The Impact of S1 Versus S2 Thinking on Automation Usage Decisions

Automation usage decisions were examined via a System 1–System 2 framework. This study examined intent errors which occurs when the operator knowingly chooses an option with a lower probability for success. The task simulated firing decisions that soldiers often make during combat. Participants were randomly assigned to conditions in a 2 (Automation: No-aid, Aid) x 3 (Credit Points: 0, 3, 6) design. Operators in the aid condition differed from those in the no-aid in that they could rely on “advice” from an error-free machine. Credit points could be earned by correct decisions. Participants in the aid condition were predicted to have shorter response latencies and make fewer errors. Additionally, operators were expected to have longer latencies when making incorrect than correct AUDs. A main effect for the credit points variable was anticipated. Results were consistent with the premise that the presence of an error-free machine produced fewer errors and shorter latencies from reliance on the S1 system. However, the availability of the machine only produced shorter response latencies when the operators’ AUD was correct. If the AUD was incorrect latencies were longer in the aid condition. The credit points variable was not statistically significant in any of the analyses.