1. Introduction
Noise is pervasive in the modern world. As an important environmental stressor that can continuously undermine human health, it has long been prevalent in factories, mines, airports, military settings, and busy traffic lines. Although great effort has been made to reduce noise in all these places, modern technological development may expose people to some new situations [
1]. For example, modern commutation technology can enable mobile working in noisy transportation facilities. The fast deployment of drones may also put people in the noisy environment, even in their daily routine. As a non-negligible environmental stressor, noise directly harms people’s hearing [
2]. Repeated exposure to a noisy environment can also harm people’s health through the deterioration of the cardiovascular and neuroendocrine systems (for reviews [
3]). Even short-term noise exposure (acute noise) has been found to undermine human cognitive functions including attention, memory, and alertness [
4,
5,
6,
7,
8,
9,
10].
Previous studies found that noise has negative effects on human information processing. For example, it can affect attentional abilities [
11,
12] and can result in poorer performance in working-memory tasks [
4,
6,
10]. Unsurprisingly, noise has also been found to affect learning tasks that require the participation of working memory and attention, such as writing [
8], reading comprehension [
5], etc.
One possible explanation for noise’s negative effect is distraction. Jones et al. [
4] argued that attention could be diverted from the information-processing task (e.g., information retained in working memory) to a noise stimulus unrelated to the task. The effect of distraction also varies with the task characteristics [
13]. In addition, noise may increase mental workload and emotional distress, reducing available cognitive resources [
14]. Therefore, individuals must either put more effort into completing the task or experience a certain decline in performance in the presence of noise. According to the maximum adaptation theory [
15], individuals can adapt to a considerable range of pressures, but task performance is affected once a certain threshold is exceeded. Consequently, noise’s effects increase with intensity and duration.
Furthermore, given noise’s destructive role in cognitive processing, it can be postulated that more complex decision-making is also subject to its influence. However, despite self-reported discomfort in making judgments under noise and the belief that noise may undermine decision-making [
16,
17,
18], hardly any concrete evidence has been found.
For example, Folscher [
17] investigated noise’s effects on medical staff’s decision-making. The results showed that, although there was no statistically significant difference in clinical decision-making performance, 65% of participants thought the noise was distracting and 88% reported experiencing some degree of noise-related stress. Lindvall and Västfjäll [
16] investigated aircraft noise’s impact on performance and decision-making. Half the participants wore noise-canceling headphones, whereas the other half did not. They then played six internal recordings of Boeing cockpit sounds made during flight. After listening to the recording, participants were asked to rate their emotional responses and how they thought they were performing different tasks. The results showed that participants believed they performed better on the task and made better decisions without noise. Syndicus et al. [
18] investigated noise’s effects on risk decision-making among 88 undergraduate students. Four experimental conditions were set up: quiet conditions, office noise (continuous background noise caused by computers, keyboard typing, and occasional ringing phones), a discussion about a museum’s anniversary celebration, and a group discussion about the Ebola virus. The subjects were evenly divided into four groups, and the corresponding sound files were played (participants in quiet conditions were also required to wear headphones). Then, subjects were asked to complete the Lottery Task, the Balloon Analog Risk Task, the Choice Dilemma Questionnaire [
19], and the Risk Scenario Questionnaire [
20]. After completing the decision-making tasks, the subjects were also asked to evaluate their perception of sound, emotion, and confidence in making decisions. The results showed that the three noise conditions did not affect participants ‘decision-making. However, the participant’s confidence in their decisions was affected when noise was present.
A possible reason behind such findings might be the result of the decision-making paradigm used in their experiments. Descriptive paradigms can best capture the essence of these tasks, which comprise most decision-making research. However, when participants make decisions based on descriptions, they only gather information about the potential outcomes of their choices and the associated probabilities by reading complete abstract descriptions of available options and making once-for-all choices [
21,
22]. Moreover, because performing this task does not require the continuous activation and updating of attention and working memory, which are the most vulnerable cognitive components in the face of noise, these paradigms can hardly detect noise’s effects.
However, in reality, most decisions are made not only by descriptions but also rely on long-term accumulated experience (e.g., doctors need to make a diagnosis based on their own experience; pilots steered the plane according to their experience). Human factor researchers are particularly interested in these types of tasks, from air traffic controllers that make aircraft maneuvers [
23] to firefighters making evacuation plans [
24]. Participants must make repeated and continuous choices in a typical experimental paradigm of experience-based decision-making. They may receive no ad hoc information about the possible payoffs of their choices but can only see the results for each turn. Many studies using this paradigm have found that experienced-based decision-making fundamentally differs from description-based decision-making. For example, the framing effect is very salient in a description-based paradigm, which can cause participants to choose the option with fewer expected values (EV) or so-called irrational choices [
21]. On the other hand, in the experience-based paradigm, although participants may choose the options by chance at the beginning, they can implicitly learn the actual payoffs of each option and gradually identify the one with the highest EV through repetition [
25,
26].
