Table 2 Normative recommendation distributions and moral disagreement effects for each dilemma, experiment, and agent.
From: The power of justifications to repair human-robot trust, even under moral disagreement
Exp | Agent | Dilemma | Share of the Sample who Recommended each Action path | Moral Disagreement Effect size \(\:{\eta}_{p}^{2}\) | |
|---|---|---|---|---|---|
Do not feed | Feed | ||||
1 | Robot | Hunger Strike | 46.5% | 53.5% | 0.22 |
2 | Robot | Hunger Strike | 50.4% | 49.6% | 0.40 |
Do not resuscitate | Resuscitate | ||||
2 | Robot | Dot Not Resuscitate (DNR) | 34.4% | 65.6% | 0.38 |
3 | Robot | DNR | 28.8% | 71.2% | 0.35 |
3 | Human | DNR | 25.8% | 74.2% | 0.32 |
Do not divert | Divert | ||||
3 | Robot | Toxic Gas | 46.7% | 53.3% | 0.27 |
3 | Human | Toxic Gas | 59.3% | 40.7% | 0.24 |
