Effects of Auditory Pre-Stimulation on Cognitive Task Performance in a Noisy Environment

Abstract

The accident rate due to human errors in industrial fields has been consistently high over the past few decades, and noise has been emerging as one of the main causes of human errors. In recent years, auditory pre-stimulation has been considered as a means of preventing human errors by improving workers’ cognitive task performance. However, most previous studies demonstrated the effectiveness of the auditory pre-stimulation in a quiet environment. Accordingly, studies on the effects of pre-stimulation in a noisy environment are still lacking. Therefore, this study aimed to empirically investigate: (1) the effects of noisy environments on the performances of cognitive tasks related to different functions of working memory and (2) the effects of auditory pre-stimulation on the performances of cognitive tasks in a field-noise environment. To accomplish these research objectives, two major experiments were conducted. In the first experiment, a total of 24 participants performed each of three basic short-term/working memory (STM/WM) tasks under two different experimental conditions (quiet-noise environment and field-noise environment) depending on the presence or absence of field noise. In the second experiment, the participants performed each of the three basic STM/WM tasks in a field-noise environment after they were provided with one of four different auditory pre-stimulations (quiet noise, white noise, field noise, and mixed (white and field) noise). The three STM/WM tasks were the Corsi block-tapping, Digit span, and 3-back tasks, corresponding to the visuospatial sketchpad, the phonological loop, and the central executive of WM, respectively. The major findings were that: (1) the field-noise environment did not affect the scores of the Corsi block-tapping and 3-back tasks, significantly affecting only the Digit span task score (decreased by 15.2%, p < 0.01); and (2) the Digit span task performance in the field-noise environment was improved by 17.9% (p < 0.05) when mixed noise was provided as a type of auditory pre-stimulation. These findings may be useful for the work-space designs that prevent/minimize human errors and industrial accidents by improving the cognitive task performance of workers in field-noise environments.

1. Introduction

During the past few decades, technological evolution has drastically changed the nature of the human factor problems that one faces in relation to industrial safety. Analyses of major accidents have concluded that human errors on the part of operators, designers, or managers have played a major role in their occurrence [1]. Bhavsar et al. [2] argued that human error was one of the major reasons for industrial accidents, emphasizing the need to introduce new technologies to prevent it. Dinges [3] also stated that unintentional human errors in the workplace, which can include mistakes by operators, maintenance, and management, were the most frequently identified root causes of accidents, contributing significantly to between 30% and 90% of all serious incidents across industries [4,5]. In a similar vein, Hals [6] found that human error was a primary causal factor in 70–80% of accidents in the oil and gas industry. According to Park et al. [7], the industrial accident rate due to human error was found to be high in Korea as well; out of 351 cases of serious accidents that occurred in Korea in 2013, 117 (33.7%) were related to human error. Hence, it is known that human error remains the most common contributing factor in fatal accidents worldwide, and the ineffective mental workload of operators plays a role in such accidents [8,9]. Indeed, a number of studies have emphasized that an increase in a worker’s cognitive workload or a decrease in their cognitive task performance can lead to human errors and serious industrial accidents [10,11,12].

Noise has been emerging as one of the main causes of negative cognitive task performance in industrial sites. Although some studies showed the positive effects of noise on cognitive performance [13,14,15,16,17,18], multiple studies have demonstrated the negative effects of noise. Fernández et al. [19] noted that noise is one factor that can increase the risk of accidents in the workplace by interfering with communication. They also claimed that noise can contribute to indirectly increasing the rate of accidents by causing stress, attention loss, and blood pressure increases, etc. Stansfeld and Matheson [20] revealed that children exposed to chronic environmental noise had poorer auditory discrimination and speech perception as well as poorer memory requiring high processing demands. In another major study, Jahncke et al. [21] stated that noise such as irrelevant speech impairs performance in proofreading [22,23], serial recall [24], mental arithmetic [25,26], reading comprehension, operation span, and tasks activating prior knowledge from long-term memory [27]. Monteiro et al. [28] have demonstrated the interactive effect of occupational noise on attention and short-term memory. More recently, several review studies [29,30] have revealed the harmful effect of environmental noise (road traffic, aircraft, and train and railway noise) on cognition.

As noted above, noise can decrease cognitive task performance and is thus a major cause of human errors. At industrial sites, noise is inevitable, and field workers are continuously exposed to it. According to 1999–2004 National Health and Nutrition Examination Survey (NHANES) data [31], hazardous noise is one of the most common occupational hazards in the United States, with over 22 million workers exposed. In relation to this survey, Kerns et al. [32] showed that the industries with the highest prevalence of self-reported occupational noise exposure were construction (51%), manufacturing (47%), and transportation and warehousing (40%). In France, a survey by the Ministry of Employment [33] indicated that approximately 7% of employed workers are exposed to excessive noise levels (more than 85 dB (A) for at least 20 h per week) and that close to 25% are exposed to hazardous noise (more than 85 dB (A) but for less than 20 h per week) [34]. In Korea, the average noise exposure level at 16,359 workplaces in 2015 was 83.6 dB (A), and more than half of the noise measurements exceeded the exposure limits in 13.7% of the workplaces tested [35]. In light of the fact that noise is prevalent in most industrial sites, as mentioned above, it is important to devise measures to mitigate the adverse noise effects on cognitive task performance in order to prevent human errors and better ensure industrial safety.

