The subject population
In this experimental study, 15 male students of Hamadan University of Medical Sciences with an age range of 18-30 years were randomly selected. Pure tone audiometry was performed for selecting subjects with normal hearing. As shown in Fig.1, three common commercial earmuffs and three earplug models (one foam formable and two 3-flange pre-molded) with technical specifications from reliable international manufacturers used in the Iranian’s work environments were examined. A prototype of the proposed molded earplug designed based on subjects’ ear shape and size was also tested. The initial ear mold was made using a soft material and the final molded earplug was made from it using silicone materials in the lab. A ceramic filter was placed inside the molded earplug so that it can produce a special channel to allow transmit normal conversation. The inclusion criteria for participating in the study included having normal hearing and vision along with Persian native language.
Experiment procedure
In each experiment session, based on the REAT method, the hearing threshold of subjects was measured by a reference noise with and without HPDs for 40 minutes. Next, the speech intelligibility was tested without noise in acoustic room. Finally, the speech discrimination score (SDS) of subjects with and without HPDs was measured at two signal to noise (S/N) ratios for 30 minutes. As mentioned, based on the REAT method, pure tone audiometery was performed to measure a person's hearing threshold with and without HPDs [15]. In this way, based on the insertion loss of HPDs in the one-octave band frequency spectrum, the actual values of the noise reduction were calculated according to ANSI S3.19-1974 [16, 17, 18]. Participants are trained and supervised by the researchers to wear HPDs. The order of testing different types of HPDs was chosen randomly for REAT and SDS tests. All experiments were performed in an acoustic booth as shown in Fig. S1.
In the next step, using the reference white noise, the ambient background noise was fixed at 70 dBA, using speakers (Pejvak Ava CO) resembles a medium level of noise emission in some offices with high mental workload. Two signal (speech) to background noise ratios (0 and +5) were considered to relatively resemble the speech level of a speaker at a normal distance to a listener in the mentioned background noise (speech= 70, 75 dBA, noise= 70 dBA).
Speech intelligibility was measured based on the speech discrimination score and using a real two-channel audiometer (Piano model; Inventis CO). According to ISO 8253-3 standard, using a reliable and accurate list of monosyllabic Persian words (25 words), the subjects are asked to repeat the words played by the speaker in the room environment. One of the researchers recorded the answers at a distance of one meter from the listener. Then, the percentage of correctly repeated words is determined [19]. For speech audiometry, the speech of the Persian word from a suitable speaker with normal and clear speech with no particular accent was recorded [20].
Statistical analysis
Data were analyzed using SPSS software (ver.22, Chicago, IL, USA). The normality of the data was tested using the Kolmogorov-Smirnov test. When data were normally distributed, they were analyzed using the paired sample and student T-tests. The significant relation among some features was analyzed using Pearson correlation. Wilcoxon’s tests will also use when data distribution is not normal. The significant level for all tests was set at 5%.