6. SAMPLING STRATEGY
To obtain the maximum amount of information during the course of a survey with a minimum number of samples, a statistical sampling strategy should be developed before conducting any survey [16]. Several pieces of information must be known in advance to plan a sampling strategy, including the size of the workforce to be sampled, the accuracy of the sampling and measurement method to be used and the confidence one wishes to have in predicting the exposure of the workforce.
For example, to determine with 90% confidence that at least one worker from a workplace subgroup will be in the top 10% of the exposures occurring in the group, the number of employees to sample would be chosen from Table 4. Other tables are given in the publication for confidence limits of 95% and for the top 20% of exposures. Again, judicious use of sampling statistics will optimize the number of samples needed.
Table 4. Minimum sample size (n) for including (@ 90% confidence level) at least one high risk employee* [16]
Size of Employee Group (N):
1 2 3 4 5 6 7 8 9 10 11-12 13-14 15-17 18-20 21-24 25-29 30-37 39-49 50 4
Minimum Number of Measured Employees (n):
1 2 3 4 5 6 7 7 8 9 10 11 12 13 14 15 16 17 18 22
- Exposure in highest 10% of N.
7. SAMPLING AND CALIBRATION TECHNIQUES
The following are suggested general techniques for active sampling using some of the more common samplers. These instructions elaborate on those given in NMAM methods. Consult individual methods for details of sample size.
a. Calibration of Personal Sampling Pumps
The accuracy of determinating the concentration of a toxic substance in air is no greater than the accuracy with which the air volume is measured. Therefore, accurate calibration of the airflow rate through the sampling train is necessary. The frequency of calibration depends on the use, care and handling to which the pump is subjected. In addition, pumps must be recalibrated after each repair and if they have been abused. Ordinarily, pumps should be calibrated in the laboratory and the field, both before field use and after each field survey.
The choice of a reference instrument will depend on where the calibration is to be performed. For laboratory use, primary standards, such as a spirometer or soap-bubble meter, are recommended [17]. Several electronic soap-bubble calibrators and one dry-cell calibrator are commercially available as primary calibrators. Other instruments, such as a wet-test,
