Professional Documents
Culture Documents
AP Statistics
Experimental Design
Free Response Quiz
Since there are a total of 300 subjects for the study and we need two groups, we could assign a person a number
from 1-300. We would then use a random number generator to decided what group each person goes in. For
instance, 1-150 would be in the treatment group with the filter, while the other 150 individuals would be in the
control group.
b) In this study, the researchers chose to include a group who only drank coffee that was made with the standard
filter. Why is it important to include a control group in this study even though cholesterol levels will be measured
at the beginning and at the end of the study?
There could be possible confounding variables, and the control group would be used as a baseline group for the
results that will be obtained. The control group allows for comparison, whether good or bad, on how the treatment
is affecting the subjects cholesterol.
c) Why would the researchers choose to use only nonsmokers in the study?
Since smoking is probably a variable that could affect cholesterol levels, it is best to control for the variable by
providing all non-smokers in the sample. This ensures that one variable, smoking in our case, is accounted for and
the variation in the treatment is not due to smoking or non-smoking. However, this will only allow us the opportunity
to generalize our findings to non-smokers.
It is not reasonable to generalize these findings because the participants are volunteers and only those strongly
having an opinion will participate.
Blocking Scheme B
Key
Block 1
Block 2
Forest
Forest
a) Which of the blocking schemes, A or B, is better fit for this experiment? Explain your answer.
Blocking scheme A is better because we are blocking on the location with respect to the forest. Plots near the
forest are probably more productive as they probably have exposure to more nutrient rich soil.
b) Even though the students have decided to block, they must randomly assign the varieties of trees to the plots
within each block. What is the purpose of this randomization in the context of this experiment?
The randomization is used to spread the differences out that may occur within each block or more specific each
plot of soil. Some soil may be exposed to more light, while other soil may be exposed to more water and wind. By
randomizing we try to ensure that we are equally spreading this variation out amongst the levels of soil.
50 male
patients
Randomize
100 male
patients
50 male
Treatment 2
No Magnets
Control
patients
We will split the 100 male patients into two treatment groups: A low intensity magnet group and a no magnet group(
control group) or even a placebo group. We will randomize which group each male will go to by flipping a coin: Heads
will be in the magnet group and tails will go to the control/placebo group. We will compare the self-reported pain
levels of both groups at the end of this study.
b) Would you modify the design above, if, instead 100 male patients, there were 50 male and 50 female patients
available for the study? If so, how would you modify your design? If not, why not?
Yes, because males and females may have different levels of pain toleration I would block on gender. A blockrandomize design would look as follows:
Randomize
50 females
Randomize
Block by Gender
100 patients;
50 males
50 females
50 males
25 males
25 males
25 females
25 females
Treatment 1:
Magnets
Treatment 2:
Control
Treatment 1:
Magnets
Treatment 2:
Control
Compare the
self-reported
pain levels of
males
Compare the
self-reported
pain levels of
females
We will block by gender. Treatment groups will be decided by a random number generator for each group. Each
male will be numbered 1-50 and each female will be numbered 1-50. 1-25 will get the treatment while 26-50 will be
placed in the control/placebo group. We will compare the results of each treatment within each block.
West
East
Window
South
In the interior of the house, each window is surrounded by a window box to capture and measure the amount of
heating coming in through the window and to isolate the heat gain for each window.
a) A randomized block experiment will be used to compare the heat gain for the two types (A and B) of windows.
How you would you group the window boxes into blocks? (Clearly indicate your blocks using the window box
numbers.) Justify your choice of blocks.
A blocked-randomized design would look as follows:
Treatment 1 (A)
2 Windows
Block 2 (East)
4 Windows
2,3,4,5
Block 3 (South)
4 Windows
6,7
Block 4 (North)
2 Windows
1,12
Treatment 2 (B)
2 Windows
Randomize
12 Windows
Block by Location
Block 1 (West)
4 Windows
8,9,10,11
Treatment 1 (A)
2 Windows
Treatment 2 (B)
2 Windows
Treatment 1 (A)
1 Window
Treatment 2 (B)
1 Window
Treatment 1 (A)
1 Window
Treatment 2 (B)
1 Window
Block 2: 8,9
Block 3: 2,3
Block 4: 4,5
Block 5: 6,7
Block 6:1, 12
We block because heat gain could be affected by a couple of elements looking at the print. The doors could affect
the heat gain as well as the side of the building because of how the sun rises and sets.
b) For the design in part (a), describe how you would assign window types (A and B) to the numbered window boxes.
Since there are two treatment groups, we could flip a coin or use an even/odd table to determine which group will
get the treatments.