3.2. Evaluation of Psychoacoustic Recognition
As a result of the arousal changes caused by listening to fire alarms (
Figure 4A,B), the young adult males were evaluated at the highest arousal at the T-3 3100 Hz (mean: 18.15 out of 21 points). Bell (18.08 points), simulated T-3 3100 Hz (18.00 points) and slow whoop (17.54 points) followed high arousal. Although the difference in arousal level was not noticeable among the four fire alarms, T-3 520 Hz and simulated T-3 520 Hz showed relatively lower arousal, 15.77 and 13.77 points, respectively. For adult females, T-3 3100 Hz was reported as the highest arousal at 20.50 out of 21 points, followed by slow whoop (20.00 points), and T-3 520 Hz (17.92 points). Simulated T-3 3100 Hz, bell, and simulated T-3 520 Hz were evaluated with relatively low arousal at 17.42, 16.58, and 14.42 points, respectively. Commonly, the T-3 3100 Hz showed the highest arousal, and simulated T-3 520 Hz evaluated the lowest arousal. In terms of frequency of the fire alarm, the T-3 3100 Hz had relatively high, but the simulated T-3 520 Hz had a lower fundamental frequency. Additionally, as for the comparison with T-3 and simulated T-3, the simulated T-3 520 and 3100 Hz evaluated relatively lower arousal than T-3 520 and 3100 Hz in adults.
In the older adults, males evaluated the highest arousal for simulated T-3 3100 Hz at 18.69 of 21 points. Additionally, bell, T-3 3100 Hz and slow whoop followed with 17.69, 17.69 and 17.62 points, respectively. Similar to the young adult males, the T-3 520 Hz and simulated T-3 520 Hz, which had low fundamental frequency, evaluated relatively low arousal; 16.85 and 15.46 points. For older adult females, T-3 3100 Hz reported the highest arousal at 18.83 points, followed by bell, slow whoop and simulated T-3 3100 Hz; 18.42, 18.42 and 18.25 points, respectively. The older females also reported relatively low arousal for simulated T-3 520 Hz and T-3 520 Hz, 16.08 and 15.50 points, respectively.
As a result of three-way analysis variance, arousal had significant main effects depending on the type of fire alarm (F: 6.845,
p < 0.01), but the main effects of age (F: 0.063,
p = 0.801) and gender (F: 1.696,
p = 0.194) were not significant (
Table 3). The partial ŋ
2 of the type of fire alarm was 0.110, which indicates a moderate effect size. Additionally, the interactions among fire alarm, age- and gender-related variables were not significant. With post-hoc analysis of the type of fire alarm, T-3 3100 Hz, which showed the highest arousal, had significantly higher arousal than T-3 520 Hz and simulated T-3 520 Hz. Additionally, slow whoop, simulated T-3 3100 Hz and bell reported significantly higher arousal than simulated T-3 520 Hz. When simulating to the young adults for presbycusis effect, the results of a paired t-test of T-3 (520 and 3100 Hz) and simulated T-3 (520 and 3100 Hz) did not confirm a significant difference between 520 Hz (t: 1.457,
p: 0.171) and 3100 Hz (t: 0.187,
p: 0.855) in males. However, the females showed a significant difference between 520 Hz (t: 3.251,
p < 0.05) and 3100 Hz (t: 2.475,
p < 0.05).
In the case of urgency according to the type of fire alarm (
Figure 4C,D), adult males had the highest urgency at slow whoop (mean: 18.15 of 21 points), followed by bell, simulated T-3 3100 Hz, and T-3 3100 Hz as 17.08, 16.92, and 16.38 points, respectively. The T-3 520 Hz and simulated T-3 520 Hz which showed low fundamental frequency were evaluated as relatively low urgency as 15.23 and 10.54 points, respectively. An adult female showed the highest urgency with T-3 3100 Hz at 19.92 points, followed by slow whoop, T-3 520 Hz and simulated T-3 3100 Hz at 19.50, 15.67 and 15.43 points, respectively. In addition, bell and simulated T-3 520 Hz reported relatively low urgency at 13.58 and 12.08 points, respectively. Considering the adult male, the T-3 520 Hz, which had low fundamental frequency, evaluated relatively higher for females than males. In terms of older adults, the older adult male evaluated as the highest urgency (21.67 points) at simulated T-3 3100 Hz, followed by slow whoop, T-3 3100 Hz, and bell at 21.85, 21.62, and 20.85 points, respectively. Similar to the adult male, the older adult male showed relatively low urgency at T-3 520 Hz and simulated T-3 520 Hz (low fundamental frequency); 17.62 and 17.16 points, respectively. The older female evaluated the bell as the highest urgency at 21.67 points, followed by simulated T-3 3100 Hz, T-3 3100 Hz and slow whoop at 19.92, 19.83 and 19.33 points, respectively. Additionally, they reported T-3 520 Hz and simulated T-3 520 Hz (low fundamental frequency) as relatively low urgency at 16.83 and 14.58 points.
