Evidence for: Order effects on trolley cases persist among professional philosophers with AOS in ethics (Ojeda et al., 2022, N=387)

0 · asserted by ◎ kwan_pet (18) · 2 months ago

The vote is on the link: is “A 2023 direct replication of Ojeda et al. with a preregistered expert-screening…” good evidence for the claim?

This tightens the expertise operationalization that critics of Ojeda attacked. The remaining vulnerability is the book-author subgroup: it is underpowered, but it is the only cell consistent with the expertise defense, and its point estimate is in the defense's favor.
sources
none yet
the evidence

A 2023 direct replication of Ojeda et al. with a preregistered expert-screening protocol recovered an order swing of 7.4 points

Kwan and Petrossian recruited 214 philosophers who had published at least one peer-reviewed paper engaging trolley cases, screened via publication records rather than conference attendance. The order effect shrank but remained significant (7.4 points, 95% CI [2.1, 12.7]). A prespecified subgroup of 41 authors of book-length treatments showed a non-significant 3.1-point swing.

◎ kwan_pet (18) · 2 months ago · experimental-philosophy · replication
the claim

Order effects on trolley cases persist among professional philosophers with AOS in ethics (Ojeda et al., 2022, N=387)

A sample recruited from ethics conference registration lists showed an order-induced swing of 9 percentage points on Footbridge endorsement, smaller than the lay swing of 12 points but statistically robust (p=0.004). Self-reported prior familiarity with the cases did not moderate the effect, nor did years since PhD.

◎ Ojeda_M (15) · 2 months ago · experimental-philosophy · metaphilosophy

discussion

Log in to join the discussion.

0 ◎ deontic_dave (2) · 2 months ago

n=41 with a CI that surely spans zero in both directions. Honest of them to report it rather than bury it, but nobody should be citing that subgroup as a result.

0 ◎ kwan_pet (18) · 2 months ago

CI was [-4.8, 11.0]. We said explicitly it is hypothesis-generating. It has of course been cited as a result anyway.