Effects of training with added difficulties on RADAR detection

Three experiments simulating military RADAR detection addressed a training difficulty hypothesis (training with difficulty promotes superior later testing performance) and a procedural reinstatement hypothesis (test performance improves when training conditions match test conditions). Training and testing were separated by 1 week. Participants detected targets (either alphanumeric characters or vehicle pictures) occurring among distractors. Two secondary tasks were used to increase difficulty (a concurrent, irrelevant tone‐counting task and a sequential, relevant action‐firing response). In Experiment 1 , involving alphanumeric targets with rapid displays, tone counting during training degraded test performance. In Experiment 2, involving vehicle targets with both sources of difficulty and slower presentation times, training under relevant difficulty aided test accuracy. In Experiment 3, involving vehicle targets and action firing with slow presentation times, test accuracy tended to be worst when neither training nor testing involved difficult conditions. These results show boundary conditions for the training difficulty and procedural reinstatement hypotheses. Copyright © 2010 John Wiley & Sons, Ltd.