Working memory (henceforth referred to as WM) is the basis of fundamental cognitive functions, such as rehearsal, reasoning, decision making, action planning, problem solving, situation awareness, and perceptual processing (bottom-up/top-down processing) [36,37,38]. In addition, WM has been found to be associated with the occurrence of various types of human errors [4,38,39]. According to Baddeley’s model [40], the WM system consists of three subcomponents, that is, the visuospatial sketchpad, the phonological loop, and the central executive. The visuospatial sketchpad and the phonological loop are short-term memory (henceforth referred to as STM) components of WM, as they are intended to temporarily hold/process visuospatial and auditory-verbal information, respectively, and the central executive provides executive control of WM by monitoring and manipulating the STM components. Note that a fourth component to the revised model [41], the episodic buffer which provides a temporary interface between the slave system (the phonological loop and the visuospatial sketchpad) and long-term memory, was not considered in this study since it is assumed to be controlled by the central executive [41].

As noted earlier, it is known that cognitive task performance can be reduced in a noisy environment [22,23,24,25,26,27]. However, to the best of the authors’ knowledge, less attention has been paid to how a noisy environment affects the performance of basic STM/WM tasks related to different functions of WM. This lack of research and knowledge hampers the ergonomic designs of work tasks and environments to minimize the adverse effects of noise at noisy industrial sites, which can eventually result in a decrease in cognitive task performance and thereby an increase in the rate of human errors/accidents.

In an effort to fill the above-mentioned research gap, this paper seeks to address the following unexplored research question:

Research Question 1. How will a noisy environment affect the performance of basic STM/WM tasks related to different functions of WM?

In recent years, numerous scholars have demonstrated the effectiveness of auditory pre-stimulation as one way to improve workers’ cognitive task performance. Several studies revealed that listening to classical music prior to a variety of cognitive tasks significantly increased the cognitive task performance by improving the levels of arousal or positive affect in the listeners [42,43,44,45]. Lee et al. [46] also demonstrated that listening to white noise in advance of a task helped to improve the work efficiency of workers by making them feel emotionally stable and improving their concentration. Hong and Kim [47] demonstrated positive effects of listening to meditation music programs before a class on attention and learning attitudes among college students. Jang and Lee [48] investigated the effects of different musical genres on mental focusing; their study showed that a group that listened to binaural beats prior to an attention test showed higher mental focusing than those who listened to classical or popular music. Judging from the results of these studies, it appears that providing workers with appropriate auditory pre-stimulation can contribute to improving their cognitive task performance, thereby preventing human errors.

Despite earlier work, however, our understanding of the effects of auditory pre-stimulation on cognitive task performance in noisy environments is very limited due to insufficient relevant research. Most previous studies did not consider situations in which workers at industrial sites are inevitably exposed to noise. Thus, it remains unknown whether the beneficial effects of auditory pre-stimulation demonstrated in the previous studies would occur even in situations where noise exists. This lack of knowledge serves as an obstacle to devising measures to reduce the decline in cognitive performance caused by noise and the resulting human errors. Therefore, in an effort to address the knowledge gap, this paper aimed to address the following second research question:

Research Question 2. How will auditory pre-stimulation affect the performance of basic STM/WM tasks in a noisy environment?

To address Research Question 1 and Research Question 2, two experiments were conducted. In Experiment 1, the effects of a noisy environment on the performance of basic STM/WM tasks related to different functions of WM were examined. Three basic STM/WM tasks related to WM functions were considered. They were the Corsi block-tapping [49], Digit span [50], and 3-back [51] tasks. Participants performed the three STM/WM tasks and mental workload ratings under two experimental conditions (quiet-noise and field-noise environments). Cognitive task and mental workload rating scores obtained under the two conditions were compared. In Experiment 2, the effects of auditory pre-stimulation on the performance of basic STM/WM tasks in a field-noise environment were investigated. After the participants were provided with each of the four different auditory stimuli (quiet noise, white noise, field noise, and mixed (white and field) noise) in advance, they performed the three basic STM/WM tasks and mental workload ratings in the field-noise environment. The cognitive task and mental workload rating scores obtained under the four conditions with different types of auditory pre-stimulation were compared with those obtained under the condition in which no auditory pre-stimulation was presented.

