Psychology Research Lesson on Bystander Intervention

Taught November 11, 2003

Developed for Psychology 100 by Melanie Cary, Bill Cerbin, Rob Dixon and Carmen Wilson

Taught by Bill Cerbin and observed by Melanie Cary, Rob Dixon and Carmen Wilson

Goal of the lesson: To develop students’ ability to analyze and explain human behavior in terms of relationships among multiple factors (or variables).

Rationale for the goal: Psychological reasoning involves thinking in terms of the relationships among variables. To explain any form of behavior we need to determine the factors, conditions, variables, and antecedents that influence the actions in question. Most people engage in this kind of thinking on a daily basis, and gradually develop "intuitive theories" to explain what makes people tick. In the general psychology course we introduce students to many psychological models on a wide range of topics. They are good at remembering the models, but have trouble using them to explain and predict human behavior. Moreover, their own "intuitive theories" of behavior tend to interfere with learning and using discipline-based models. Consequently an important goal of the course is to help students move beyond common sense explanations and develop an understanding of empirically-tested models of human behavior.     

Research Lesson: "Bystander Intervention: Why Do Onlookers Come to the Aid of Strangers?"

The lesson focused on the psychology of "bystander intervention" (i.e., why strangers help or do not help a person in need). A number of factors influence how bystanders interpret an incident and whether they assume responsibility to help (e.g., the number of bystanders present, whether the victim appears to need and deserve help, etc.).

Research Lesson Design

    Before the class. One week before the lesson, students took a Bystander Intervention Pre-test consisting of seven (7) scenarios depicting people in need of assistance. They predicted whether bystanders would help or not help in the situations, and what factors would influence bystanders' behavior.

    During the class

·       The instructor described the lesson and gave written directions to students.

·       In small groups students created a list of factors that influence bystander intervention.

·       The instructor circulated among groups to answer questions.

·       Students prepared a written summary and an overhead transparency with their group's list of factors.

·       The instructor led a class discussion comparing the groups' responses.

·       The instructor introduced the research-based model of bystander intervention.

·       Students analyzed their factors against the research-based bystander intervention model.

·       Students wrote an individual analysis of how their group's factors fit the bystander model.  

    After the class. Students completed a Bystander Intervention Post-test in which they predicted bystander behavior in four scenarios. At the next class period the instructor leads a class discussion comparing students' post-test answers to the actual research results.

Rationale for the Lesson Design

We wanted students to have direct experience analyzing and explaining behavior in terms of relationships among variables. To do this, the lesson involved students in three analytical activities: 1) generating initial ideas about why bystanders help or do not help, 2) comparing their explanations of bystander intervention to the research-based model of bystander intervention and 3) predicting and explaining how bystanders would act in several novel situations.

Results and Analysis of Research Lesson Post-Test

We used a Bystander Intervention Post-Test to evaluate students' understanding of the factors that influence bystander intervention. The test consisted of four scenarios depicting a person in need. Students were asked to predict and explain bystander behavior for each scenario. We assigned scores of 0-3 to their predictions and explanations based on the extent to which the answer was consistent with the research model of bystander intervention.

Below are examples of each score level for Post-test Item 1. The scenarios are based on actual research studies of bystander intervention.

Post-test Item 1. Research subjects are asked to fill out a questionnaire in a waiting room. Before long, smoke started to come into the room through a small wall vent. The researchers were interested in determining whether subjects would investigate the situation to see if it was an emergency and how long it would take before doing so. The researchers varied the number of subjects in the waiting room—sometimes there was just one subject in the room alone, sometimes three subjects were in the room, and sometimes there was one subject and two other people (who were confederates of the experimenter and who ignored the smoke and continued working on the questionnaire). Predict how the subjects responded. In which situation were the research subjects most likely to check out whether the smoke posed a threat (alone, with 2 other subjects, with 2 confederates who ignored the smoke)? In which situation were they least likely to check out whether the smoke posed a threat? Explain the reasons for your predictions.

Consistent explanation (Score = 3)

The situation in which the research subjects were most likely to check out the smoke threat was when the subject was in the room alone. They were probably least likely to check out the smoke threat when in the room with two confederates. When alone a person realizes that it is up to them completely to check out the smoke, with no pressure from anyone and no help from anyone else either. When in a room with 2 unconcerned confederates, the research subject is least likely to check it out because he/she may feel that they are overreacting since it doesn’t seem to concern the other 2 people in the room.

Partially consistent explanation (Score = 2)

If the person is alone, I think that they would investigate it since no one else is there. If there is just three people I think they would be less likely to respond since they can diffuse the responsibility. However, with two confederates ignoring the smoke the subject would be forced to investigate the smoke to see if there was any real danger.

Inconsistent explanation (Score = 1)

The lone subject would have been very suspicious, and may have been cautious investigating the smoke, but not very confident, being alone and in possible danger. With two other subjects, more confidence would have been observed as more individuals would take notice along with the subject. This would provide a dispersion of confidence to examine the smoke. The two confederates would have been a little suspicious to a subject, and cause skepticism within the subject about the whole survey situation.

Table 1 reports the mean score for each of the four post–test items. Students’ explanations were most consistent with the research model for Item 3 (man collapsing on subway), probably because the “common sense” answer to the scenario is consistent with the research-based answer. Their explanations were least consistent with the research model for Item 4 (student who has seizure), which produced the greatest number of inconsistent explanations (8). In this case students thought that “everyone” would help a fellow student having a seizure, implying that helping in this situation is the norm. For example, one student argued that subjects would be more likely to help when other bystanders were present because the person “wants to live up to the expected social norm.”   

Table 1: Mean Score for Predictions and Explanations on the Bystander Post-Test

(maximum score = 3.0; N=25)

Table 2 reports the percentage of students scoring at each level on the items. The percentages indicate that a large majority of students’ predictions and explanations were consistent with the research model of bystander intervention for each post-test item.

Table 2: Percentage of Students at Each Score Level on the Bystander Post-Test

Discussion and Conclusions

Because the pre and post tests for the Research Lesson are not comparable, we cannot quantify the differences between students’ understanding prior to and after the lesson. However, the patterns of responses indicate that students’ understanding of bystander intervention improved in three key areas:

1. Understanding Diffusion of Responsibility. Prior to the Research Lesson students tended to think a bystander is more likely to help when other bystanders are present in a situation (i.e., “strength in numbers”). On the post-test a large majority of students indicated that help is more likely when only one bystander is present, based on the idea of "diffusion of responsibility."

2. Understanding how social context affects individual behavior. On the pre-test students often based their predictions on the bystander’s character. Bystanders who are empathic, sympathetic, nice, and caring are more likely to help than those who lack empathy and do not care what happens to other people. On the post-test, few students referred to the bystander’s personality or disposition.

3. Analyzing behavior. The Research Lesson helped develop students' ability to analyze and explain human behavior in terms of relationships among multiple factors (or variables). On the post-test students were able to clearly state causal connections between factors in the situation (e.g., number of bystanders) and subsequent behavior (i.e., whether a person would help or not). 

Research Lesson Report updated March 5, 2004