Experimental support for a trust heuristic

Abstract

Scholars have argued that trust acts as a decision-making heuristic, aiding people in assessing the risks of various technologies. This may be especially likely for social trust, or integrity (vs. calculative trust, or competence), and among people with limited experience with a given technology. The goal of this research is to provide an experimental test of the trust heuristic. In three experiments, which were identical except for the technology of focus (Study 1: autonomous vehicles, N = 1691; Study 2: nuclear power, N = 844; Study 3: airplane travel, N = 853), participants recruited from an online platform completed measures of trust, and then were randomly assigned to respond to measures of perceived risk in either a treatment (time pressure) or control condition. Competence and integrity were significantly negatively associated with risk perceptions. In Studies 1–2, the relationship between integrity (not competence) and risk was significantly stronger in the timed vs. untimed condition. However, Study 3, which focused on a context characterized by greater experience, did not find this effect. Results support the role of integrity, but not competence, as a decision-making heuristic, and suggest that greater levels of experience weaken the influence of integrity on risk.

Keywords:

• Trust
• heuristics
• judgement and decision making
• technology acceptance

Acknowledgments

The authors would like to thank The Ohio State University Center for Automotive Research for funding the present research.

Data availability

The data that support the findings of this study are openly available in Mendeley Data, doi: 10.17632/pcsvfnbv66.1

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 Time pressure manipulations produce similar effects on judgments as other divided-attention manipulations, as they both impair people’s ability to correct their judgments (Gilbert and Gill 2000). For example, a time pressure manipulation and digit search task (in which participants were under time pressure or had to count “5”s on a screen while reading a report) produced the same impacts on social judgment, in which participants were not able to unlearn false information they read about a target individual in the report (Gilbert, Tafarodi, and Malone 1993). Similarly, having participants hold a number in their head and time pressure had the same effects on whether people preferred favorable or unfavorable feedback (Swann et al. 1990).

2 We also tested H1 in Studies 1-3 following the methods of (Finucane et al. 2000), presented in Appendix G. Results align with those presented here, with one exception. That is, in Study 2, the correlation between competence and risk differed significantly across conditions based on the alternate modeling approach, such that the coefficient was significantly more negative in the timed vs. untimed condition. This result was not observed when using PROCESS.

3 As in Study 1, we ran exploratory analyses for Studies 2 and 3 to test the possibility that competence mediates the relationship from integrity to risk (Appendix H). Results align with those presented here. Additionally, Study 2 found a significant indirect effect of integrity on risk, mediated by competence. Study 3 found the same, but only among those with limited experience.

Additional information

Funding

This work was funded by OSU Center for Automotive Research.