2. Methods

2.1. Participants

A total of 24 participants in their 20 s (11 males and 13 females) participated in the experiment. Their mean age was 24.4 years (range: 21~29; standard deviation: 2.1). These participants did not suffer from any cognitive disorders or hearing loss/impairment. Each participant signed an informed consent form prior to the participation, and the experiment was carried out in accordance with the Declaration of Helsinki. All the participants were given plenty of time for rest between trials in consideration of recovery and protection from potential fatigue/injury. In addition, they were allowed to immediately cease the participation whenever feeling any discomfort/inconvenience during the experiment.

2.2. Study Design

2.2.1. Experiment 1

In Experiment 1 (to examine the effects of a noisy environment on the performance of basic STM/WM tasks), the participants performed the three STM/WM tasks without being provided with auditory pre-stimulation. Depending on the presence or absence of noise in the experimental environment, the experimental conditions were classified into the following two categories (denoted here as Conditions 1 and 2).

Condition 1. An experimental condition for performing tasks in Quiet-noise Environment (${\mathit{E}}_{\mathit{Q}}$) without auditory pre-stimulation (${\mathit{P}}_{\mathit{N}}$) (henceforth referred to as ${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{Q}}$).

Here, ${\mathit{E}}_{\mathit{Q}}$ refers to a general laboratory environment without any additional noise, with the noise intensity level set to 40~45 dB (A). This corresponds to the noise level in workplaces that involve mental concentration (e.g., library, office) (BS 8233:2014). The ${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{Q}}$condition was considered as an environment for measuring the basic cognitive ability of each participant.

Condition 2. An experimental condition for performing tasks in Field-noise Environment (${\mathit{E}}_{\mathit{F}})$ without auditory pre-stimulation (${\mathit{P}}_{\mathit{N}}$) (henceforth referred to as ${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{F}}$).

${\mathit{E}}_{\mathit{F}}$was an environment in which field noise prevalent in actual workplaces was provided. Field noise includes noise both from manufacturing sites (noise produced by the direct recording and mixing of sounds from machines/tools such as grinders, sandpaper machines, and hammers, for instance) and from construction sites (noise produced by the direct recording and mixing of sounds from machines/tools such as vehicles and drills, among others). In this case, the noise intensity level was set to 80 dB (A), which corresponds to a minimal action level of hearing protection (for an eight-hour workday) as stipulated by the EU (EU Directive 2003/10/EC). The ${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{F} }$ condition was regarded as a typical working environment for industrial workers.

2.2.2. Experiment 2

In Experiment 2 (to examine the effects of auditory pre-stimulation on the performance of basic STM/WM tasks in a noisy environment), the participants performed the three STM/WM tasks in Field-noise Environment (${\mathit{E}}_{\mathit{F}}$) after they were provided with auditory pre-stimulation for five minutes. According to the type of auditory pre-stimulation provided, the experimental conditions were classified into the following four categories (denoted here as Conditions 3 through 6).

Condition 3. An experimental condition for performing tasks in Field-noise Environment (${\mathit{E}}_{\mathit{F}}$) with the auditory pre-stimulation of Quiet noise (${\mathit{P}}_{\mathit{Q}}$) (henceforth referred to as ${\mathit{P}}_{\mathit{Q}} {\mathit{E}}_{\mathit{F}}$).

Quiet noise was chosen as the type of pre-stimulation on the basis of the results of Mohan et al. [52], who found that quiet sitting with meditation improved cognitive function/performance. The noise intensity level of Quiet noise was 40~45 dB (A), which corresponds to that of quiet workplaces that involve mental concentration.

Condition 4. An experimental condition for performing tasks in Field-noise Environment (${\mathit{E}}_{\mathit{F}}$) with the auditory pre-stimulation of White noise (${\mathit{P}}_{\mathit{W}}$) (henceforth referred to as ${\mathit{P}}_{\mathit{W}} {\mathit{E}}_{\mathit{F}}$).

White noise was adopted as the type of pre-stimulation for this condition based on the previous studies indicating that it may effectively improve cognitive task performance [46,53,54]; it was provided by Genius Mate P model, which is a white noise generator. The noise intensity level of white noise was set to 60 dB (A) based on the results of Eschenbrenner [54] that low-intensity white noise helped to improve cognitive task performance.

Condition 5. An experimental condition for performing tasks in Field-noise Environment (${\mathit{E}}_{\mathit{F}}$) with auditory pre-stimulation of Field noise (${\mathit{P}}_{\mathit{F}}$) (henceforth referred to as ${\mathit{P}}_{\mathit{F}} {\mathit{E}}_{\mathit{F}}$).

Field noise was chosen as the type of pre-stimulation on the basis of Smith et al. [55], who showed that the effects of office noise could be reduced by prior exposure to this type of noise; the noise intensity level was set to 80 dB (A), as in ${\mathit{E}}_{\mathit{F}}$.

Condition 6. An experimental condition for performing tasks in Field-noise Environment (${\mathit{E}}_{\mathit{F}}$) with auditory pre-stimulation of Mixed noise (${\mathit{P}}_{\mathit{M}}$) (henceforth referred to as ${\mathit{P}}_{\mathit{M}} {\mathit{E}}_{\mathit{F}}$).

