In: Statistics and Probability
A botanist measures the sepal widths (in cm) of 150 iris flowers, 50 of type iris setosa, 50 of type iris virginica, and 50 of type iris versicolor. The data is noted below. Of interest is whether sepal width differs, on average, among the three iris types. You may assume that, for each iris type, the sample may be treated as a simple random sample.
a. What are the null and alternative hypotheses?
b. What is the value of the test statistic?
c. What is the approximate p-value?
d. State and verify (using plots or descriptive statistics) the additional two assumptions required for the p-value in c) to be valid.
e.
Sepal Width Class
3.5 Iris-setosa
3 Iris-setosa
3.2 Iris-setosa
3.1 Iris-setosa
3.6 Iris-setosa
3.9 Iris-setosa
3.4 Iris-setosa
3.4 Iris-setosa
2.9 Iris-setosa
3.1 Iris-setosa
3.7 Iris-setosa
3.4 Iris-setosa
3 Iris-setosa
3 Iris-setosa
4 Iris-setosa
4.4 Iris-setosa
3.9 Iris-setosa
3.5 Iris-setosa
3.8 Iris-setosa
3.8 Iris-setosa
3.4 Iris-setosa
3.7 Iris-setosa
3.6 Iris-setosa
3.3 Iris-setosa
3.4 Iris-setosa
3 Iris-setosa
3.4 Iris-setosa
3.5 Iris-setosa
3.4 Iris-setosa
3.2 Iris-setosa
3.1 Iris-setosa
3.4 Iris-setosa
4.1 Iris-setosa
4.2 Iris-setosa
3.1 Iris-setosa
3.2 Iris-setosa
3.5 Iris-setosa
3.1 Iris-setosa
3 Iris-setosa
3.4 Iris-setosa
3.5 Iris-setosa
2.3 Iris-setosa
3.2 Iris-setosa
3.5 Iris-setosa
3.8 Iris-setosa
3 Iris-setosa
3.8 Iris-setosa
3.2 Iris-setosa
3.7 Iris-setosa
3.3 Iris-setosa
3.2 Iris-versicolor
3.2 Iris-versicolor
3.1 Iris-versicolor
2.3 Iris-versicolor
2.8 Iris-versicolor
2.8 Iris-versicolor
3.3 Iris-versicolor
2.4 Iris-versicolor
2.9 Iris-versicolor
2.7 Iris-versicolor
2 Iris-versicolor
3 Iris-versicolor
2.2 Iris-versicolor
2.9 Iris-versicolor
2.9 Iris-versicolor
3.1 Iris-versicolor
3 Iris-versicolor
2.7 Iris-versicolor
2.2 Iris-versicolor
2.5 Iris-versicolor
3.2 Iris-versicolor
2.8 Iris-versicolor
2.5 Iris-versicolor
2.8 Iris-versicolor
2.9 Iris-versicolor
3 Iris-versicolor
2.8 Iris-versicolor
3 Iris-versicolor
2.9 Iris-versicolor
2.6 Iris-versicolor
2.4 Iris-versicolor
2.4 Iris-versicolor
2.7 Iris-versicolor
2.7 Iris-versicolor
3 Iris-versicolor
3.4 Iris-versicolor
3.1 Iris-versicolor
2.3 Iris-versicolor
3 Iris-versicolor
2.5 Iris-versicolor
2.6 Iris-versicolor
3 Iris-versicolor
2.6 Iris-versicolor
2.3 Iris-versicolor
2.7 Iris-versicolor
3 Iris-versicolor
2.9 Iris-versicolor
2.9 Iris-versicolor
2.5 Iris-versicolor
2.8 Iris-versicolor
3.3 Iris-virginica
2.7 Iris-virginica
3 Iris-virginica
2.9 Iris-virginica
3 Iris-virginica
3 Iris-virginica
2.5 Iris-virginica
2.9 Iris-virginica
2.5 Iris-virginica
3.6 Iris-virginica
3.2 Iris-virginica
2.7 Iris-virginica
3 Iris-virginica
2.5 Iris-virginica
2.8 Iris-virginica
3.2 Iris-virginica
3 Iris-virginica
3.8 Iris-virginica
2.6 Iris-virginica
2.2 Iris-virginica
3.2 Iris-virginica
2.8 Iris-virginica
2.8 Iris-virginica
2.7 Iris-virginica
3.3 Iris-virginica
3.2 Iris-virginica
2.8 Iris-virginica
3 Iris-virginica
2.8 Iris-virginica
3 Iris-virginica
2.8 Iris-virginica
3.8 Iris-virginica
2.8 Iris-virginica
2.8 Iris-virginica
2.6 Iris-virginica
3 Iris-virginica
3.4 Iris-virginica
3.1 Iris-virginica
3 Iris-virginica
3.1 Iris-virginica
3.1 Iris-virginica
3.1 Iris-virginica
2.7 Iris-virginica
3.2 Iris-virginica
3.3 Iris-virginica
3 Iris-virginica
2.5 Iris-virginica
3 Iris-virginica
3.4 Iris-virginica
3 Iris-virginica
Using a significance level of = 0.05, state your conclusions in the language of the problem.
a)
ho: there is no significant difference in the mean sepal between
three types of iris.
h1: at least one of the mean sepal between three types of iris
differ significantly.
b)
F test statistic = 47.36
c)
p-value = 0.000
d)
With F=47.36, p<5%, I reject ho and conclude that at least one
of the mean sepal between three types of iris differ
significantly.
The same is observed from the interval plot. the sepal width for iris Sentosa is the highest as compared to versi-color and virginica.
assumption 1:
Ho: there is equality of variances between three types of iris. V/s h1: there is in-equality of variances between three types of iris. With Levenes, W=0.65, p>5%, I fail to reject ho and conclude that there is equality of variances between three types of iris.
assumption 2:
There are no outliers in the data set.
procedure 1:
stat -> ANOVA -> one way
response -> sepal width
factor -> class
ok
output 1:
Procedure 2:
stat -> ANOVA -> test with equal variances
response -> sepal width
factor -> class
ok
Output 2: