It is statistically incorrect to use the changes of the most variable 5% of a population as a guideline to decide whether any individual’s spirometric response is meaningful, clinically significant, or SS because the variability in the population may not be normally distributed. Four relevant publications reported the mean and SD of FEV1 and FVC in their respective populations.2-5Table 1 illustrates the problem with extracts from original publications. To obtain the key volume and percent changes in each population, these authors reported the highest measured FEV1 and FVC of each subject from three preintervention (at baseline) and, often, postintervention (after rest,4 placebo,3,5 or drug2 administration) measurements. The reported population change in mean ± SD of FEV1 and FVC are shown in Table 1 (columns 1 and 2 and columns 3 and 4, respectively). Table 1 also shows the calculated 95% CLs (columns 5-8) that the study authors considered as the minimum changes of the most responsive 5% of the population. The CLs were calculated for the one-tailed or two-tailed z scores (1.65 and 1.96, respectively) times the SD of the preintervention (and sometimes postintervention values) and usually added the change in mean of the intervention.2,3,5 Each publication suggested that for another individual to have a positive bronchodilator response, the difference between the highest postintervention and preintervention ΔFEV1 or ΔFVC should exceed the 95% CL seen in the tested population (Table 1, columns 5-8). In summary, each study2-5 used its population change in mean ± SD to statistically calculate the overall abnormal, positive, or meaningful population changes of the most variable 5% of its population. Guidelines derived from the 95% CL1,2 of a population should not be used to define an individual positive responder because this would imply that only 5% of the individuals would be meaningful responders.