq 5[/latex] and [latex]n \times (1-p) \geq 5[/latex].
- Use a binomial distribution when at least one of [latex]n \times p \lt 5[/latex] or [latex]n \times (1-p) \lt 5[/latex].
Attribution
“9.3 Distribution Needed for Hypothesis Testing“ in Introductory Statistics by OpenStax is licensed under a Creative Commons Attribution 4.0 International License.