Mixed noise was literally a mixture of Field noise and White noise. It was selected as the type of pre-stimulation based on the results of existing studies that cognitive task performance could be improved in an environment with field noise masked with white noise [35,56,57]. The noise intensity level in this condition was set to 80 dB (A) for both the Field noise and White noise.

Table 1. Experimental conditions for Experiment 1 and Experiment 2.

Experiment	Condition	Pre-Stimulation	Environment
1	1. ${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{Q}}$	Non	Quiet-noise
	2. ${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{F}}$		Field-noise
2	3. ${\mathit{P}}_{\mathit{Q}} {\mathit{E}}_{\mathit{F}}$	Quiet noise	Field-noise
	4. ${\mathit{P}}_{\mathit{W}} {\mathit{E}}_{\mathit{F}}$	White noise
	$5. {\mathit{P}}_{\mathit{F}} {\mathit{E}}_{\mathit{F}}$	Field noise
	6. ${\mathit{P}}_{\mathit{M}} {\mathit{E}}_{\mathit{F}}$	Mixed noise

P (Pre-stimulation): ${\mathit{P}}_{\mathit{N}}$ (Non), ${\mathit{P}}_{\mathit{Q}}$ (Quiet noise), ${\mathit{P}}_{\mathit{W}}$ (White noise), ${\mathit{P}}_{\mathit{F}}$ (Field noise), ${\mathit{P}}_{\mathit{M}}$ (Mixed (White and Field) noise). E (Environments): ${\mathit{E}}_{\mathit{Q}}$ (Quietnoise), ${\mathit{E}}_{\mathit{F}}$ (Field-noise).

summarizes all of the experimental conditions for Experiment 1 and Experiment 2.

2.3. Experiment Variables

The independent variable was the experimental condition; it had two levels (${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{Q}}$and ${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{F}})$ in Experiment 1 and five levels (${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{F}}$, ${\mathit{P}}_{\mathit{Q}} {\mathit{E}}_{\mathit{F}}$, ${\mathit{P}}_{\mathit{W}} {\mathit{E}}_{\mathit{F}}$, ${\mathit{P}}_{\mathit{F}} {\mathit{E}}_{\mathit{F}}$, and ${\mathit{P}}_{\mathit{M}} {\mathit{E}}_{\mathit{F}}$) in Experiment 2. The scores of the three STM/WM tasks and the mental workload ratings were employed as the dependent variables in this study. Detailed explanations of the STM/WM tasks and mental workload ratings are given below.

2.3.1. STM/WM Tasks

The current study employed three basic STM/WM tasks: the Corsi block-tapping [49], Digit span [50], and 3-back [51] tasks. The three STM/WM tasks differ in terms of how they load the WM system, as they involve different subsystems of Baddeley’s WM model [40]. The Corsi block-tapping, the Digit span, and the 3-back tasks correspond to the visuospatial sketchpad, the phonological loop, and the central executive subsystem, respectively.

The Corsi block-tapping task [49] was adopted to evaluate each participant’s visuospatial STM performance, as this task has been widely used to test the function of the visuospatial sketchpad [58,59,60,61,62,63,64]. Here, twelve square frames were generated on a computer screen, and five visual stimuli (black circles) were sequentially flashed in five different square frames. The participants were instructed to memorize the sequence and location of the blinking black circle, wait for 15 s, and then reproduce them by pointing on the answer sheet presented on the monitor screen. Square frames and black circles were randomly presented for each condition. The number of correct matches was used as a performance measure; the score range for the task was 0 to 5 points.

The Digit span task [50] was selected to evaluate the auditory–verbal STM performance, as this task has been used to test the phonological loop function in many previous studies [65,66,67,68]. Participants were presented with ten single-digit numbers from 0 to 9 by voice in a random order, were instructed to memorize the ten numbers in the order presented, to wait for 15 s, and then to write the numbers on the provided answer sheet. The number of correct matches was used as a performance measure; the score range for this task was 0 to 10 points.

The 3-back task (the N-back task with N = 3) [51] was employed to evaluate the central executive WM performance, as this task has been widely used to test the central executive function [69,70,71,72]. This task requires encoding the incoming stimuli, monitoring, maintaining and updating the material, and matching the current stimulus to the target stimulus [73,74,75]; decision, selection, inhibition, and interference resolution processes are also involved [71,76]. In each 3-back task trial, the participants received a target number, which was a random number of 0 to 9. Then, a sequence of 16 single-digit numbers from 0 to 9 were presented auditorily; each sequence contained four target numbers at random positions. The participants were then instructed to write down the number that was presented three numbers earlier if the current number matched a predefined target number. As a performance measure, the number of correct matches was employed in this study; the score range for the task was 0 to 4 points. Figure 1 provides an example STM/WM task trial.

