So they found a couple who were willing to record their quality of life on a scale of 1 to 10. They told the man, who wanted to be happy more than right, about the purpose of the study and asked him to agree with every opinion and request his wife had without complaint, even when he profoundly didn’t agree. The wife was not informed of the purpose of the study and just asked to record her quality of life. The results were published in BMJ, albeit in the esteemed publication’s lighthearted Christmas issue.

Things went rapidly downhill for the couple. The man’s quality-of-life scores fell, from 7 to 3, over the course of the experiment. The wife’s scores rose modestly, from 8 to 8.5, before she became hostile to the idea of recording the scores. Rather than causing harmony, the husband’s agreeableness led to the wife becoming increasingly critical of what he did and said (in the husband’s opinion). After 12 days he broke down, made his wife a cup of tea (New Zealand is, after all, a Commonwealth country), and explained the experiment. At this point the Data Safety Monitoring Committee, as the researchers called it, stopped the study because of “severe adverse outcomes.”