Page:NIOSH DM DFM respirator evaluation draft.pdf/51

This page has been proofread, but needs to be validated.
WORKING DRAFT 9.15.92—Performance Evaluation of DM and DFM Filter Respirators
43

would have no external validity. External validity includes topics such as possible non-sampling errors and biases[1] in WPF results. It requires that the respirator-use tasks and conditions of use are representative of actual conditions in typical workplaces. Unlike internal validity, for which there are objective statistical computations to justify conclusions, evaluating external validity is largely a subjective judgment.
  For those researchers that wish to generalize their respirator-performance study findings to larger groups, a two-stage process in involved during which external validity problems can arise.[2][3] First, researchers must define a target population of persons, settings, or times (e.g., efficacy of respirators worn by most users in the U.S. for specific respirator classes under actual working conditions of typical respirator programs). Second, researchers must draw respirator-wearer samples to represent these populations. However, samples usually cannot be drawn systematically in a formal randomized manner and are drawn instead because they are convenient and give an intuitive impression of representativeness. However, the settings and conditions of any given research study may severely hamper the generalizability of the results.
  Cook and Campbell have suggested that it is useful to distinguish between (1) target populations, (2) formally representative samples that correspond to known populations, (3) samples actually achieved in field research, and (4) achieved populations.[4] They have noted:

To criticize the study because the achieved sample of settings was not formally representative of the target population may appear unduly harsh in light of the fact that financial and logistical resources for the experiment were limited, and so sampling was conducted for convenience rather than formal representativeness. . . . it is worth noting that accidental samples of convenience do not make it easy to infer the target population, nor is it clear what population is actually achieved.[5]

  NIOSH has concluded that the majority of respirator-performance studies reported in the professional literature have not considered the uncertainty or margin or error

  1. ↑ Biases are effects that deprive a statistical result of representativeness by systematically distorting it.
  2. ↑ Bracht, G. H. and G. V. Glass: The External Validity of Experiments. Amer. Educ. Res. J. 5:437–474 (1968).
  3. ↑ Cook, T. D. and D. T. Campbell: Quasi-Experimentation-Design and Analysis Issues for Field Settings, Houghton Mifflin Company, Boston, MA (1979), pp. 70-80.
  4. ↑ Ibid., p. 71.
  5. ↑ Ibid., p. 71.
This article is issued from Wikisource. The text is available under Creative Commons Attribution-Share Alike 4.0 unless otherwise noted. Additional terms may apply for the media files.