2.3.2. Mental Workload Ratings

Subjective mental workload ratings were assigned immediately after each task trial; they included evaluations of each task-specific mental workload experienced while performing the three STM/WM tasks as well as the overall mental workload. For each of the task trials, each participant subjectively rated their mental workload using the Borg CR10 scale [77] based on a visual analogue scale (VAS), as shown in Figure 2.

2.4. Experimental Environment and Procedures

2.4.1. Experimental Environment

The experiment was conducted in the Human Factors and Ergonomics Laboratory of Incheon National University. To ensure optimal cognitive performance, room temperature and humidity were maintained at 22 °C and 40–45%, respectively. A soundproof chamber 2.7 m × 5.2 m (width × length) in size was created in order to exclude the effects of extraneous noise and light. A monitor screen (27 inches) placed in front of the participant was used to provide the experimental tasks. A speaker providing Field noise was placed approximately two meters behind the participant. A digital noise meter (DT-8852) was installed about 30 cm from the participant′s ear to maintain the noise intensity level of the four auditory pre-stimulation $\left({\mathit{P}}_{\mathit{Q}}, {\mathit{P}}_{\mathit{W}}, {\mathit{P}}_{\mathit{F}}, \mathrm{and} {\mathit{P}}_{\mathit{M}}\right)$ and the two task environments $\left({\mathit{E}}_{\mathit{Q}} \mathrm{and} {\mathit{E}}_{\mathit{F}}\right)$. Figure 3 shows the overall experimental environment.

2.4.2. Procedures

Prior to the experimental trials, adequate training sessions were provided to the participants to allow for familiarization with the experimental tasks. Experiment 1 and Experiment 2 were conducted on different days in order to minimize/prevent the effect of mental fatigue on the experiment results. The presentation orders of the two experimental conditions (${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{Q}}$and ${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{F}}$) in Experiment 1 and the four experimental conditions (${\mathit{P}}_{\mathit{Q}} {\mathit{E}}_{\mathit{F}}$, ${\mathit{P}}_{\mathit{W}} {\mathit{E}}_{\mathit{F}}$, ${\mathit{P}}_{\mathit{F}} {\mathit{E}}_{\mathit{F}}$, and ${\mathit{P}}_{\mathit{M}} {\mathit{E}}_{\mathit{F}}$) in Experiment 2 were randomized for each participant. In each experimental condition, each participant performed all three STM/WM tasks; each task trial lasted for an average of approximately two minutes and 30 s, and plenty of rest time was given between the trials. Note that in Experiment 2, the participants were provided with auditory pre-stimulation for five minutes before they performed the three STM/WM tasks. Immediately after each task trial, each participant conducted subjective ratings of the mental workload.

2.5. Statistical Analyses

In analyzing the data collected from Experiment 1, for each of the three STM/WM tasks, a paired t-test was conducted to test the effect of the experimental condition (${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{Q}}$ and ${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{F}}$) on the corresponding STM/WM task score and mental workload rating scores.

In analyzing the data collected from Experiment 2, for each of the three STM/WM tasks, one-way repeated-measures ANOVA was conducted to test the effect of the experimental condition (${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{F}}$, ${\mathit{P}}_{\mathit{Q}} {\mathit{E}}_{\mathit{F}}$, ${\mathit{P}}_{\mathit{W}} {\mathit{E}}_{\mathit{F}}$, ${\mathit{P}}_{\mathit{F}} {\mathit{E}}_{\mathit{F}}$, and ${\mathit{P}}_{\mathit{M}} {\mathit{E}}_{\mathit{F}}$) on the corresponding STM/WM task score and mental workload rating scores. Mauchly’s test was utilized to assess the sphericity of the data for each ANOVA. In cases where the sphericity was violated, the degrees of freedom were corrected. Greenhouse–Geisser correction was used when the Greenhouse–Geisser estimate of sphericity (e) was less than 0.75; otherwise, Huynh–Feldt correction was used [78,79].

In the case of statistically significant ANOVA results, planned comparisons were conducted in which each of the four experimental conditions (${\mathit{P}}_{\mathit{Q}} {\mathit{E}}_{\mathit{F}}$, ${\mathit{P}}_{\mathit{W}} {\mathit{E}}_{\mathit{F}}$, ${\mathit{P}}_{\mathit{F}} {\mathit{E}}_{\mathit{F}}$, and ${\mathit{P}}_{\mathit{M}} {\mathit{E}}_{\mathit{F}}$) in which auditory pre-stimulation was presented was compared with the baseline condition (${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{F}}$) where no auditory pre-stimulation was presented. All statistical tests were conducted using IBM SPSS Statistics 25.0 and were based on an alpha level of 0.05.

