1. This study makes use of random sampling. Because it takes a sample of random Americans to
generalize all Americans and they were not randomly assigned to their explanatory variable
groups.
2. H0 : π 2002 - π 2004 = 0
Ha : π2002 - π 2004 ≠ 0
Null hypothesis
donators in 2004
: The proportion of donators in 2002 equals to the proportion of
Alternative hypothesis : The proportion of donators in 2002 isn't the same as the proportion in
2004. (Two-sided)
3. 2002 proportion
: 0.154
2004 proportion
: 0.172
Difference in proportions
kalimat)
: -0.018Comment tulis jmlh pendonor n tidak (jlsin tabel dlm
4. Here’s the segmented graph
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
2002
Donated blood in previous 12 months
2004
Did not donate blood in previous 12 months
In 2002, the number of people that donated blood in previous 12 months is 210. The
number of people that did not donate is 1,152. The proportion of adults American who donated
blood in previous 12 months is 0.154
In 2004, the number of people that donated blood in previous 12 months is 230. The
number of people that did not donate is 1,106. The proportion of adults American who donated
blood in previous 12 months is 0.172.
The difference in proportion 2002 & 2004 is -0.18
5.
We can suspect that the data which is input matches the segmented bar graph and the two
proportion applet.
6.
a. First we calculate the 2002 proportion then we calculate the 2004 proportion. Note the
difference between the two proportions or just look at the observed difference counted by the
applet. Insert it as the count samples and select the beyond option since the alternative
hypothesis is two-sided. Then click count, the red statement right below it is the approximate pvalue.
b. In this case the p-value is 0.2110
c. The data does not provide enough evidence to allow us to reject the null hypothesis.
7. Yes, the data is approximately normal
Mean = 0.000
SD = 0.014
8. a. Z = -1.26
b. We can see that our observed difference in sample proportion is more than 1 standard
deviation below zero.
c. The p-value according to the standardized statistic is 0.2065. So, just like the simulationbased result, we still don't have enough evidence to reject the null hypothesis.