As a result of the three-way analysis variance, the main effects of type of fire alarm (F: 10.119,
p < 0.000) and age (F: 36.843,
p < 0.000) were significant, but the main effect of gender was not confirmed (F: 1.038,
p: 0.039;
Table 4). The partial ŋ
2 by type of fire alarm and age was identified at 0.115 and 0.118, respectively. This means a moderate or large effect size. Similar to arousal, the interaction of all variables was not significant. Specifically, the slow whoop and T-3 3100 Hz confirmed significantly higher than T-3 520 Hz and simulated T-3 520 Hz. Additionally, bell and simulated T-3 3100 Hz confirmed to significantly higher urgency than simulated T-3 520 Hz. As a result of post-hoc analysis for age, the older adults showed significantly higher urgency levels than the young adult group. In terms of presbycusis simulation, a paired t-test for T-3 (520 and 3100 Hz) and simulated T-3 (520 and 3100 Hz) confirmed significant differences (t: 3.184,
p < 0.05) in 520 Hz among males. However, 3100 Hz confirmed no significant differences (t: −0.442,
p: 0.666). For the female, significant differences were observed at both 520 Hz (t: 2.305,
p < 0.05) and 3100 Hz (t: 3.822,
p < 0.05).
As a result of immersion evaluation by listening to fire alarms, adult males gave the highest immersion (mean: 18.92 out of 21 points) for slow whoop (
Figure 4E,F). Additionally, T-3 520 Hz, T-3 3100 Hz and bell followed in order of high immersion; 16.62, 16.38, and 16.15 points, respectively. Simulated T-3 3100 Hz and 520 Hz induced relatively low immersion; 16.08 and 15.38 points. Adult females evaluated slow whoop (19.92 points) as the highest immersion, followed T-3 3100 Hz, simulated T-3 3100 Hz and T-3 520 Hz; 17.83, 17.50, and 15.75 points, respectively. Additionally, simulated T-3 520 Hz and bell were evaluated as a low immersion at 15.25 and 13.42 points, respectively. In terms of older adults, older males reported simulated T-3 3100 Hz (23.31 points) as the highest immersion, followed by T-3 3100 Hz, bell, and slow whoop; 23.23, 23.00, 22.62 points, respectively. Relatively, the T-3 520 Hz and simulated T-3 520 Hz were evaluated as low immersion; 21.92 and 21.77. In the older female, bell was evaluated as the highest immersion at 22.83 points, followed by simulated T-3 3100 Hz, T-3 520 Hz, and T-3 3100 Hz at 22.75, 22.33 and 21.75 points, respectively. The slow whoop and simulated T-3 520 Hz reported relatively low immersion at 21.50 and 20.67 points. Overall, immersion cannot confirm the tendency for the low fundamental frequency to correspond to low arousal and urgency.
As a result of the three-way analysis variance, immersion confirmed significant main effects of type of fire alarm (F: 2.867,
p < 0.05) and age (F: 182.481,
p < 0.001) as well as urgency (
Table 5). The partial ŋ
2 of type of fire alarm and age was 0.049 and 0.398, respectively. The type of fire alarm showed a small to moderate effect size, and age was identified as a very large effect size. Additionally, unlike other factors, immersion confirmed a significant interaction (F: 2.948,
p < 0.05) between the type of fire alarm and age with a small-to-moderate effect size (partial ŋ
2: 0.051). As a result of post-hoc analysis of the type of fire alarm, slow whoop, which showed the highest immersion, confirmed a significant difference with simulated T-3 520 Hz. The results of the paired t-test for immersion of T-3 and simulated T-3 showed no significant difference at both 520 Hz (t: 0.810,
p: 0.424) and 3100 Hz (t: 0.298,
p: 0.771). Similarly, no significant difference was found in both 520 Hz (t: 0.589,
p: 0.568) and 3100 Hz (t: 0.251,
p: 0.806).
