Evidence Relationship: Supports
Vote on whether "Trolley verdicts show test-retest reliability of r=0.34 at six months, versus r=0.71 for the same subjects' abortion and capital punishment attitudes" is good evidence that supports the claim "Order of presentation reverses majority verdicts on Footbridge in 31% of subjects (Lindqvist & Ojeda, 2020, N=4,812)"
Evidence Claim
Trolley verdicts show test-retest reliability of r=0.34 at six months, versus r=0.71 for the same subjects' abortion and capital punishment attitudes
Bramwell (2021) tracked 1,140 participants across two waves, administering four trolley variants alongside a standard applied-ethics attitude battery. Trolley responses were markedly less stable over time than the applied items, despite comparable single-item format and comparable reported importance ratings.
Main Claim
Order of presentation reverses majority verdicts on Footbridge in 31% of subjects (Lindqvist & Ojeda, 2020, N=4,812)
In a preregistered multi-site study, participants who saw Switch before Footbridge endorsed pushing at 21%, while those who saw Footbridge first endorsed it at 12%; the reverse asymmetry appeared for Switch endorsement. Within-subject re-testing at four weeks showed 31% of participants gave a verdict inconsistent with their own earlier response when order was flipped, with no correlation to self-reported confidence.
Log in to join the discussion.
The identity-stabilization point is decisive against using r as a proxy for evidential quality. Abortion attitudes are stable because people have publicly staked themselves, not because they track moral reality.
Agreed that stability isn't sufficient for evidential value. My claim is narrower: instability of this magnitude is a defeater, whatever explains the contrast case.