The Wary Guerrillla, Part IV: An Experiment

Participants were recruited from a large midwestern university. The undergraduates were mostly traditional college age (93% were between 19 and 23). There were 90 males and 87 females. The experiment was designed and implemented using Medialab software. The participants answered questions and played the economic games on a computer. Steps were taken to create the false impression that participants were playing against other humans. At the beginning of the experiment, we asked participants if we could take their picture to use as an avatar for game play. This deception was designed to increase the perception that participants were playing against other human beings.

The game was organized as follows. Participants began the survey and were informed they would earn extra credit points for their actions. A series of distractor questions was asked and a distractor public goods game was played. Near the end of the experiment, the students were informed another student was given five additional extra credit points to divide and chose to keep four and give away one. Participants were informed that they could accept or reject this, where accepting meant the split stood as it was and rejecting meant that neither the participant nor the confederate would earn these additional points. Students who accepted the unfair split are considered either egoists or altruists. Students who rejected were considered wary cooperators.

After the rejection, an opportunity for punishment was given to the students. They were informed that they could use the points they had already gained by participating in the experiment in order to deduct the same amount of points from the other player. Participants who also accepted this second opportunity for punishment – those who would incur acute personal costs in order to punish injustice – are classified as Wary Guerrillas. Additionally, students who accepted the unfair split were also given the opportunity to punish. Participants that accepted and then punished are termed “Other Guerrillas” and are not covered by our theory.

The experiment is visually laid out below:

This paper focuses primarily on the Wary Guerrilla type as a sub-type of the Classic Wary Cooperator. The Egoist and Altruist types have been extensively explored elsewhere in the literature (Becker, 1976; Monroe, 1994). A few words exploring the “Other Guerrillas” are offered at the end of the results section.

The Wary Guerrilla, a tdaxp series
1. Abstract
2. Terrorism
3. Predictions
4. An Experiment
5. Results
6. Absolute Guerrilla
7. Those Who Cause Less Pain
8. Future Research
9. Political Implications
10. Bibliography

One thought on “The Wary Guerrillla, Part IV: An Experiment”

  1. The Wary Guerrilla is being published one section per day, every day (except Christmas) until December 29th. If you wish I can email you the whole thing… but what's the fun in that? :-p

    Please keep commenting — feedback is the best part of blogging!

  2. Apparently the comment I left a couple of days ago was eaten.

    Anyway, I had a question. I was wondering if the idea of anonymity had been considered in the experiment and then rejected?

    As I understand it the participants are not only connected to their choices by points toward their grade, but potentially by their pictures.

  3. “Apparently the comment I left a couple of days ago was eaten.”

    God bless blogspirit's reliability. I opened a ticket on blogspirit's support with your problem — who knows if it will do anything.

    “Anyway, I had a question. I was wondering if the idea of anonymity had been considered in the experiment and then rejected?

    As I understand it the participants are not only connected to their choices by points toward their grade, but potentially by their pictures.”

    The pictures were taken for a separate, distractor game whose results were used by another group. When they got to the wary guerrilla detector, participants were given the following information:

    'Player 1 has already made his or her decision. Player 1 may still be in the lab finishing up the survey, or he/she may have already left the lab. Neither you nor Player 1 will ever know each other's identities. Neither you nor Player 1 will have an opportunity to communicate with each other The only thing you will know about Player 1 is how he/she decided to split the extra credit points. The only thing Player 1 will know about you is how you decide to react.'

    Interestingly, the other group had games were photos and names were visible for half of their subjects, and half were contestants were anonymous to each other. Their was not statistically significant difference between those conditions.

    That group also showed no difference on the sex of the player — the same result we found.

    Kurzban, DeScioli, and O'Brien's “Audience Effects of Moralistic Punishment” [1] imply that levels of anonyminity should matter, but we did not dupliate their environment and perhaps the conditions we used were too similar to each other to matter.

    It certainly is an area worthy of future research. I hope to run another version of this experiment this semester! 🙂


  4. That is interesting!

    You know, I was thinking that especially since computers were used that the anonymity of the participants wouldn't be guranteed, but maybe it is the other way around. Perhaps there is a difference in how people relate to one another by screen and keyboard that is at work here.

    Could the experiment be replicated where participants are face-to-face / hidden, but seen before or after the experiment / and anonymously but without the computer interface?

    I also wonder if this particular factor might have application in the real world. A real guerilla fighter may be aware others know he is supposedly secretly an insurgent, does that have bearing on his actions toward operations in his community?

  5. Aherring,

    Part VIII, which should be up tomorrow (Wednesday the 27th) will deal with real-world applications, and the follow-up experiments my partner and I propose. He is interested in the angle of religious extremism, while I want to see how this applies to student behavior.

    When I first imagined this project, I was thinking of ways that an occupying power could abort an insurgency before it began. If you could determine who the people /likely/ to become actual guerrillas were on their behavior type in some sort of encounter, you might be able to save a lot of lives.

Leave a Reply

Your email address will not be published. Required fields are marked *