Variability with Time
In the end of the 1980s, a round-robin test was undertaken in France regarding the measurement and prediction of the spatial sound level decay in a room . Due to the high number of participating teams, the measurements were spread from June to November with the institute opening the process. On completion it was found that the results of the last months differed from the earlier ones. Being suspicious of the accuracy of the participating technicians involved, the institute performed a new set of measurements; it discovered that its results also differed! This was eventually traced to the hygrometry and temperature.
Lesson Learned: Always write down the measurement conditions.
An engineering outfit had to assess the noise levels generated by a large transportation corridor near its pet project. Due to the urgency, an occasional technician (i.e., an office helper who had been shown the basic sound level meter measurement procedure) was instructed to perform the measurements. On looking at the results, the engineers were puzzled: The sound level values were within 2 dB(A) of each other. Interrogating the technician, they eventually found that the measurements had been performed from the car due to the low temperature. One engineer then went for the kill: Had the engine been stopped during measurements? A dumbfounded technician then answered that one engine among so many others would not change much, would it?
Lesson Learned: Always explain the reasons for performing the measurements a given way; try to have a rough idea of the range in which the measurement results should be found.