2. Methodology
2.1. Sample selection practices
Even though a sample of about two persons per thousand can statistically describe the country’s population within a confidence limit of 90% [6], a much greater sample was selected because of the inherent variation of noise exposure and the expected high rejection rate due to possible pre-existing noise-induced hearing problems. More specifically, a sample of about 1% — five times the statistically acceptable norm, of the industrial working population was selected for participation to allow for those two factors and reduce the risk of biased measurements. The distribution of the workers in their respective industrial sector was obtained from official statistics [7], while the selection of each one of the workers within the same industry was random. Each one of the selected workers had to fill out a questionnaire providing information about his/her age, years at work, habits, military training and other factors that allowed the exclusion of people with pre-existing conditions from further testing. Based on those answers people from the following groups were excluded from the audiometric tests: (1) people with known hearing diseases, (2) people who did their military service in artillery, (3) hunting and gaming sportsmen [1], and (4) people who suffered head injuries/accidents.
Table M1, shows the distribution of people selected to participate in this project as well as the number of people tested during the various phases of the plan.