You are on page 1of 5

The following is a modified version of the Chapter 4 Investigation.

Complete the following problems in place


of the Chapter 4 Investigation from your textbook.

All problems are worth 1 point each unless otherwise stated.

Social psychologists throughout the years have shown that an aggression-sexuality link exists not only in
various animal species but also in humans. Dutton and Aron (Journal of Personality and Social Psychology,
1974) set out to show that a more general link exists in humans, namely emotional arousal of all kinds and
sexual attraction. They set up their study to compare men in a high emotional arousal situation to men in a low
emotional arousal situation.

Researchers wanted to test the notion that an attractive female is seen as more attractive by males who
encounter her in a fear-arousing situation than by males who encounter her in a non-fear-arousing situation. In
the high emotional arousal group, men deemed to be between the ages of 18 and 35 who crossed a suspension
bridge 230 feet above rocks and shallow rapids in Capilano Canyon, North Vancouver, British Columbia,
Canada, and were not accompanied by a female were approached on the bridge by an attractive female inter-
viewer. The same interviewer also approached men who fit the same criteria but crossed a solid wood bridge 10
feet above a rivulet that ran into the main river. Both groups of men were interviewed on the bridge and were
told that the interview was for a psychology class project on the effects of exposure to scenic attractions on
creative expression. The men filled out a short questionnaire after which the interviewer wrote her phone
number down on a slip of paper and said that if they were interested in the results of the experiment they could
call her. The researcher talked to 18 men on the suspension bridge and 16 men on the wooden bridge.

Step 1: Ask a research question. The researcher is now interested in seeing whether a higher proportion of
men in the group that crossed the Capilano Canyon bridge will call her versus the proportion of calls she will
receive from the group that crossed the solid wooden bridge.

Step 2: Design a study and collect data.

1. Is this an observational study or an experiment? Select one.


A. Observational study because the researchers observed whether the men would call or not.
B. Experiment because the researchers decided which bridges to use for the study.
C. Observational study because the researchers did not tell the men which bridge to cross.
D. Experiment because this is not a random sample of men.

2. What are the observational units in the study? Select one.


A. Men aged 18 to 35 who were crossing a bridge without a female companion
B. Bridges in British Columbia, Canada
C. The female interviewer
D. Students in a psychology class

3. We have two variables: bridge type and whether the subject called or not. Which of these variables is
explanatory and which is response?

You will learn in Chapter 5 how to state null and alternative hypotheses for studies like this. Here they are for
this study:

H0: The long-run proportion of men who will call the researcher is the same under both conditions.
Ha: The long-run proportion of men who will call the researcher is higher for those who cross the
suspension bridge compared to those who cross the solid wood bridge.

Of the 18 men on the suspension bridge that accepted her phone number, 9 called her. Of the 16 men on the
solid wooden bridge that accepted her phone number, 2 called her.

Step 3: Explore the data

4. Fill in the two-way table with the data from the study.
Suspension Bridge Wooden Bridge Total
Subject called interviewer 11
Subject did not call interviewer 23
Totals 18 16 34

5. Create a segmented bar graph to display these data, either by hand or using Tableau. Put the explanatory
variable on the x-axis and segment by the response variable. See page 263 of your textbook for a review
of segmented bar graphs.

6. What proportion of men on the suspension bridge called the interviewer?

7. What proportion of men on the wooden bridge called the interviewer?

Step 4: Draw inferences. In Chapter 5, we will learn how to use the 3S strategy to help us investigate how
much evidence the sample data provide to support our conjecture that the long-run probability of men who call
is greater for those on the suspension bridge than for those on the wooden bridge. As a result of the simulation
analysis method used in Chapter 5 we received a p-value of 0.023.

8. [2 pts] Use the null and alternative hypotheses stated above and the p-value of 0.023 to write a
conclusion about strength of evidence in this study. (Hint: Even though the study design is different, the
way you interpret p-values in terms of the null and alternative hypothesis is still the same.)

9. How would the p-value have changed if the alternative hypothesis was stated as: Alt: The long-run
proportion of men who will call the researcher among those who cross the suspension bridge is different
from that among those who cross the solid wood bridge. Select one.
A. It would stay the same since the data has not changed.
B. We cannot predict how the p-value would change in this situation.
C. It would approximately be cut in half.
D. It would approximately double.

10. [2 pts] In studies comparing two groups, the parameter of interest is often stated as the true difference in
the long-run proportions in the two groups. In this study, this is the long-run difference in the
proportions of men calling the researcher comparing the men crossing the Capilano suspension bridge to
the men crossing the wooden bridge. A 95% confidence interval for this parameter is (0.025, 0.725).
Explain how the fact that the confidence interval does not include 0 corresponds to the size of the p-
value obtained in #9.

Step 5: Formulate conclusions.

