Statistical and Research Design Considerations Concerning Industrial Respirator Research.
NIOSH 1980 Sep:247-256
Problems in past investigations of the physiological and psychological aspects of respirator wear are reviewed and recommendations for improving research and statistical analyses in these areas are presented. Basic differences in the approaches of statisticians and physiologists to research involving respirator wear are defined. Recommendations with respect to research design, delineation of variables, and use of previous research and findings are presented. Analysis strategies to determine the percentage changes in the physiological and psychological functions and the relationships between laboratory measurements and actual work data are discussed. The authors conclude that coordinated, cooperative, large sample studies are needed to determine the effects of industrial respirators on work performance and that sophisticated multivariate statistics applied to data gathered from large samples provide the best possibility or obtaining definitive research answers.
Medical-research; Analytical-methods; Equipment-design; Respiratory-protective-equipment; Safety-engineering;
Papers from the NIOSH International Respirator Research Workshop, September 9-11, 1980, Division of Safety Research, NIOSH, Morgantown, WV