Replicating Milgram (Sampling)

The following post will critique the sampling strategy and sample size used in Burger (2009). An appropriate sampling strategy supports a study's generalizeability and part of the value of a study is its ability to make inferences about the population from which the sample is taken (Frankfort-Nachmias & Nachmias, 2008). A sample is representative of the population of focus and is used to make studying a larger population feasible.

Rather than focusing on creating a sample that was accurately representative of the general public, Burger's (2009) focus was to recreate an older study, and because of ethical constraints, was forced to utilize a careful and extensive process of participant selection. Miller (2009) suggested Burger's efforts may have been overdone, lacked effectiveness and cost the experiment too much in terms of findings and generalizeability.

Burger (2009) sought to replicate real life in an experimental setting, but his sampling strategy weakened the overall findings and limited their generalizeability to wider populations. In this laboratory experiment, participants cannot be considered representative of the population about which the research intends to generalize (Frankfort-Nachmias & Nachmias, 2008). The Burger study did not randomly select its participants; they were selected according to several criteria.

It is difficult to determine the effect of the selection and screening process in Burger (2009). A relatively large percentage of individuals were screened out because of psychological or emotional problems and other issues determined by the clinical psychologists who did the final screening. In all, out of the self-selected group of 180 (approximately) volunteers for participation, 30% were excluded. Then, out of the 123 who remained, 70 were chosen to participate, which ends up to be approximately 38% strategically selected from the initial volunteers for participation. Burger's (2009) sample could not be considered representative of the general population because of this highly selective process. Understandably, Burger did not use a random sampling strategy because of ethical concerns about the effects of exposure to the treatment. However, Miller (2009) suggested that perhaps Berger's ethical concerns were overemphasized and by screening out more than half of the volunteers, the cost to the study's precision was immense. After the extensive screening process, however, Berger used a random process to assign participants to one of two treatment groups.

It would appear that the sampling frame for Burger's study was not well-defined, and could not have been due to the constraints imposed by ethics of the psychological field. Although Burger identified his population as people in general, he could not select a sample that represented the general population. And although Burger defined his population, at least in theory, he could not simply randomly select his participants. His primary goal was to replicate Milgram.

Burger (2009) justified his sample size, claiming that although he failed to find significant difference between his and Milgram's study, it was not on account of a lack of power. He believed his sample size was within reasonable limits to the Cohen's recommendations for medium effect. Although the findings of Burger's study are legitimately comparable to Milgram's, his findings are not generalizeable to the general population.

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

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

Miller, A. G. (2009). Reflections on 'Replicating Milgram' (Burger, 2009). American Psychologist, 64(1), 20-27. doi:10.1037/a0014407