I noticed someone mentioned Rhythm 0—here's my humble opinion: Marina Abramović's experiment actually had significant methodological biases:
Her setup inherently screened out morally principled individuals from the start, attracting only those who viewed it as "just a lighthearted prank, no harm done."
As Abramović grew increasingly distressed, the bar for "moral compromise" kept escalating—the more ethical a participant was, the less willing they became to continue the mistreatment.
This created a self-selecting cycle that ultimately filtered for the most morally deficient individuals. While it superficially mirrors the broken windows theory, it was essentially a purification process for unethical behavior.