Effects of Increases in Mental Workload on Avoidance of Ground Hazards
New sensor and display technologies are expected to enhance the performance of soldiers by providing them more information about the battlefield. However, there is concern that greater quantities of information and increases in mental workload might cause distraction, reduce attention to dangers in the immediate environment, and threaten soldier survival. The purpose of this laboratory investigation was to quantify the effects of increases in mental workload on one of the soldier's most basic tasks --- avoiding ground hazards while walking. The participants were 12 U.S. Army infantry soldiers. The study was conducted on a treadmill that was modified to provide the participants a view of impending ground hazards up to 5 meters forward of their walking position. The study was a 2X3 fixed factor design with two levels of terrain difficulty (No Hazards and Hazards) and three levels of mental workload (No Load, Moderate load, and High load), all as within-subject effects. Mental workload was increased from the "No Load" to a "Moderate" level by requiring the participants to perform a mental arithmetic task while walking. Mental workload was increased from the "Moderate" to the "High" level of load by increasing the difficulty of arithmetic problems. The dependent variables included time and error in the performance of the mental arithmetic task, the mean and standard deviation in step length and step rate, the number of ground hazards contacted, and subjective ratings of workload. The participants" scores on the Armed Forces Qualification Test (AFQT) and subtests of the Armed Services Vocational Aptitude Battery (ASVAB) related to arithmetic skills were also obtained. The results of the investigation indicated that when the participants were required to avoid hazards, step length decreased and step rate increased, as was expected. Both measures of gait increased in variability. Subjective ratings of physical demand and effort obtained across the three levels of mental workload increased significantly, along with perceptions of workload associated with a perceived decline in performance. Subjective ratings obtained across the two levels of terrain difficulty indicated that ratings of mental demand and effort increased with each increase in level of mental workload. When the participants were confronted with the more difficult arithmetic problems at the "High" level of mental workload, time and error in performing the mental arithmetic task increased as did ratings of temporal demand, frustration, and workload attributable to a perceived decline in performance; however, subjective ratings of physical demand decreased. Interactions found between terrain difficulty and mental workload indicated that differences in ratings of performance and overall workload scores between the two levels of terrain difficulty decreased significantly between the "No Load" and the "Moderate" level of mental workload, and converged at the "High" level of mental load. Although relationships were found between perceived workload, gait, and performance of the mental arithmetic and hazard avoidance tasks, the analysis did not reveal a significant effect of mental workload on the number of hazards contacted. Some participants tended to contact more hazards at the "High" level of mental workload than at the "No Load" or the "Moderate" levels, as expected. However, other participants tended to contact more hazards at the "Moderate" level of mental load than at either of the two extremes. Still other participants tended to contact more hazards at the "No Load" level of mental workload than at the "Moderate" or the "High" levels. Correlations were found between subjective ratings of workload, mental arithmetic performance, and scores on the AFQT and subtests of the ASVAB related to arithmetic skills, but no relationships were found between test scores and performance of the hazard avoidance task. However, when test scores were used as covariates in the analysis of mental arithmetic performance, the findings revealed that the number of correct responses to the arithmetic problems decreased when the participants were required to avoid hazards. The results of the study may support the belief that the allocation of limited resources will vary based on past experience and other individual differences, and that the amount of resources allocated to a task may be influenced by the difficulty of the task, criteria for performance, and the motivation of the individual.