Cognitive Requirement Analysis to Derive Training Models for Controlling Complex Systems.

Although simulators are widely used for training of process control tasks, simulator trainings are conducted with little consideration o f research results on cognition, training design or effectiveness (Salas et al., 200 6). Therefore, a hierarchical task analysis described through the sub-goal templates m ethod, a cognitive reliability and error analysis method and a protocol analysis were applied to the Cabin Air Management System (CAMS), a process control simulat ion. Thirty-nine apprentices participated in a CAMS training and a test session a week later. System stabilisation and fault diagnosis as criteria as well as person-r elated variables such as general mental abilities, cognitive style, self-efficacy an d personality traits were measured and think aloud protocols collected. Main results o f the first study of the research project are presented. Each of the task analysis me thods separately contributed to describe cognitive requirements. There were qualita tive differences between good and poor performers, particularly with respect to d ecision errors, errors in sequencing and timing as well as explaining, formin g rules and planning. Furthermore, person-related variables explained an additional proportion of variance. Based on these findings and the results o f raining research, methods of drill and practice, overlearning and error training are recommended.