3.3. Brain Activity by EEG Measurement
The present study applied EEG to objectively measure emotional changes expressed in response to listening to fire alarms. Based on the frequency band, EEG variables were divided into theta (4–8 Hz), alpha (8–13 Hz), SMR (12–16 Hz), beta (mid-beta; 16–21 Hz and high-beta; 21–30 Hz), and gamma (30–50 Hz) (please see
Figure 5).
3.3.1. EEG Data: Theta (4–8 Hz)
In the electrode placement, the theta power (4–8 Hz) over frontal electrodes (Fp1, Fp2, Fz, F3, F4, F7, F8) was greater during listening to fire alarms than other electrodes. As a result, changes of theta power on the listening condition for adult males were highly decreased at T-3 520 Hz (−11.61), bell, T-3 3100 Hz, simulated T-3 520 Hz, and slow whoop also decreased with −5.78, −3.75, −3.38 and −1.69, respectively. On the other hand, simulated T-3 3100 Hz increased in listening conditions than background at 1.16. In terms of adult females, the theta of adult females increased in listening conditions for all fire alarms. Specifically, slow whoop increased the most at 12.62, followed T-3 520 Hz, and simulated T-3 520 Hz at 11.09 and 7.78. The T-3 3100 Hz, simulated T-3 3100 Hz and bell, which had high fundamental frequency, measured relatively small increases at 3.90, 1.19 and 0.15, respectively. In older adults, older males showed increased theta power similar to adult females. Specifically, the T-3 520 Hz increased the most at 18.65, followed by simulated T-3 3100 Hz and T-3 3100 Hz, simulated T-3 520 Hz, and T-3 520 Hz at 14.87, 13.87, 11.94 and 10.81, respectively. The slow whoop measured the least increase at 9.02. In older females, the change of theta power differentiated according to the fire alarm. Simulated T-3 520 Hz, T-3 3100 Hz and slow whoop increased at 17.44, 6.97 and 1.20, respectively. On the other hand, the bell, simulated T-3 3100 Hz and T-3 520 Hz decreased at −18.47, −10.91 and −3.60, respectively. The statistical analysis of theta according to type of fire alarm, age and gender showed that all variables had no significant effect. In other words, the three variables did not have a statistically significant effect on the change in theta. In terms of interaction, interactions between age and gender were significant (F: 8.960, p < 0.05) and showed a small effect size (partial ŋ2: 0.030). Additionally, other significant interactions between variables were not founded.
3.3.2. EEG Data: Alpha (8–13 Hz)
As a result of confirming the distribution of the alpha power according to the placement of the electrode, the younger group was weak, but the changes of the alpha power increased as the electrode progressed to the occipital area of the brain. However, the overall tendency showed widely distributed responses in all electrodes. The older group also found that the alpha power variations were widely distributed regardless of the placement of the electrodes. The changes in alpha power in listening conditions tended to decrease regardless of age and gender. Specifically, for young males, the alpha decreased the most at −18.56 at simulated T-3 3100 Hz, followed by −16.56 and −15.70 at bell and T-3 520 Hz, respectively. T-3 3100 Hz and slow whoop reduced alpha power at −14.91 and −12.99, and simulated T-3 520 Hz showed the least reduction (−11.02). Young females, similar to males, measured the most reduction in alpha at simulated T-3 3100 Hz (−8.26), followed by T-3 3100 Hz and bell at −5.71 and −4.61. Subsequently, the simulated T-3 520 Hz and slow whoop also decreased to −3.72 and −3.09 respectively; the lowest reduction in alpha was T-3 520 Hz at −2.44. For older males, the slow whoop decreased the most in alpha at −3.04, followed by simulated T-3 520 Hz and T-3 520 Hz at −2.90 and −2.74, respectively. Additionally, the bell and simulated T-3 3100 Hz decreased at −2.43 and −2.08, and the T-3 3100 decreased the least at −1.80. Older females showed the largest decrease in alpha at −6.74, followed by simulated T-3 3100 Hz and simulated T-3 520 Hz at −5.08 and −2.71. T-3 520 Hz and slow whoop, also decreased, at −2.67 and −2.65. The lowest reduction in alpha was identified at T-3 3100 Hz (−1.79), the same as in older males. Statistical analysis of changes in alpha according to listening condition showed that the main effect of type of fire alarm was not significant (F: 0.516, p: 0.764). In other words, the acoustic characteristics of the fire alarms did not significantly affect the changes in alpha power. However, the main effects of age (F: 20.014, p < 0.000) and gender (F: 9.329, p < 0.01) were significant. These results suggest that age and gender were affected by alpha wave changes, with small-to-moderate effect size (partial ŋ2: 0.064 and 0.031, respectively). In addition, significant interaction between age and gender was identified (F: 14.358, p < 0.000), with small-to-moderate effect size (partial ŋ2: 0.047). The post-hoc analysis confirmed a significant difference (young adult: −9.80, older adult: −3.30). It also showed a significant difference for gender (male: −8.96, female: −4.16).
