"People said they were 14.2% more likely to buy this chocolate bar when they were shown the version with a cost breakdown." What people say they would do and what they actually do are 2 very different things. The conversion rate of actual behavior needs to be tested for this to mean something useful.
You are right regarding the chocolate bar experiment.
That's why the researchers ran multiple experiments, some of which measured actual behavior:
- People were 16.1% more likely to bid for a gift card for an Everlane backpack (vs a J.Crew one) when they saw cost information about it
- Sales of chicken noodle soup bowls ($4.95) in Harvard’s campus canteen increased 21.1% when costs were disclosed