Evaluating Burger (2009)

Burger (2009) chose an experimental design that compared an experimental group and a control group. Burger's choice of experimental design was appropriate because it made a comparison between two groups; the first, the group exposed to the treatment, and the second, the control group . In effect, an experimental design is a planned interference of responses (Frankfort-Nachmias & Nachmias, 2008). In the Burger study, the dependent variable was the point at which the participants stopped shocking the learner. He wanted to see if manipulating the participants' exposure to a refusing individual (independent variable), produced a change in the point at which participants stopped shocking the learner.

An experimental design administers a treatment to one group of participants for the purpose of observing their responses (Frankfort-Nachmias & Nachmias, 2008). This type of design manipulates the conditions imposed upon one group of individuals and maintains a control group, upon which the conditions are not imposed. For the Burger (2009) study, an experimental design was more appropriate than a cross-sectional or a quasi-experimental design because of the need to manipulate variables. This manipulation allows the researcher to determine the direction of causation (Frankfort-Nachmias & Nachmias, 2008). A cross-sectional or quasi-experimental design is chosen when the researcher seeks to make observations rather than determine causality. Pre-experimental designs have no control group and are not appropriate for determining causation (Frankfort-Nachmias & Nachmias, 2008).

Threats to Validity

Experimental designs can be weak in external validity because real-life circumstances cannot be replicated and laboratory environments can be construed and difficult to generalize to larger populations (Frankfort-Nachmias & Nachmias, 2008). The Burger (2009) study was somewhat weak in external validity for this reason. In addition, in a laboratory environment, participants may not be representative of a population about which the research may wish to generalize (Frankfort-Nachmias & Nachmias, 2008). The Burger study did not randomly select its participants; they were selected according to several criteria. This was important, however, because of ethical considerations. At the start, participants were self-selected, in that they responded to a call for participants. The lack of randomization is a threat to external validity.

Typically, when seeking evidence of causation (internal validity), which is typical in experimental design, generalizeability/external validity is sacrificed (Frankfort-Nachmias & Nachmias, 2008). Burger extended great control over the independent variable, which strengthened internal validity. He was careful to control internal validity by controlling for the effects of extraneous variables, such as gender, personality traits, education, and ethnicity.

Other threats to internal validity include maturation, issues in testing, such as when the pre-test affects the results of the post-test, using imprecise or inappropriate measures, non-random selection (selection bias), and participant drop-out, altered behavior of participant(s) from being studied.

Burger, J. M. (2009). Replicating Milgram: Would people still obey today? American Psychologist, 64(1), 1-11. doi:10.1037/a0010932

Frankfort-Nachmias, C., & Nachmias, D. (2008). Research methods in the social sciences (7th ed.). New York: Worth.