3. Results and Discussion

3.1. Experiment 1

3.1.1. Results

The mean scores of the Corsi block-tapping, Digit span, and 3-back tasks were 4.50, 7.96, and 2.21 in the ${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{Q}}$ condition and 4.63, 6.75, and 1.83 in the ${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{F}}$ condition, respectively. Paired t-test results indicated that the experimental condition affected only the Digit span task score, t (23) = 2.87, p < 0.01, r = 0.51; the ${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{F}}$ condition showed a significantly smaller mean than the ${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{Q}}$ condition. For each of the Corsi block-tapping and 3-back task scores, no significant mean difference was found between the experimental conditions. The mean (standard deviation) values of each experimental condition for each dependent variable are shown in Figure 4; asterisks indicate statistical significance in the paired t-test, and error bars represent one standard error above and below the mean.

The mean scores for each task-specific mental workload experienced during the Corsi block-tapping, Digit span, and 3-back tasks and the overall mental workload were 0.54, 1.71, 2.67, and 2.04 in the ${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{Q}}$ condition and 1.13, 3.00, 4.21, and 3.38 in the ${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{F}}$ condition, respectively. Paired t-test results showed that the experimental condition affected all four mental workload rating scores (p < 0.05); the ${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{F}}$ condition showed a significantly larger mean than the ${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{Q}}$ condition in all trials. The mean and standard deviation values of each experimental condition for each dependent variable are shown in Figure 5 with asterisks and error bars.

3.1.2. Discussion

In Experiment 1, each participant performed the three STM/WM tasks for two experimental conditions (${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{Q}}$ and ${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{F}}$) to examine the effects of a noisy environment on the performance of basic STM/WM tasks. Paired t-test results showed that the experimental condition did not affect the scores of the Corsi block-tapping and 3-back tasks and only significantly affected the Digit span task score (p < 0.01). The ${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{F}}$ condition showed a 15.2% smaller mean than the ${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{Q}}$ condition. It can be inferred from these results that ${\mathit{E}}_{\mathit{F}}$ has a greater effect on the phonological loop than on the visuospatial sketchpad and the central executive of the WM system. In other words, noise acted as a type of interference/distraction in the phonological analysis and articulatory rehearsal process during the Digit span task. The observed field-noise effects on the Digit span task performance are in good agreement with the findings of other studies that showed that extraneous auditory stimuli/noise deteriorated task performance. Way et al. [80] revealed that the performance on an auditory processing task requiring the phonological loop function was poorer in a noisy environment compared with a quiet environment. Similarly, Salame and Baddeley [81] showed that the performance on a serial recall task involving phonological STM was lower in a noisy condition of 75 dB than in a quiet condition of 37 dB.

Another notable observation from Experiment 1 was that the mean scores for each task-specific mental workload experienced during the Corsi block-tapping, Digit span, and 3-back tasks and the overall mental workload were significantly higher in the ${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{F}}$ condition than in the ${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{Q}}$ condition (Figure 6). These findings suggest that ${\mathit{E}}_{\mathit{F}}$ increased the participants’ mental workload, congruent with the results of other studies showing that noise increased participants’ mental workloads [82,83]. Given that ${\mathit{E}}_{\mathit{F}}$was an environment in which actual field noise was provided, the results of Experiment 1 suggest that field noise prevalent in actual workplaces can negatively impact the worker’s phonological loop function in addition to increasing the mental workload.

3.2. Experiment 2

3.2.1. Results

For each of the three STM/WM tasks and set of mental workload ratings, the mean (standard deviation) values of each experimental condition are summarized in

Table 2. The mean and standard deviation values of all dependent variables for each experimental condition in Experiment 2.

	Condition	${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{F}}$	${\mathit{P}}_{\mathit{Q}} {\mathit{E}}_{\mathit{F}}$	${\mathit{P}}_{\mathit{W}} {\mathit{E}}_{\mathit{F}}$	${\mathit{P}}_{\mathit{F}} {\mathit{E}}_{\mathit{F}}$	${\mathit{P}}_{\mathit{M}} {\mathit{E}}_{\mathit{F}}$
Measure
STM/WMTask Score
Corsi block-tapping		4.63 (1.05)	4.79 (0.58)	4.42 (1.21)	4.67 (0.76)	4.42 (1.13)
Digit span		6.75 (2.50)	6.50 (2.46)	6.58 (2.87)	7.54 (2.30)	7.96 (1.82)
3-back		1.83 (1.23)	1.88 (1.07)	1.92 (1.05)	1.83 (1.00)	1.96 (0.95)
Mental Workload
Corsi block-tapping		1.12 (0.90)	1.45 (1.61)	1.50 (1.98)	1.70 (1.45)	1.91 (1.79)
Digit span		3.00 (1.47)	3.87 (2.07)	3.91 (2.26)	3.62 (1.63)	3.66 (2.39)
3-back		4.20 (2.02)	5.45 (2.39)	5.12 (2.64)	5.16 (2.40)	5.16 (2.79)
Overall		3.37 (1.49)	4.25 (1.67)	4.04 (2.19)	3.83 (1.71)	3.83 (2.23)

. The results of the statistical analyses showed that the experimental condition significantly affected only the Digit span task score, F(4, 92) = 2.61, p < 0.05, ${\mathsf{\eta }}_{\mathit{p}}^2$ = 0.10, as in Experiment 1; thus, the analysis and discussion of the results for Experiment 2 mainly focus on the Digit span task.

