Suchergebnisse

Objective: Two experiments are described that examined operators' perceptions of decision aids. Background: Research has suggested certain biases against automation that influence human interaction with automation. We differentiated preconceived biases from post hoc biases and examined their effects on advice acceptance. Method: In Study 1 we examined operators' trust in and perceived reliability of humans versus automation of varying pedigree (expert vs. novice), based on written descriptions of these advisers prior to operators' interacting with these advisers. In Study 2 we examined participants' post hoc trust in, perceived reliability of, and dependence on these advisers after their objective experience of advisers' reliability (90% vs. 70%) in a luggage-screening task. Results: In Study 1 measures of perceived reliability indicated that automation was perceived as more reliable than humans across pedigrees. Measures of trust indicated that automated "novices" were trusted more than human "novices"; human "experts" were trusted more than automated "experts." In Study 2, perceived reliability varied as a function of pedigree, whereas subjective trust was always higher for automation than for humans. Advice acceptance from novice automation was always higher than from novice humans. However, when advisers were 70% reliable, errors generated by expert automation led to a drop in compliance/ reliance on expert automation relative to expert humans. Conclusion: Preconceived expectations of automation influence the use of these aids in actual tasks. Application: The results provide a reference point for deriving indices of "optimal" user interaction with decision aids and for developing frameworks of trust in decision support systems.

Automation users often disagree with diagnostic aids that are imperfectly reliable. The extent to which users' agreements with an aid are anchored to their personal, self-generated diagnoses was explored. Participants (N = 75) performed 200 trials in which they diagnosed pump failures using an imperfectly reliable automated aid. One group (nonforced anchor, n = 50) provided diagnoses only after consulting the aid. Another group (forced anchor, n = 25) provided diagnoses both before and after receiving feedback from the aid. Within the nonforced anchor group, participants' self-reported tendency to prediagnose system failures significantly predicted their tendency to disagree with the aid, revealing a cognitive anchoring effect. Agreement rates of participants in the forced anchor group indicated that public commitment to a diagnosis did not strengthen this effect. Potential applications include the development of methods for reducing cognitive anchoring effects and improving automation utilization in high-risk domains.

Objective: We tested the hypothesis that automation errors on tasks easily performed by humans undermine trust in automation. Background: Research has revealed that the reliability of imperfect automation is frequently misperceived. We examined the manner in which the easiness and type of imperfect automation errors affect trust and dependence. Method: Participants performed a target detection task utilizing an automated aid. In Study 1, the aid missed targets either on easy trials (easy miss group) or on difficult trials (difficult miss group). In Study 2, we manipulated both easiness and type of error (miss vs. false alarm). The aid erred on either difficult trials alone (difficult errors group) or on difficult and easy trials (easy miss group; easy false alarm group). Results: In both experiments, easy errors led to participants mistrusting and disagreeing more with the aid on difficult trials, as compared with those using aids that generated only difficult errors. This resulted in a downward shift in decision criterion for the former, leading to poorer overall performance. Misses and false alarms led to similar effects. Conclusion: Automation errors on tasks that appear "easy" to the operator severely degrade trust and reliance. Application: Potential applications include the implementation of system design solutions that circumvent the negative effects of easy automation errors.

Objective: The purpose of the current work was to clarify how subjective trust determines response behavior when interacting with a signaling system. Background: In multiple theoretical frameworks, trust is acknowledged as a prime mediator between system error characteristics and automation dependence. Some researchers have operationally defined trust as the behavior exhibited. Other researchers have suggested that although trust may guide operator responses, trust does not completely determine the behavior. Method: Forty-four participants interacted with a primary flight simulation task and a secondary signaling system task. The signaling system varied in reliability (90%, 60%) and error bias (false alarm, miss prone). Trust was measured halfway through the experimental session to address the criterion of temporal precedence in determining the effect of trust on behavior. Results: Analyses indicated that trust partially mediated the relationship between reliability and agreement rate. Trust did not mediate the relationship between reliability and reaction time. Trust also did not mediate the relationships between error bias and reaction time or agreement rate. Analyses of variance generally supported specific behavioral and trust hypotheses, indicating that the paradigm employed produced similar effects on response behaviors and subjective estimates of trust observed in other studies. Conclusion: These results indicate that strong assumptions of trust acting as the prime mediator between system error characteristics and response behaviors should be viewed with caution. Application: Practitioners should consider assessing factors other than trust to determine potential operator response behaviors, which may be more predictive.