Statistical analysis of indoor bacterial air concentration and comparison of four RCS Biotest samplers

The bacterial air concentration expressed as number of colony-forming units (CFU) per m 3 is widely used as a microbiological parameter in indoor air quality (IAQ) assessments. This paper presents a study case for an office area performed by four independent teams using four identical air-sampling instruments. The potential sources of error are identified, and a statistical analysis of their influence upon the data reproducibility is presented. Like in many other environmental measurements, it results that the log-normal distribution fits the data better than the normal distribution. It is concluded that the log-normal distribution should be used in the statistical hypotheses testing, in the design of the experimental programs, and in the computation of the confidence intervals.