3.3.3. EEG Data: Beta (12–30 Hz); SMR (12–16 Hz), Mid-Beta (16–21 Hz) and High Beta (21–30 Hz)
In the present study, the beta power (12–30 Hz) was subdivided as a frequency domain, such as SMR (sensorimotor rhythm, 12–16 Hz), mid-beta (16–21 Hz) and high-beta (21–30 Hz). As a result of confirming the distribution of the beta power according to the electrode placement, the sub-beta powers were distributed widely regardless of electrode placement.
For the SMR power, the young male decreased in all listening conditions regardless of the type of fire alarm. Specifically, bell measured the most decrease to SMR power at −1.79, and simulated T-3 3100 Hz, and T-3, 3100 Hz followed at −1.46 and −1.45, respectively. Slow whoop and simulated T-3 520 Hz also decreased at −1.23 and −1.11, respectively, and the T-3 520 Hz identified as a least decreased fire alarm at −1.01. For young females, all fire alarms except T-3 520 Hz identified a decrease. Specifically, slow whoop showed the greatest decrease (−1.01). Simulated T-3 3100 Hz, bell, T-3 3100 Hz, and simulated T-3 (520 Hz) were also decreased; −0.99, −0.79, −0.75, and −0.11, respectively. On the other hand, the T-3 520 Hz increased to 0.27 based on the listening condition. In older males, it was observed an increasing tendency, except slow whoop, unlike young adults, who showed a decreasing tendency to SMR. Among them, the SMR increased the most at 1.12 in bell, followed by T-3 3100 Hz, simulated T-3 520 Hz, T-3 520 Hz and simulated T-3 3100 Hz at 0.78, 0.54, 0.30 and 0.27, respectively. In contrast, the slow whoop was measured at −0.16, which was the SMR power decreased. The older females, similar to young adults, confirmed the tendency to decreased SMR power. Specifically, bell and simulated T-3 3100 Hz showed a relatively high decrease with −1.61 and −1.45, followed by slow whoop and simulated T-3 520 Hz; −1.36 and −1.35. In contrast to the previous fire alarms, the T-3 520 Hz increased the SMR power at 0.27. As a result of statistical analysis for the changes of the SMR power according to the listening condition, the main effect for age was significant (F: 32.882, p < 0.05) with small effect size (partial ŋ2: 0.02). However, the other main effects according to the type of fire alarm (F: 32,882, p: 0.521) and gender (F: 2.218, p: 0.137) were not significant. Nonetheless, significant interactions were confirmed for age and gender (F: 111.881, p < 0.000) with moderate effect size (partial ŋ2: 0.65). As a result of post-hoc analysis according to age, a significant difference (young adult: −0.96, older adult: −0.28) was confirmed within age.
