Statistics of moral commitments

Deontic and Consequentialist Commitments Among Study Subjects.

In order to connect results with ethical theory, it is important to measure the degree to which subjects are more or less disposed to endorse consequentialist or deontic principles. The standard method for measuring such commitments is by so called Trolley Case and Sophie's Choice moral dilemmas. In our survey, questions Q3.1-3.5 posed such dilemmas. There are a number of conceptual problems introduced by these measures, some of which our work investigates. Here we simply report summary measures.

The summary measures diagnose implicit commitment to various deontic or consequentialist principles. This commitment need not be conscious: one should not infer from the fact that, e.g., a subject has a high D_E score that he or she would endorse the doctrine of double effect if it were explicitly formulated, or consciously employed the principle in reason to his or her responses. Rather, a high D_E score, for example, indicates that those responses are rationally explicable by strong commitment to the doctrine of double effect, i.e. the subject is disposed to issue moral judgments which are consistent with the doctrine of double effect, whether or not the subject consciously employed any such principle in his or her reasoning. Similarly, the two general measures of commitment to deontic versus consequentialist principles are a summary measure of the extent to which subjects are disposed to judge in ways that are consistent with deontic or consequentialist principles, and do not, by themselves, indicate whether such principles were consciously employed by subjects.

Commitment to the doctrine of Double Effect can be measured by D_E=Q3.2-Q3.11.
D_E was distributed among our subjects as:

graph 1

with a mean of 2.2222 and an SD of 2.04315. High values on this measure indicate that subjects’ judgments about moral propriety are strongly consistent with a the doctrine of double effect, while low values indicate that the subjects judgments are not at all consistent with the doctrine of double effect.

Commitment to the Doing/Allowing distinction is measured by D_A=Q3.3. D_A (i.e. Q3.3) was distributed among our subjects as follows:

graph 2

with a mean of 3.8286 and an SD of 1.83165 .

MT3 is a Sophie’s Choice scenario in which all actions have an identical consequence, but on one option the agent brings these consequences about himself or herself, while on the other option the consequence is caused by another (along with further undesirable consequences). High values again indicate judgments consistent with a strong commitment to a Doing/Allowing distinction, while low values are inconsistent with a commitment to Doing/Allowing.

The relevance of personal versus impersonal features of context can be measured by P_I=MT4-MT1. P_I was distributed among our subjects as follows:
graph 3

with a mean of 3.9906 and an SD of 2.03538.

Subjects’ judgments are consistent with the high relevance of personal/impersonal features of context when P_I has a high value, and not consistent when P_I has a low value.

The general extent to which subjects’ judgments are consistent with the consequentialist idea that right action maximizes good consequences while minimizing bad consequencies, or conversely with the deontic idea that certain kinds of actions are required or prohibited, regardless of their consequences, can be measured in two ways.

Average relevance of consequences for a subject can be measured by AVG_C_D=Q3.3-|Q3.2-Q3.1| . AVG_C_D is distributed among our subjects as:
graph 4

with a mean of 1.6095 and a SD of 2.74736 .

AVG_C_D gives a measure of the extent to which subjects judge otherwise impermissible actions as permissible when the bad consequences of the actions are unavoidable and the good consequences can be achieved only by doing the action, discounted by subjects’ willingness to differentially judge passive versus active killing as permissible, when killing saves lives. The higher the score, the more consistent a subject’s judgments are with consequentialism, while the lower the score, the more consistent the judgments are with doentic principles.

AVG_C_D is not, however, very good at diagnosing the strongest deontic commitments or the strongest consequentialist commitments. Subjects whose judgments follow very closely to a pure conseuqentialist or pure deontic theory are better otherwise.

Strong deontic and consequentialist commitments can be diagnosed by S_D_C=Q3.1+Q3.2+Q3.3+Q3.5. S_D_C is distributed among our subjects as:
graph 5

with a mean of 14.5825 and a SD of 5.61602.

Those with pure deontic commitments will score all of Q3.1-Q3.3 and Q3.5 as highly inappropriate (4£S_D_C£8), while subjects with pure consequentialist commitment swill score all of Q3.1-Q3.3 and Q3.5 as highly appropriate (24£S_D_C£28). In our sample there were 16 subjects whose judgments were consistent with a strong commitment to deontic principles, while 7 subjects had judgments that were consistent with a strong commitment to conseuqentialism.