Choice time, or disjunctive reaction time (RT), has been found to be a linear increasing function of the average amount of information transmitted (H t ) per stimulus-response event in a variety of tasks (e.g., Gregg, 1954;Hick, 1952;Hyman, 1953). Hick found that the same function fitted the data when errorless performance was required and also when S speeded up his responses to the point where a substantial number of errors occurred. Hyman, demanding errorless performance, found that RT 1 This report is a consolidation of data from two master's theses by Brainard and Irby; these were completed in 1956 and 19S7 as part of a general research program for which P. M. Fitts and E. A.

No abstract

SIMNET, an acronym for simulator networking, was initiated in 1983 as a project on large-scale simulator networking by the Defense Advanced Research Projects Agency (DARPA; Order AO 4739, signed February 15, 1983). It is a proof-of-principle technology demonstration of interactive networking for real-time, person-in-the-loop battle engagement simulation and war-gaming. Intended for military collective training, SIMNET is also adaptable for training or exercising commanders and staffs at higher echelons, usable in the development of military concepts and doctrine, and suitable to the testing and evaluation of alternative weapon system concepts prior to acquisition decisions. This paper summarizes the technical history of SIMNET development and identifies lessons learned that could contribute to the success of future efforts to develop training technologies and systems, especially for collective training. It concludes with a discussion of the implications and challenges of SIMNET for the human factors and training technology communities.

The application of synthetic tasks to the assessment of complex performance is discussed in relation to the trade-offs involved in achieving adequate levels of face validity and in specifying the exact changes in psychological functions that may result from particular environmental manipulations. It is argued that the multiple-task performance battery approach can provide levels of face validity adequate to maintain the motivation of subjects while at the same time permitting the identification of changes in specific performance functions. The characteristics of this approach are discussed in relation to a program of research on the effects of confinement and demanding work-rest schedules on crew performance.

Thirteen investigations were carried out as a part of an 8-year program of research on the performance effects of various work/rest schedules during confinement to a simulated aeorspace vehicle crew compartment. A total of 139 subjects were tested using a standard battery of performance tasks. The synthetic work approach used provided a reliable, face-valid, and sensitive technique for assessing complex operator performance. It was found that a man can work 12 hours per day on a 4-hours work/4-hours rest schedule for periods of at least 30 days. For shorter periods, a man can work 16 hours per day on a 4/2 schedule but at a significant cost to his reserves for meeting emergencies such as sleep loss. Circadian periodicities are found in psycho-physiological functions paralleled by similar periodicities in performance functions, the latter being subject to modification by special motivational instructions.