The mean scores for the Digit span task were 6.75, 6.50, 6.58, 7.54, and 7.96 in the ${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{F}}$, ${\mathit{P}}_{\mathit{Q}} {\mathit{E}}_{\mathit{F}}$, ${\mathit{P}}_{\mathit{W}} {\mathit{E}}_{\mathit{F}}$, ${\mathit{P}}_{\mathit{F}} {\mathit{E}}_{\mathit{F}}$, and ${\mathit{P}}_{\mathit{M}} {\mathit{E}}_{\mathit{F} }$ conditions, respectively. The ANOVA results showed that the experimental condition affected the Digit span task score (p < 0.05). Planned comparisons, which compared the baseline condition ${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{F}}$ with the other four experimental conditions (${\mathit{P}}_{\mathit{Q}} {\mathit{E}}_{\mathit{F}}$, ${\mathit{P}}_{\mathit{W}} {\mathit{E}}_{\mathit{F}}$, ${\mathit{P}}_{\mathit{F}} {\mathit{E}}_{\mathit{F}}$, and ${\mathit{P}}_{\mathit{M}} {\mathit{E}}_{\mathit{F} }$), revealed a significant difference only between ${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{F}}$ and ${\mathit{P}}_{\mathit{M}} {\mathit{E}}_{\mathit{F}}$; the ${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{F}}$ condition showed a significantly smaller mean than the ${\mathit{P}}_{\mathit{M}} {\mathit{E}}_{\mathit{F}}$ condition, and no significant mean differences were found in the other comparisons. Figure 6 provides a graphical representation of the mean (standard deviation) of the Digit span task scores for each experimental condition; asterisks indicate statistical significance in the planned comparisons and error bars represent one standard error above and below the mean.

The mean rating scores of mental workload experienced by the participants while performing the Digit span task were 3.00, 3.88, 3.92, 3.63, and 3.67 in the ${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{F}}$, ${\mathit{P}}_{\mathit{Q}} {\mathit{E}}_{\mathit{F}}$, ${\mathit{P}}_{\mathit{W}} {\mathit{E}}_{\mathit{F}}$, ${\mathit{P}}_{\mathit{F}} {\mathit{E}}_{\mathit{F}}$, and ${\mathit{P}}_{\mathit{M}} {\mathit{E}}_{\mathit{F} }$ conditions, respectively. The ANOVA results found no significant mean difference between the experimental conditions. Figure 7 shows the mean (standard deviation) values for the mental workload experienced by the participants while performing the Digit span task.

3.2.2. Discussion

In Experiment 2, each participant performed the three STM/WM tasks for four experimental conditions (${\mathit{P}}_{\mathit{Q}} {\mathit{E}}_{\mathit{F}}$, ${\mathit{P}}_{\mathit{W}} {\mathit{E}}_{\mathit{F}}$, ${\mathit{P}}_{\mathit{F}} {\mathit{E}}_{\mathit{F}}$, and ${\mathit{P}}_{\mathit{M}} {\mathit{E}}_{\mathit{F} }$) to examine the effects of auditory pre-stimulation on the performance of basic STM/WM tasks in ${\mathit{E}}_{\mathit{F}}$. For the Digit span task scores, planned comparisons that compared the baseline condition ${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{F}}$ with the other four experimental conditions revealed a significant difference only between the ${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{F}}$ and ${\mathit{P}}_{\mathit{M}} {\mathit{E}}_{\mathit{F}}$ conditions; the ${\mathit{P}}_{\mathit{M}} {\mathit{E}}_{\mathit{F}}$ condition showed a 17.9% larger mean than the ${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{F}}$ condition, confirming that mixed noise was effective as a type of auditory pre-stimulation (Figure 6). These results are in good agreement with those of Smith et al. [55], who confirmed that the effects of office noise can be reduced by prior exposure to this type of noise.

Interestingly, the mean score on the Digit span task in the experimental condition of ${\mathit{P}}_{\mathit{M}} {\mathit{E}}_{\mathit{F}}$ was identical to that in ${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{Q}}$, which was adopted as the environment for measuring the basic cognitive ability of each participant in Experiment 1 (Figure 4 and Figure 6). Given that ${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{F}}$ showed a 15.2% smaller mean than ${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{Q}}$, the observed results would suggest that if Mixed noise is provided in advance, cognitive task performance degradation due to Field-noise Environment can be restored to a level equivalent to the basic competency of each individual.

Another interesting finding was that while ${\mathit{P}}_{\mathit{M}} {\mathit{E}}_{\mathit{F}}$ showed a significantly larger mean than ${\mathit{P}}_{\mathit{N}} {\mathit{E}}_{\mathit{F}}$ for the Digit span task scores (Figure 6), no significant difference was found in the mental workload score (Figure 7). This is thought to be due to the fact that the task execution time in Experiment 1 and Experiment 2 differed from each other at two minutes and 30 s and seven minutes and 30 s, respectively, depending on whether or not auditory pre-stimulation (five-minute time duration) was presented. In relation to this, some studies have shown that an increased task duration increased the mental workload [84,85]. Given the beneficial effects of auditory pre-stimulation, future studies can devise an accurate measurement method of the mental workload experienced only during the main task.

