Study D Professors at Harvard University and the University of California-San Di
ID: 3470286 • Letter: S
Question
Study D
Professors at Harvard University and the University of California-San Diego wanted to know whether leaders of countries who issue verbal warnings to potential invaders were: a) more likely to be attacked; and, b) gauge the leaders’ willingness to defend themselves (rather than back down). To answer this question, the professors recruited 16 undergraduates to come into the lab and play a stylized game with each other. Players rotated between the role of “defender” or “invader” and received pay-offs for each round played depending how well their side did in the potential war (basically, playing “fight” was costly, especially if the other side did the same). At the end of the day, the pay-offs from each game were translated into cash payments that were handed to the players. During each round of play, it was randomly determined whether the players could communicate with each other. Specifically, when players were allowed to communicate with each other, the defender could tell the potential invader “I plan to defend myself.”
The professors found that defenders who stated that “they will defend themselves” were 10 percentage points less likely to be attacked (compared to the games when communication was not allowed) and 4 percentage points more likely to fight when attacked. They conclude that these verbal warnings change the behavior of erstwhile invaders and strengthen the resolve of defending countries.
Comment on the construct validity of the experiment. In particular, do you think the statement “I plan to defend myself” accurately measures the type of talk national leaders engage in?
Comment on the internal validity of the experiment. Does the experiment provide an unbiased estimate of the effect of communicating “I will defend myself” on game play?
Suppose the researchers decided to conduct an observational study as well on the topic. They collected public statements from countries on the brink of war and coded whether the statements were bellicose (e.g., “We will defend ourselves”) or negotiating (e.g., “war is a last resort” or “we still hope for peace”). The researchers found that leaders making bellicose public statements were 10% LESS likely to defend themselves when fighting over the disputed territory. Provide a possible explanation for this finding.
What is the chief advantage of the observational study over the lab experiment?
What is the chief disadvantage of the observational study compared to the experiment?
Explanation / Answer
Observational study is done in a natural set-up, so the results would be more authentic. In the experimental study, the participants are briefed on the whole process and aim of the experiment thus giving room for the participants to be more aware of how they will be behaving. Participants might dodge the responses which will change the results.
The problem with the observational study is that people may start behaving differently if they become aware that they are being observed. Experimental design will have two groups, one experimental and the other control to compare the data but in observational study, we will have only one group thus making it difficult for a comparative study.