Chapter 8 Minitab Statistical decision making Confidence intervals

Chapter 8 Minitab: Statistical decision making

Confidence intervals for the population mean • Choose Basic Statistics from the Stat menu and 1 -Sample t from the sub -menu.

• In the top right of the new window select Summarized data.

• For Example 8. 6 enter 40 as the Sample size, 3. 47 as the Mean, and 0. 82 as the Standard deviation. • Click Options.

• In the Options window type 90. 0 in the space to the right of Confidence level. • Click OK then OK in the One-Sample t for the Mean window.

• The confidence interval appears in the session window under 90% CI for .

Hypothesis tests for differences between population means • Select Basic Statistics from the Stat menu then 2 -Sample t from the sub-menu.

• In the new window select Summarized data from the menu to the top right. • For the pooled estimate of the standard error tick Assume equal variances.

• Type in the Sample size, Sample mean and Standard deviation of the ‘own brand’ sample in Example 8. 19 (14, 34. 9 and 1. 4) in the Sample 1 boxes and the figures for the premium sample (17, 33. 4 and 1. 1) in the Sample 2 boxes.

• Click Options and check that 0 is in the space beside Test difference. • Change Alternative hypothesis to Difference > hypothesized difference. • For the pooled estimate of the standard error tick Assume equal variances.

• Click OK then OK in the Two-Sample t for the Mean window. • The output includes the test statistic, T-value and the probability that the test statistic occurs if the null hypothesis is true, the P-value.

Paired t tests • Enter the two sets of data in Example 8. 20 into C 1 and C 2.

• Choose Basic Statistics from the Stat menu then Paired t from the sub-menu.

• Check that Each sample is in a column is selected. • Type C 2 for the location of Sample 1 and C 1 for the location of Sample 2.

• Click Options, type 8 as the Hypothesized difference and change the Alternative hypothesis to Difference < hypothesized difference. • Click OK then OK in the Paired t for the Mean window.

• The output in the session window includes the test statistic, Tvalue and the probability that the test statistic occurs if the null hypothesis is true, P -value.