A comparison between the three largest and three last of five forced expiratory maneuvers in a population study

Five acceptable forced expiratory maneuvers were obtained with a portable spirometer from each person in a population of 1,670 selected from a stratified random sample of a community. The largest 3 of 5 tracings were identifiable as a significantly different subser of the 5 tracings. In 83 per cent of comparisons, the variance of the last 3 was greater than that of the largest 3. Among the largest 3 tracings the highest single value was also identified as significantly different from the other two tracings. For epidemiologic purposes the largest 3 of 5 acceptable tracings is preferable to the last 3 of 5.