An inter-laboratory investigation of ANSI standard fitting protocols, sample size, subject and experimenter gender, and trial on the real-ear attenuation of two types of earplugs
Identical experiments were conducted between two acoustical-testing laboratories to determine the inter-laboratory differences of using two different hearing protection device (HPD) fitting procedures for testing the real-ear attenuation at threshold (REAT) of a popular vinyl foam earplug and a multi-sized premolded PVC single-flanged earplug. The first fitting procedure tested in the experiment is included in the revision of the American National Standards Institute (ANSI) standard S12.6-1984 by the ANSI Working Group ANSI S12/WG11, Field Effectiveness and Physical Characteristics of Hearing Protectors. This fitting procedure, “subject fit,” is intended to estimate “...the attenuation obtained in the top 10-20% of today’s industrial and military hearing conservation programs, i.e. the attenuation that should be obtained by an informed and motivated work force” (ANSI S12.6-199X, Draft 1.4, p. 4). The subject-fit procedure employs HPD-naive subjects, minimizes experimenter involvement, enforces subject-selection controls, and requires subjects to fit the HPD with reasonable comfort using only the manufacturer’s fitting instructions. The subject-fit method differs from the second procedure tested in this investigation, experimenter fit, in both procedure and objective. In the ANSI S3.19-1974 “experimenter-fit” method, which is the procedure currently required by the Environmental Protection Agency (EPA) for the testing and labeling of HPDs (EPA, 1990), the experimenter fits the HPD to the subject (comfort is not a consideration) to determine the optimum attenuation of the HPD. The development of the subject-fit protocol was motivated by the large discrepancy between the attenuation achieved in the field and that claimed by manufacturers of HPDs using experimenter fit from ANSI S3.19-1974. Some experts have developed schemes to derate manufacturers’ laboratory data to approximate attenuation typically achieved in the field.
In addition to investigating the differences between the two fitting protocols, other factors relevant to the revision of ANSI S12.6-1984 were studied: subject and experimenter gender effects, ear canal size effects, inter-laboratory differences, and the number of replications and subjects needed for REAT tests.
Results indicated that the subject-fit method provided statistically significantly less attenuation than the experimenter-fit method. Subject-fit tended to overestimate in-field attenuation, but not by as much as experimenter-fit. No consistent subject-gender effects were found in the analysis. Experimenter gender did not have a significant effect on subject-fit foam-earplug attenuation. The lack of significant trial effects indicated that the goodness of fit did not change for either fitting condition or across trials. Ear canal size and attenuation effects were documented with mixed results.