Hypothesis Tests

Discussion of Other Tests

The following descriptions provide an overview of other hypothesis tests available in the Analyst Application.

One-Sample Z-Test for a Mean

In the One-Sample Z-Test for a Mean task, you can test whether the mean of a population is equal to the value you specify in the null hypothesis. This test is appropriate when the population standard deviation or variance is known, and your data are either normally distributed or you have a large number of observations. Generally, a sample size of at least 30 is considered to be sufficient.

The default output from the test includes summary statistics for the selected variable, the Z statistic, and the associated p-value.

One-Sample Test for a Proportion

In the One-Sample Test for a Proportion task, you can test whether the proportion of a population giving a certain response is equal to the proportion you specify in the null hypothesis.

The default output from this test provides a frequency table of responses versus the analysis variable, the observed proportion, the Z statistic, and the associated p-value.

One-Sample Test for a Variance

In the One-Sample Test for a Variance task, you can test whether the variance of a population is equal to the value you specify in the null hypothesis.

The default output from this test includes summary statistics for the selected variable, the chi-square statistic, and the associated p-value.

Two-Sample t-Test for Means

In the Two-Sample t-Test for Means task, you can test whether the means of two populations are equal or, optionally, whether they differ by a specified amount. Two-sample data arise when two independent samples are observed, possibly with different sample sizes. Note that, if the two samples are not independent, the two-sample t-test is inappropriate and you should use instead the Two-Sample Paired t-Test for Means task (see the section "Paired t-test" for more information).

The default output from the test includes summary statistics for the two samples, two t statistics, and the associated p-values. The first t statistic assumes the population variances of the two groups are equal; the second statistic is an approximate t statistic and should be used when the population variances of the two groups are potentially unequal.

Top of Page