Learning Evolved, Part III: Coalitionary Education

Altruistic punishment is another tool to rely on. People will forfeit rewards if it means they can punish free-riders and others who have treated them unfairly (Sanfey et al, 2003, 1755). Give students a meaningful opportunity to punish free-riders in their group, such as being able to deduct a point from another’s students project at a cost to one point to the punishing student. The more free-riders free-ride, the more they will be punished, and such a system will significantly increase contributions from students as a whole (Fehr and Gachter, 2000). The benefits of punishing increases with each cheater deterred from free-riding (Boyd et al, 2003, 3533). This sort of punishment is absolutely critical to group work, because without it the human drive to avoid an unfair deal (Smith, 2006) may cause students to give up rather than labor under unfair conditions. Likewise, the presence of free-riders can increase tension and complaints within a group (Price, 2006, 32). Just as we eschew types of instruction that destroys self-efficacy, and so lessons motivation (Chemers, Hu, and Garcia, 2001, 56), we must view free-riding as serious and organize groups such that students can punish, and thereby eliminate, shirking behavior.

Another threat to motivation is the outbreak of intergroup peace. That is, if one group believes the other will not try hard, it may not try hard. Worse, this peace can snowball, because cooperation reduction is most common between groups which have made it a habit (Sapolsky, 2004). Fortunately, human competitiveness should minimize the chances of this. All a teacher must do is make sure that student groups perceive a threat from another groups, and the behavior will follow automatically. A “perception of threat makes [some people] more likely to perceive threat-based messages as credible” (Lupia, 2002, 321), and such messages can be facilitated by having public group meetings at the end of class or in other ways. Humans are naturally biased thinkers (Lodge and Taber, 1-2), and teachers should design classes to exploit the natural human bias towards group competition, and against intergroup stasis.

Of course, there are problems here too. Certain populations are more averse to punishing cheaters than others (Kotulak 2006; Singer et al, 2006), or more be more aggressive than others McDermott 35 2006). Fully a quarter of the population does not punish cheaters in laboratory experiments (Kurzban, DeScioli, and O’Brien, 2006) and these types that do not are stable — they perform the same behavior in experiment after experiment (Smith et al, 2004). Interestingly, it appears to be possible to quickly identify who will act cooperatively and who won’t based on simple games (Kurzban and DeScioli, 2005), which implies teachers can use quick “fun-day” session to know how to group students for maximal group output. Likewise, in general required Classes, students may be sorted by major as certain fields of study attract more cooperative students than others (Guth and Tietz, 1990). This can be done by making sure that enough punishers exist in every group to get the most best effort out of the free-riders (Orbell et al, 2004) — to motivate the unmotivated, in other words. If predictive performance sorting such as standardized test scores (Robbins et al, 2005, 262) GPA (Weissberg and Owen, 2005, 308), and standardized test scores (Robbins, Le, and Lauever, 2005, 411) are accepted as valid , then gameplay and majors should be, as well.

Throughout series I have argued for an approach in which, while the system is designed by the teachers, much of the everyday work is done by the students. This does not take away old methods, such as having a well defined syllabus (Barker, 2002, 382), but rather complements those methods with new knowledge.. Sometimes the best thing that teachers can do is to sit back and not pretend to know every answer (Roth, 1996, 203), or for that matter to know the best way to motivate in every circumstance. Researchers in other fields have now confirmed this, demonstrating that decentralized reward-and-punishment models do not need an all-seeing governing authority (Orbell et al, 2004, 1). Likewise, complex group assignments will allow students to specialize their tasks to their learning styles (Halonen ,2002). Some types of behavior can be adaptive in one setting but ordinarily disruptive in classrooms (Ding, 2002; Harpending and Cochran, 2002), and by allowing the same sort of specialization we see everywhere else in our society, we can design classrooms to get the best motivation out of everyone while still delivering concrete projects, reports, and other academic goods.

Learning Evolved, a companion series to Classroom Democracy
1. Darwinism-Cognitivism
2. Social Motivation
3. Coalitionary Education
4. Bibliography

One thought on “Learning Evolved, Part III: Coalitionary Education”

  1. 'The benefits of cheating increases with each cheater deterred from free-riding'

    is that a misprint? if not, can you explain to me how it makes sense in the flow of that paragraph?

  2. Thank you, O my editor 🙂 The benefits of punishing, not cheating, increase with each cheater deterred. The inspiration of this line [1]:

    “Contributors have higher fitness than defectors if punishers are sufficiently common that the cost of being punished exceeds the cost of cooperating (py > c).”

    The text should be corrected now.

    Interestingly, as punishment goes up with deviation from the norm and one's own contribution, the situation for cheaters gets worse with every cheater deterred in another way, too.

    [1] http://www.tdaxp.com/archive/2006/09/06/don-t-get-suckered.html

Leave a Reply

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