Download Spotfire S+ User`s Guide for UNIX

Transcript
Chapter 6 Statistics
We are interested in two alternative hypotheses: the two-sided
alternative that μ H – μ L = 0 and the one-sided alternative that
μ H – μ L > 0 . To test these, we run the standard two-sample t-test
twice, once with the default two-sided alternative and a second time
with the one-sided alternative hypothesis greater.
1. Open the Two-sample t Test dialog.
2. Type weight.gain in the Data Set field.
3. Select gain.high as Variable 1 and gain.low as
Variable 2. By default, the Variable 2 is a Grouping
Variable check box should not be selected, and the Assume
Equal Variances check box should be selected.
4. Click Apply.
The result appears in the Report window:
Standard Two-Sample t-Test
data: x: gain.high in weight.gain , and y: gain.low
in weight.gain
t = 1.8914, df = 17, p-value = 0.0757
alternative hypothesis: true difference in means is
not equal to 0
95 percent confidence interval:
-2.193679 40.193679
sample estimates:
mean of x mean of y
120
101
The p-value is 0.0757, so the null hypothesis is rejected at the 0.10
level but not at the 0.05 level. The confidence interval is (– 2.2,40.2) .
In other words, we conclude at the 0.05 level that there is no
significant difference in the weight gain between the two diets.
To test the one-sided alternative that μ H – μ L > 0 , we change the
Alternative Hypothesis field to greater in the Two-sample t Test
dialog. Click OK to perform the test and see the output shown below.
240