One way ANOVA
In Exercise
10.2 (BPS, Chapter 10, page 505), different amount of corns were
planted per acre, the yields are shown in worksheet.
In this experiment, we want to test the null
hypothesis that there is no difference between mean yield of corns
in five different ratio of planting. The alternative hypothesis
is that there is some difference, that not all five population
means are equal.
Ho:u1=u2=u3=u4=u5
Ha
: Not all of u1, u2,
u3, u4 and
u5 are equal
We call
analysis of variance F test to solve this problem.
Before proceeding
with the one way ANOVA test, it's important to check to
see if the assumptions of one-way ANOVA are satisfied.
Specifically, the populations are normal with possibly different
means and the same variance. Select Stat->Basic Statistics->
Display Descriptive Statistics to see the summarized data.
Since the ratio of the largest to the smallest standard deviation
is less than 2, it's safe to do one-way analysis of variance.
Select Stat->ANOVA->One-way(Unstacked),
enter all 5 planting ratio in Response box as shown.
Click Graphs
button to specify the boxplot graph to been drawn for these five
population. From side by side boxplot we can visually check the
assumptions.


After clicking
OK, we get the output of ANOVA table. The columns
in this table are labeled Source, DF( degree of
freedom), SS(sum of squares), MS(mean square), F
statistic and P-value. The P-value is 0.736, which
is pretty large. So we failed to reject null hypothesis in this
example.
For information
purpose, the output from the test provides the mean and standard
deviation for each group and plots individual 95% confidence intervals
for the means.