In the mid-beta, the young males showed a tendency to decreased mid-beta power. Specifically, simulated T-3 3100 Hz was the largest decrease at −1.12, followed by T-3 3100 Hz, bell and simulated T-3 520 Hz and T-3 520 Hz at −0.82, −0.81, −0.58 and −0.48, respectively. In particular, slow whoop showed the least decrease. For young females, the T-3 3100 Hz and simulated T-3 3100 H decreased the mid-beta power at −0.19 and −0.08, respectively. However, bell, T-3 520 Hz, slow whoop and simulated T-3 520 Hz tended to increase the mid-beta power at 0.65, 0.53, 0.14 and 0.12, respectively. For older males, slow whoop only decreased at −0.17 in the mid-beta power, and other fire alarms were increased. T-3 3100 Hz showed the largest increase was 0.81, and bell, simulated T-3 520 Hz and T-3 520 Hz followed at 0.51, 0.47 and 0.30, respectively. In addition, simulated T-3 3100 Hz showed the smallest increase at 0.05. Older females, similar to young males, decreased in all fire alarms. Specifically, simulated T-3 3100 Hz, and bell decreased to −1.02 and −0.99 respectively. Additionally, slow whoop, simulated T-3 520 Hz, T-3 3100 Hz and T-3 520 Hz were decreased to −0.45, −0.44, −0.28 and −0.22, respectively. Statistical analysis of changes in mid-beta according to type of fire alarms, age and gender showed that all main effects of type of fire alarm, age and gender were not significant. It means that mid-beta variations had not been affected by the type of fire alarm, age and gender. In terms of interaction, significant interactions between age and gender were identified (F: 19.634, p < 0.000), with a moderate effect size (partial ŋ2: 0.063). Additionally, all post-hoc analyses were not significant.
As a result of measuring high-beta power for young males, high-beta showed a tendency to reduce in all fire alarms. The slow whoop decreased the most at −0.60, followed by simulated T-3 3100 Hz, simulated T-3 520 Hz and bell at −0.46 −0.34 and −0.33, respectively. For young females, it was found, contrary to males, that all fire alarms tended to increase in high-beta. Specifically, T-3 520 Hz had the largest increase at 1.45, while bell had the smallest increase. Other fire alarms were measured in order T-3 3100 Hz, simulated T-3 520 Hz, and simulated T-3 3100 Hz; 1.10, 1.05, 0.73 and 0.65, respectively. In the case of older males, it was confirmed the high-beta increased with all fire alarms, similar to young females. The bell showed the largest increase at 1.98, and slow whoop showed the smallest increase at 0.26. For other fire alarms, T-3 3100 Hz, simulated T-3 520 Hz, T-3 520 Hz, and simulated T-3 3100 Hz measured at 1.83, 1.41, 1.05 and 0.87, respectively. For older females, T-3 3100 Hz increased at −0.62, but all other fire alarms showed decreased tendency. Specifically, simulated T-3 3100 Hz measured the largest decrease at −0.49, followed by bell, slow whoop, simulated T-3 520 Hz and T-3 520 Hz at −0.41, −0.26, −0.24 and −0.07, respectively. Statistical analysis of changes in high-beta showed no significant main effects in all variables. The interaction between age and gender (F: 30.520, p < 0.000) showed only significant results with large effect size (partial ŋ2: 0.094). As a result of post-hoc analysis, there were no significant differences in all variables.
3.3.4. EEG Data: Gamma (30–50 Hz)
Finally, the gamma power according to the type of fire alarms, age and gender indicates that gamma power increased for young adults regardless of the type of fire alarm and gender. Specifically, young males showed the largest increase at T-3 520 Hz (0.96) and the smallest increase at T-3 3100 Hz (0.04). Other fire alarms measured, in order of bell, simulated T-3 520 Hz, simulated T-3 3100 Hz, and slow whoop, at 0.50, 0.39, 0.37 and 0.29, respectively. Young females also measured the largest increase at simulated T-3 3100 Hz (2.26) and the smallest increase at simulated T-3 520 Hz (0.89). Other fire alarms measured, in order of T-3 520 Hz, T-3 3100 Hz, slow whoop, and bell, at 1.87, 1.65, 1.24 and 0.98, respectively. Older males, similar to young adults, tended to have increased gamma power on all fire alarms. The highest increase was 3.19 at bell, and the smallest increase was 1.08 at slow whoop. The other fire alarms, which are T-3 3100 Hz, simulated T-3 520 Hz, T-3 520 Hz, and simulated T-3 3100 Hz, were measured at 2.80, 2.69, 1.95 and 1.18, respectively. In older females, slow whoop was decreased only at −0.24, but the other fire alarms were increased. Specifically, T-3 3100 Hz measured the highest increase at 1.86, followed by simulated T-3 3100 Hz, T-3 520 Hz, simulated T-3 520 Hz and bell at 0.77, 0.67, 0.61, and 0.11, respectively. The statistical analysis of changes in gamma power did not show main effects in all variables, similar to high-beta. In terms of interaction, a significant interaction between age and gender (F: 12.074, p < 0.001) was confirmed with p: 0.039). Additionally, post-hoc results were not found to be significant.