To further investigate the relative excellence of Mixed noise as a type of auditory pre-stimulation, an additional one-way repeated-measures ANOVA was conducted on an ad hoc basis. This analysis aimed to determine whether the mean scores on the Digit span task were statistically significantly different between the four experimental conditions (${\mathit{P}}_{\mathit{Q}} {\mathit{E}}_{\mathit{F}}$, ${\mathit{P}}_{\mathit{W}} {\mathit{E}}_{\mathit{F}}$, ${\mathit{P}}_{\mathit{F}} {\mathit{E}}_{\mathit{F}}$, and ${\mathit{P}}_{\mathit{M}} {\mathit{E}}_{\mathit{F} }$). The ANOVA results showed that the experimental condition significantly affected the Digit span task score (p < 0.05); planned comparisons that compared the ${\mathit{P}}_{\mathit{M}} {\mathit{E}}_{\mathit{F}}$ condition with the other three experimental conditions revealed that the ${\mathit{P}}_{\mathit{M}} {\mathit{E}}_{\mathit{F}}$ condition showed a statistically significant higher mean than the ${\mathit{P}}_{\mathit{Q}} {\mathit{E}}_{\mathit{F}}$ and ${\mathit{P}}_{\mathit{W}} {\mathit{E}}_{\mathit{F}}$ conditions (p < 0.05). Figure 8 depicts the mean (standard deviation) of the Digit span task score for each experimental condition; asterisks indicate statistical significance in the planned comparisons, and error bars represent one standard error above and below the mean.

The results showing the relative excellence of Mixed noise compared to Quiet noise and White noise are in line with the results of Smith et al. [55], who found that the adverse effects of noise could be reduced by prior noise exposure, suggesting the importance of adaptation. Based on these results, it can be inferred that masking Field noise with White noise was helpful for improving participants’ Digit span task performance. This inference appears to be well supported by previous works demonstrating that masking noises increased cognitive task performance and reduced stress [35,56,57,86].

4. Conclusions

This study empirically elucidated the effects of auditory pre-stimulation on the performance of basic STM/WM tasks in a noisy environment. The major findings were that: (1) the ${\mathit{E}}_{\mathit{F}}$ did not affect the scores of the Corsi block-tapping and 3-back tasks, significantly affecting only the Digit span task score; and (2) the Digit span task performance in the ${\mathit{E}}_{\mathit{F}}$ was improved when mixed noise was provided as a type of auditory pre-stimulation.

The current study results have some theoretical and practical implications. First, redesigning existing work can be suggested so that it consists of tasks mainly involving the visuospatial sketchpad and the central executive rather than tasks requiring the phonological loop in an environment with field noise. Second, work tasks requiring the phonological loop could be adversely affected by noise; hence, it is suggested that these work tasks be performed in a less noisy environment. It is expected that using anti-noise tools such as earplugs or noise-cancelling headphones to reduce/eliminate noise can help prevent the deterioration of cognitive task performance. Lastly, if it is inevitably required to perform a task that mainly involves the phonological loop in a noisy environment, providing the noise mixed/masked with white noise prior to the task could contribute to maintaining/recovering the phonological loop function.

Some limitations of the current study are acknowledged here together and future research ideas are presented. First, in this study, the presentation time of auditory pre-stimulation was set to five minutes, and the effectiveness was confirmed within a task duration of two minutes and 30 s. Future studies can examine the effects of auditory pre-stimulation while varying the presentation time of the pre-stimulation and task duration in order to further enhance our understanding of the pre-stimulation effects on cognitive task performance. Second, this study used analogue white noise provided by a white noise generator. Future work can consider natural white noise sources, such as nature sounds. In addition to white noise, various types of sounds, such as classical music and participants′ preferred music can be included as types of auditory pre-stimulation. In relation to this, as mentioned in the Section 1, several studies demonstrated that listening to classical music prior to a main task increased cognitive task performance [43,44,45]. Third, this study recruited participants in their 20s; future studies can recruit older participants in order to understand the interaction effects of auditory pre-stimulation and age. Fourth, while the current study employed the 3-back task which used a series of auditory stimuli, future studies need to consider different variants of the n-back task, including various input modalities (visual, auditory, etc.), in order to improve our understanding. Lastly, this study considered only the Borg CR10 scale for mental workload ratings; different types of evaluation methods can also be used for a better understanding of auditory pre-stimulation effects on mental workloads in future studies; for example, galvanic skin resistance (GSR), heart rate variability (HRV), electroencephalogram (EEG), and NASA-TLX can be utilized to evaluate mental workloads.