The effect of bronchitis, smoking, and occupation on ventilation.

3 This method of statistical analysis is used throughout this paper and is fully explained in Principles of Medical Statistics by Bradford Hill. Briefly, if the observed difference between the two means is at least three times the standard error of the difference between the means, the possibility of this being a chance occurrence is one out of 370. This difference may be regarded as significant and will be so reported in this paper. This study was confined to men working full time. Men who had become disabled and had retired were excluded from the study. As a result, only the most fit subjects by chance may have been included. Selection of the men depended to a great extent on their availability, resulting in a haphazard sampling. However, there is no reason to believe that this has introduced any particular bias. Bronchitis was defined as daily cough for at least the preceding six months, productive of at least one teaspoonful of sputum per day. The number of cigarettes smoked per day was multiplied by the number of years during which the subject had smoked. For instance, a man smoking 20 cigarettes a day for 20 years would have a smoking value of 400. On this basis, the men were divided into four groups: