ABSTRACT
In almost any military occupational specialty soldiers must monitor and react to stimuli, store and process information, make and communicate decisions, while using different in- and output modalities. When assessing influences on performance, e.g., by sleep-deprivation or heat-stress in intervention designs, researchers face a dilemma: They can either examine highly trained military experts in their jobs, thus risking limited generalizability of their results, or use standard measures of cognitive performance with little ecological relevance. To solve this dilemma, we developed a multidimensional job simulation by abstracting a wide range of military jobs into a generic military simulation of a complex workplace (GEMS COW). It had to meet the following requirements: 1) validly and reliably measure the relevant psychological constructs of the abstracted jobs, 2) require only minimal training, and 3) score high acceptance with military specialists. GEMS COW assesses attention, short-term memory (sm), and working memory (wm) in a lifelike setting. Three studies were conducted focusing on psychometric properties (Study 1, N = 74), overall task complexity in an intervention design (Study 2, N = 54), and acceptance (Study 3, N = 22) of GEMS COW. Psychometric properties proved satisfactory (construct validities: wm = .58, attention = .45, sm =.68; retest reliability: .61-.90). The simulation has a medium to high mental difficulty and soldiers accept it as a realistic military task. GEMS COW is easy to learn and reliably measures psychological constructs in a complex generic simulation. It can be used in intervention studies and may easily be implemented in other languages.
