P/S 03 · Conformity, bystanders and research ethics
Going Along and Standing By
Interpret three tables from studies of conformity and helping.
Passage
Social psychologists study how the presence of other people changes the way individuals think, feel and act. Two well-known areas are conformity, which is a change in a person’s behavior or belief to match a group, and helping behavior, which is influenced by how many other people are present. Researchers often study both in the laboratory by using confederates, who are people who appear to be participants but are actually working for the experimenter.
Part 1. In a study of conformity, a volunteer sits at a table with a number of confederates and is asked to judge which of three comparison lines matches a standard line in length. The correct answer is always obvious when the volunteer is alone. On each critical trial, the confederates answer aloud, one after another, and all give the same wrong answer. The volunteer answers last. The researchers vary the number of confederates from 0 to 10 and record the percentage of critical trials on which the volunteer gives the same wrong answer. Table 1 shows the results from 40 volunteers for each group size.
Confederates are trained to give their answers in a natural voice, and they are told which trials are critical. On the other trials, called neutral trials, the confederates give the correct answer, so that the wrong answers do not seem strange to the volunteer. Only the critical trials are counted in the tables below.
Table 1. Percentage of critical trials with the group’s wrong answer.
| Number of confederates | Percentage of conforming answers |
|---|---|
| 0 | 2 |
| 1 | 8 |
| 3 | 30 |
| 6 | 33 |
| 10 | 32 |
In a second group of 40 volunteers, six confederates are used and the researchers change one condition at a time. Table 2 shows the results.
Table 2. Conformity with six confederates under different conditions.
| Condition | Percentage of conforming answers |
|---|---|
| Volunteer answers aloud after all the confederates (as in Table 1) | 33 |
| Volunteer writes the answer privately | 12 |
| One of the confederates gives the correct answer | 9 |
Part 2. In a study of helping, volunteers are told that they will take part in a group discussion over an intercom, each in a separate room. During the discussion, one speaker appears to have a medical emergency and calls for help. The volunteer believes that either no one else, one other person or four other people are listening. The researchers record the percentage of volunteers who leave the room to get help within two minutes. Table 3 shows the results for 40 volunteers in each group.
The intercom keeps the volunteers from seeing one another, so the number of other listeners is changed only through what the volunteer is told. Each speaker’s voice and script are recorded in advance and are the same in every group.
Table 3. Percentage of volunteers who sought help within two minutes.
| Number of other listeners the volunteer believes are present | Percentage who sought help |
|---|---|
| 0 | 85 |
| 1 | 60 |
| 4 | 40 |
All volunteers were recruited from the same introductory psychology course and received course credit. After the sessions, each volunteer was told about the true purpose of the study, the role of the confederates and the fact that no real emergency occurred, and was given the chance to withdraw their data.
Questions
Question 1 of 6
What does Table 1 suggest about the relationship between the number of confederates and conformity?
The values rise from 0 to 3 confederates but then stay about the same (30, 33, 32).
Conformity depends on group size in the data, even though it never reaches 50%.
One confederate produced only 8% conformity, well below the larger groups.
Why the answer is what it is
Conformity increases from 2% with no confederates to 30% with three, and then stays between 30% and 33% for larger groups. A small majority is enough to produce most of the effect.
Review: Social Interaction & Group Behavior · Social Structure, Inequality, Health, & Research Methods
Question 2 of 6
In Table 2, writing the answer privately lowers conformity from 33% to 12%. Which type of social influence is most strongly suggested?
If conformity came from informational influence alone, the answer would be the same in private. The drop when answers are private points to social pressure.
The confederates do not give orders. They only state their own answers.
Deindividuation involves a loss of self-awareness in a group, usually with anonymity. Here the volunteer is publicly identifiable and still conforms.
Why the answer is what it is
When responses are public, volunteers conform partly to avoid standing out, which is normative influence. When the answer is private, that pressure is reduced, and conformity drops to 12%.
Review: Social Interaction & Group Behavior · Social Structure, Inequality, Health, & Research Methods
Question 3 of 6
When one of the six confederates gives the correct answer, conformity falls to 9%. The best interpretation is that:
Random answers would not drop to a level below that of single confederates; the drop is large and systematic.
The correct answer is obvious when the volunteer is alone, so the volunteer does not need to be taught.
Conformity falls to 9%, but not to the 2% seen with no confederates, so some influence remains. The group is not irrelevant.
Why the answer is what it is
Breaking the unanimity of the group, even with one dissenter, lowers conformity sharply. Social support for one’s own judgment makes it easier to resist the majority.
Review: Social Interaction & Group Behavior · Social Structure, Inequality, Health, & Research Methods
Question 4 of 6
In Part 2, the percentage of volunteers who sought help falls as the number of other listeners rises. This pattern is best explained by:
The fundamental attribution error is about explaining other people’s behavior, and it is not the process described by the data.
Social facilitation concerns performance on tasks in the presence of others. It does not explain failing to help.
Mere exposure is about liking things that are familiar, and it has no clear link to helping.
Why the answer is what it is
When more people could help, each person feels less responsible for acting, which is diffusion of responsibility. Helping falls from 85% when alone to 40% with four other people.
Review: Social Interaction & Group Behavior · Social Structure, Inequality, Health, & Research Methods
Question 5 of 6
Which feature of the procedure is most important for the ethics of a study that involves deception, as in this passage?
Ethical studies pay attention to the volunteers’ reactions, and asking about them is part of debriefing.
Using a convenience sample affects generalizability, not the ethics of the deception.
The use of confederates is the deception itself, so it is not what makes the study ethical.
Why the answer is what it is
Deception can be justified only if the study has significant value, if no real harm results, and if participants are fully debriefed afterward. The passage describes the debriefing and the option to withdraw data.
Review: Social Interaction & Group Behavior · Social Structure, Inequality, Health, & Research Methods
Question 6 of 6
All volunteers were introductory psychology students who took part for course credit. Which concern is most directly raised by this fact?
The problem described is the sample, not the assignment. The passage does not say that assignment was non-random.
Nothing in the passage suggests that the students knew the researchers.
Each volunteer is tested in one condition, so repeated testing is not the concern.
Why the answer is what it is
A sample of students from one course may differ from other populations in age, background and attitudes. The results may therefore not generalize beyond that group, which is a question of external validity.
Review: Social Interaction & Group Behavior · Social Structure, Inequality, Health, & Research Methods
Built from these lessons