is thanked for his/her participation. The confederate should return to the classroom momentarily, ending the demonstration.
d. While the participants complete the scrambled sentence tasks, the students in class can be given a general description of the neutral scrambled sentence task (not revealing that there are two versions of it) and can be shown the items on the mood questionnaire.
The goal of the post-demonstration discussion is to have students critically examine the research demonstration and to refresh or introduce methodological topics and terms. The discussion might proceed by posing the following questions to the students:
1. Before students receive additional information about the experiment, what do they think the goal and hypothesis of this experiment might have been? What was the independent variable? The dependent variable? Usually students will think the type of instruction was the independent variable and mood was the dependent variable. Explore as many hypotheses as they can generate.
2. Introduce that the type of instruction was not intended to be manipulated. Is this a potential problem for the experiment? Discuss types of error, the importance of experimental control, and the benefits of experimental scripts and standardization.
3. Give students copies of the two versions of the scrambled sentence task and explain that the manipulation of the independent variable is actually contained in that task. What was manipulated in the task? They will begin to notice that some words are different. Is there a theme present in the words? What could the manipulation be doing? Describe the conceptual variable and its operationalization.
4. Revisit what the hypothesis might be given this new knowledge of the independent variable. Students will still refer to the measure of mood as the dependent variable.
5. Tell students that the mood questions were not actually the dependent variable. Ask the confederate to describe her role in the study and ask the students again to revise their hypothesis given the new dependent variable. Move students toward the hypothesis that the elderly prime should have an effect on behavior by decreasing walking speed compared to the neutral task. Discuss the value of different types of dependent measures (e.g, behavioral, self-report).
6. Have the confederate provide the data (the two walking times for the two participants). Regardless of whether the hypothesis is supported (which it almost always is), ask students what can be concluded from the data. Is the hypothesis supported? Points to discuss include the inadequacy of sample size and the need for statistical analysis (in this case, a t-test). A main point to make is that the prime manipulation is confounded by the type of instruction given, therefore an unequivocal causal conclusion cannot be made regarding the effectiveness of the prime. Additional discussion of the role of random assignment and blind experimenters (and confederates) is also valuable.
7. Have students brainstorm alternative ways to operationalize the independent and dependent variables in the study, and then explore alternative ways that the research question could be studied using other stereotypes and behaviors.
8. Provide students with a copy of the original Bargh, Chen, and Burrows (1996) article and ask them to read the method section for Study 2. Discuss the differences between the demonstration and the actual study. Discuss the level of detail contained in the Method section and how the study could be easily replicated based on the information provided. Examine the results and Figure 2, and discuss the value of the replication. Also discuss that studies may include additional measures (like the mood measure) to help rule out alternative explanations for the results.
9. Walk students through the method and results of studies 1 and 3, discussing the value of conceptual replication and different ways of operationalizing the independent and dependent variables.
10. Discuss the ethics of this research, including the use of deception and cover stories, informed consent, and the “faux” debriefing and real debriefing procedures.
11. As a follow-up to the in-class demonstration, instructors can also use a paper assignment intended to have students extend the Bargh, Chen, and Burrows (1996) study in some way. Students can choose to design a conceptual replication, to design a study that addresses an alternative explanation for one of the three studies, or to design a study that addresses one of the issues raised in the article’s discussion. Students write a short introduction to their research study, and then write a thorough method section, modeling that of Bargh and colleagues, to describe their study. Students then write a “Reflection” section to specifically address design decisions they made with regard to impact, control, random assignment and ethics.
Assessment and Evaluation
To assess the effectiveness of the activity in refreshing methodological terms and understanding the topics discussed, students in an advanced social psychology laboratory course (N = 10) were asked to define or provide examples of 8 research methods terms (e.g., independent variable, random assignment, confound, conceptual replication) and to indicate levels of confidence in their understanding of each of the terms. Students defined or provided correct examples for significantly more of the items at the post-test (M = 6.4, SD = 1.6) than at the pretest (M = 4.5, SD = 1.6), t (9) = 3.77, p < .01. Students were also more confident in their responses on the post-test (M = 6.1, SD = 1.0) than on the pretest (M = 4.6, SD = 1.1), t (9) = 9.87, p <. 01. Informal comments about the activity suggest that it is an engaging and useful way to reintroduce these concepts at the beginning of a semester.
Bargh, J. A., Chen, M., & Burrows, L. (1996). Automaticity of social behavior: Direct effects of trait construct and stereotype activation on action. Journal of Personality and Social Psychology, 71 (2), 230-244.
For questions or additional information, contact: email@example.com
Presented at the Teaching Pre-conference at the 8th Annual Meeting of the Society for Personality and Social Psychology, Memphis, TN, January 2007