The most parsimonious explanation for the results is that humans behave more rationally and less cooperatively when under cognitive load, at least when it comes to retribution. This echoes (Bazerman, White, & Lowenstein, 1995), which found the same increased rationality and decreased fairness when decisions had to be made simultaneously, and fits other findings (Robert & Carnevale, 1997) that socially complex situations create more rational behavior.
The interaction effect shows that cognitive load primarily effects those who would behave cooperatively, but that it has little effect on those who would behavior selfishly. This finding supports the view that humans cooperative warily. Cognitive load did not effect those who acted selfish because they had already ceased cooperating, so could only effect those who have cooperated somewhat already.
Interestingly, a one-way ANOVA with student belief on whether rulebreakers should be forgiven as a factor showed equivalent means across neutral cooperation (F = .253, p = .616), negative cooperation (F = .007, p = .935), and cognitive load (F = .029, p = .864) if the rulebreaker was sorry. Nearly identical results (F = 504, p = .479; F = .504, p = .479; F =.005 , p = 944; sic) were found without the “sorry” clause. In other words, descriptive knowledge about punishment did not translate into procedural knowledge about punishment and, unlike procedural knowledge, did not vary by cognitive load.
The Wary Student, a tdaxp research project
2. Cognitive Load
3. Cooperative Behavior
5. The Experiments
7. Main-Effect Results
8. Interaction-Effect Results
10. Future Research