The authors goal for this research was to train participants to react to alarms using sensor activity patterns.
In Experiment 1, 80 undergraduates monitored a simulated security screen while completing a primary word search task. They received spatial, temporal, single sensor, or no training to respond to alarms of differing reliability levels.
Analyses revealed more appropriate and quicker reactions when participants were trained and when the alarms were reliable.
In Experiment 2, 56 participants practiced time estimation by simple repetition, performance feedback, or performance feedback and temporal subdivision. They then reacted to alarms based on elapsed time between sensor activity and alarm onset.
Surprisingly, results indicated that participants did not benefit differentially from temporal interval training, focusing instead on advertised system reliability. Researchers should replicate these findings with realistic tasks and real-world complex task operators.
Bliss, J. P., & Chancey, E. T. (2014). An investigation of training strategies to improve alarm reactions. Applied Ergonomics, 45(5), 1278-1284.