Problem #6: the situations are almost invariably underspecified. (Problem 2 is a special case of this.) The moral judgments elicited depend on features that are not explicit, about which the reader can only make assumptions. Such as, how did the five people get on the tracks? Kidnapped and tied up by Dick Dastardly? Do they work for the railroad (and might they then also be responsible for the maintenance of the trolley)? And so on.
When a researcher uses contrived problems to test people’s moral intuitions, it would help to include a free-form question inviting the respondent to say what other information they need to form a moral judgment. That way, the next time the “trolley problem” is trotted out, the researchers will be in a better position to understand which features make a difference to the moral verdicts.
Problem #6: the situations are almost invariably underspecified. (Problem 2 is a special case of this.) The moral judgments elicited depend on features that are not explicit, about which the reader can only make assumptions. Such as, how did the five people get on the tracks? Kidnapped and tied up by Dick Dastardly? Do they work for the railroad (and might they then also be responsible for the maintenance of the trolley)? And so on.
When a researcher uses contrived problems to test people’s moral intuitions, it would help to include a free-form question inviting the respondent to say what other information they need to form a moral judgment. That way, the next time the “trolley problem” is trotted out, the researchers will be in a better position to understand which features make a difference to the moral verdicts.
ETA: didn’t see MatthewW’s similar point until after I replied.