Prediction of dishonest behavior using the CNI model of moral dilemma judgment: a registered report

ABSTRACT

Drawing on moral philosophy, research in moral psychology has used hypothetical sacrificial dilemmas to understand how moral judgments are made by laypeople. Although heavily influential, a frequent question raised is whether responses to hypothetical scenarios are informative about instances of morally relevant behavior. Using the CNI model to quantify sensitivity to consequences, sensitivity to moral norms, and general preference for inaction versus action in responses to sacrificial dilemmas, two preregistered studies examined whether specific factors underlying sacrificial dilemma judgments predict dishonest behavior for personal monetary gain (Study 1) and the greater good (Study 2). Testing the hypothesis that specific factors underlying sacrificial dilemma judgments are linked to other morally relevant behaviors via broader underlying moral dispositions, confirmatory tests provide strong support the predicted associations between sensitivity to moral norms and dishonest behavior. No support was found for a predicted association between sensitivity to consequences and dishonest behavior for the greater good. The findings contribute to ongoing debates about the value of individual differences in sacrificial dilemma judgments for predicting instances of morally relevant behavior.

KEYWORDS:

• CNI model
• deontology
• dishonest behavior
• moral judgment
• utilitarianism

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1. In actuality, all participants received the larger sum of money in compensation for their participation.

2. Some studies have compared moral dilemma judgments on standard text-based formats to behaviors in corresponding dilemmas simulated using virtual reality (VR) technology (e.g. Francis et al., 2016; Patil et al., 2014). However, this work has focused mainly on mean-level differences between judgment and behavior rather than associations between judgment and behavior.

3. The reported average correlations were based on a meta-analysis of the six studies from our group using Levenson et al.’s (1995) PPS (Gawronski et al., 2017; Körner et al., 2020; Luke & Gawronski, 2021b). We did not have any unpublished data from studies using the PPS at the time we conducted the current research.

4. Power analyses were conducted using G*Power 3.1.9.2 (Faul et al., 2007).

5. To avoid artifacts from lack of realism and plausibility (see, Kneer & Hannikainen, 2022; Körner et al., 2019), the CNI model dilemmas have been designed to be more realistic and plausible compared to the trolley problem and its variants (see, Gawronski et al., 2017; Körner et al., 2020). Nevertheless, a considerable number of studies using the CNI model dilemmas has found effects on preference for utilitarian over deontological judgments that replicate earlier findings using the trolley paradigm, including (but not limited to) effects of cognitive resources (Gawronski et al., 2017), incidental happiness (Gawronski et al., 2018), empathic concern (Körner et al., 2020), psychopathy (Luke & Gawronski, 2021b), and political ideology (Luke & Gawronski, 2021a).

6. Körner et al.’s (2020) dilemma battery for research using the CNI model includes 12 basic scenarios in 4 different variants, summing up to a total of 48 dilemmas. However, an item-based analysis of these dilemmas revealed low construct validity of the moral-norms manipulation for one of the 12 basic dilemmas (Gawronski et al., 2020). To ensure high construct validity of our moral judgment measure, this dilemma was not included in the current studies.

7. Participants were also told that, in the case of ties, we would randomly choose among tied participants to receive the $50 bonus payment.

8. Because p(action) = 1 – p(inaction), there are only four non-redundant equations in the full set of eight equations in the Appendix.