2 Variancesoverview how to example data see also
2 VariancesStat > Basic Statistics > 2 Variances
The 2 Variances procedure performs hypothesis tests and computes confidence intervals for the ratios between two populations' variances and standard deviations; a ratio of 1 suggests equality between populations. Use this test to determine if one treatment condition has more variability than the other.
For example, a lumber distributor wants to compare the variation of beam lengths that are cut by two different sawmills. The distributor measures the length of the beams from each sawmill to determine whether the consistency of the beam lengths differs.
By default, Minitab uses two tests: Bonett's test and Levene's test. For each test, the null hypothesis states that the two variances are equal (H0: s21 / s22 = 1). The alternative hypothesis can be left-tailed (H1: s21 / s22 < 1), right-tailed (H1: s21 / s22 > 1), or two-tailed (H1: s21 / s22 ≠ 1). Optionally, test ratios other than 1 (equality) can be specified.
Both samples are in one column: Choose if the sample data are in a single column, and are differentiated by sample IDs in a second column.
Samples: Enter the column that contains the data for both samples.
Sample IDs: Enter the column that indicates which sample each observation belongs to.
Each sample is in its own column: Choose if the data for each sample is in a different column.
Sample 1: Enter the column that contains the data for the first sample.
Sample 2: Enter the column that contains the data for the second sample.
Sample standard deviations: Choose to enter summary values for the sample size and standard deviation of each sample. For summary data, only the F-test can be performed. The F-test is accurate only for normally distributed data.
Sample size: Enter the sample size for each sample.
Standard deviation: Enter the standard deviation for each sample.
Sample variances: Choose to enter summary values for the sample size and variance of each sample. For summary data, only the F-test can be performed. The F-test is accurate only for normally distributed data.
Sample size: Enter the sample size for each sample.
Variance: Enter the variance for each sample.