11. Generalization. Now, lets step back a bit and think about the scope of our inference. Does this study
represent all people? How about all men? Is there any population to which we can generalize our
conclusion? Explain.
12. Causation. What about cause-and-effect conclusions? Are we able to conclude that the cause of the
difference between the two groups of men was the type of bridge they crossed? Explain.

Step 6: Look back and ahead. The researchers acknowledged the fact that there were confounding variables
involved in this study. They tried to minimize these by doing other studies to try to answer the same research
question. In another study, they again placed an interviewer on a suspension bridge. After selecting a man that
fit the criteria to be in the study, the researcher flipped a coin to decide whether to interview the subject on the
suspension bridge just as before or at least 10 minutes after they crossed the suspension bridge when they were
walking or sitting in the nearby park.

13. Select the confounding variable the researchers are attempting to eliminate in this second study.
A. Whether the subject calls the interviewer or not
B. Length of the questionnaire
C. Personality traits like outgoingness of the subject
D. Location of the interview

14. [2 pts] Explain how your answer to problem #12 meets the definition of a confounding variable.

Lets see how this second study turned out. Of those on the suspension bridge, 13 of the 20 men called the
female interviewer. Of those that had already crossed the bridge, 7 out of 23 men called the female interviewer.

15. In this study, what proportion of those on the suspension bridge called the interviewer?

16. In this study, what proportion of those who had already crossed the bridge called the interviewer?

The hypotheses are modified slightly for this new study:

H0: The long-run proportion of men who will call the researcher is the same under both conditions

Ha: The long-run proportion of men who will call the researcher is the higher among men who are
approached by the female while on the suspension compared to that among men who are approached by
the female after they have recently crossed the bridge.

17. [2 pts] Write a conclusion about the strength of evidence in this new study based on obtaining a p-value
of 0.024.

18. [2 pts] Generalization. Have the researchers addressed the concerns about generalization (discussed in
#10) with this new study? Explain.

19. [2 pts] Causation. Is a cause-effect conclusion possible in this new study? Why or why not?
The following questions will review your understanding of Chapter 4 but are not from your textbook.

20. Reuters (June 24, 1997) reported on a study published in the Journal of the American Medical
Association (1997, vol. 277, pp. 19401944) in which researchers recruited 276 volunteers (aged 18 to
55) and used nose drops to infect them with a cold virus. The volunteers were then quarantined and
observed to see whether they came down with a cold or infection. They also answered questions about
their social lives, including the number of different types of contacts they had (family, work,
community, religious groups, and so on). Reuters reported that those with one to three types of social
contacts were 4.2 times as likely to come down with cold symptoms and signs of infection compared to
those with six or more contacts.
a. Is this an observational study or an experiment? Select one.
A. Observational study because number of types of social contacts was not randomly assigned.
B. Observational study because the sample is from volunteers.
C. Observational study because researchers observed whether the volunteers came down with a cold
or not.
D. Experiment because the researchers used nose drops to infect volunteers.

b. [2 pts] Identify both the response variable and the explanatory variable in this study.

21. An article in Science magazine (Service, 1994) discussed a study comparing the health of 6000
vegetarians and a similar number of their friends and relatives who were not vegetarians. The
vegetarians had a 28% lower death rate from heart attacks and a 39% lower death rate from cancer, even
after the researchers accounted for differences in smoking, weight, and social class. In other words, the
reported percentages were the differences remaining after adjusting for differences in death rates due to
those factors.
a. [2 pts] Is this an observational study or an experiment? Explain your answer.

b. [2 pts] Identify a potential confounding variable and explain why it meets the definition of a
confounding variable.

c. On the bases of this information, can we conclude that a vegetarian diet causes lower death rates
from heart attacks and cancer?
A. Yes, because the difference in death rates was so high between the two diets.
B. Yes, because researchers accounted for smoking, weight, and social class in their analysis.
C. No, because it does not say this is a random sample of vegetarians and non-vegetarians.
D. No, because diet was not randomly assigned.

(Investigation continued on next page)


Twenty students agreed to participate in a study on colds. Ten were randomly assigned to receive vitamin C,
and the remaining 10 received a tablet that looked and tasted like vitamin C but in fact contained only sugar
and flavoring. The students did not know whether they were taking vitamin C or not, but the investigators
did. The students were followed for 2 months to see who came down with a cold and who didnt. At the end
of the study, 4 students taking the sugar tablet came down with a cold and only 1 student taking vitamin C
came down with a cold.
d. [3 pts] Select all of the following which applies to the study.
A. Placebo
B. Random sample
C. Observation study
D. Double-blind
E. Random assignment
F. Randomized experiment

e. What is the appropriate scope of inference for this study?


A. Vitamin C causes fewer colds in the group of students who are similar to our sample.
B. Vitamin C causes fewer colds in the population of all students.
C. Vitamin C is associated with fewer colds in the group of students who are similar to our sample.
D. Vitamin C is associated with fewer colds in the population of all students.

You might also like