In addition, in certain variants of this paradigm, researchers have investigated how experience-based decision-making might be influenced when certain information is presented. For example, Weiss-Cohen et al. [
27] found that conflicts with the actual distribution make an important difference. Participants would choose more high-EV options throughout the decision-making process if there were consistent descriptions. However, if there is a conflicting description, in the beginning, they would choose lower EV options in the direction predicted by the misleading information provided. With the accumulation of correct experience information, the ratio of choosing the higher-EV option would increase. This paradigm and related findings have significant implications for designing supportive decision-making systems for human factor researchers. In many information-scarce situations, human operators have to make decisions based on their own experiences. They may also need to rely on the suggestions provided by the decision support systems. When these tools make mistakes, it is difficult for human operators to establish the correct experience base for judgment.
Therefore, this study attempted to follow the paradigm of experience-based decision-making and investigate how the presence of noise could influence it. We suggest that experience-based decisions are undermined by noise, showing three patterns.
First, in contrast to description-based decisions, experience-based decisions rely heavily on the attentive acquisition of new information and continuous updating of working memory [
28,
29]. Noise can draw participants’ attention toward irrelevant stimuli and undermine memory updating [
4,
12]. We hypothesized that noise would undermine the overall performance of experience-based decision-making (H1).
Second, as evidence accumulation is a necessary part of experience decision-making, it provides a good opportunity to examine the continuous effect of noise. According to the maximum adaptation theory, noise’s duration increases its negative impact. A longer duration results in accumulated damage to an individual’s cognitive system because it takes time to surpass the adaptation threshold. Therefore, we hypothesized that noise’s negative impact on experience-based decision-making would increase as the duration increases (H2).
Third, when provided the correct information, although participants’ attention is distracted, they would not receive strong enough information to make any changes to their initial choices, so their performance would not be reduced. However, when provided incorrect information about the actual payoffs (experience-incongruence suggestion), the participants need to refute the suggestions to make more favorable decisions. Therefore, they must be very attentive to detect discrepancies between the description and their experience and make a change in their default options. However, when noise is present, it may distract the participants from detecting such discrepancies. As a result, we further hypothesized that noise would be more likely to undermine decisional performance when conflicting information was presented, compared to situations in which experience-congruent suggestions were presented (H3). However, this effect can be attenuated when the discrepancies are more salient (H4).
We conducted two studies to test our hypotheses. Study 1 compared noise’s effects on experience-based decision-making across three conditions (no suggestion, experience-congruent suggestion, and experience-incongruent suggestion). To further test the attentional mechanisms, Study 2 examined whether the noise effect under experience-incongruent information could be reduced when such incongruence was salient to be detected.
4. General Discussion
The purpose of the present study was to investigate how noise can influence experience-based decision-making and whether this influence is different when different types of information are provided. In Study 1, participants were offered no information, experienced congruent information, and experienced incongruent information. In Study 2, we further examined possible differences when incongruent information was either salient or not salient. Several findings of this study are worth discussing.
First, we found that noise could undermine the overall performance of experience-based decision-making in both Studies 1 and 2 (H1). This finding contrasts with previous studies using a description-based decision-making paradigm [
16,
17,
18]. As suggested, the reason for this difference might result from the fact that all the information needed to make a description-based decision is fully presented to the participants. However, in experience-based decision-making, which more closely resembles everyday judgment, an option’s real payoffs can only be learned after repeated exposure. During this prolonged process, attentional effort and information updating are heavily involved; because noise has been found to influence these processes, it is reasonable to observe its negative effects.
Second, both studies (H2) found that noise’s negative effect increased as the duration increased. The gradually-worsening effect is consistent with the prediction of maximum adaptation theory [
15]. The theory argues that although noise is a negative stressor, it may not have a prompt effect. Individuals can adapt to a considerable range of pressure; cognitive performance becomes undermined only after a certain threshold is exceeded. In the case of noise, a certain amount of time is required for the negative effect to exceed the adaptive threshold.
In addition, while a decreased performance was observed in the experience-incongruent condition and no suggestion condition, there are no differences in the experience-congruent condition. Therefore, according to “the framework of attentional reduction and maximum adaptation theory of noise”, noise could affect the accumulation of the experience information as it may distract attention and undermine the working memory function. However, this is not a problem when the description is congruent with the experience. This is because even when they are distracted from forming a correct experience evaluation, they can still use the pre-informed description (which can truly reflect their actual experience, the congruent condition) to make a proper judgment. However, when there is no pre-informed description (no suggestion) or the pre-informed description was wrong (the incongruent condition), to make a correct judgment, participants must do a heavy job of memorizing and examining their experience or even make a comparison between what they experienced and pre-informed. Therefore, anything that could undermine the accumulation of experience (such as noise) would affect their decisions.
Negative emotions arising during prolonged noise exposure cannot explain this effect. We did not find any changes in emotions due to the presence of noise and taking emotion into account in the analyses did not alter the major findings of this study.
Finally, we found that the noise effect was attenuated when erroneous information was easily detected (H4, Study 2). This finding deepens our understanding of the boundary conditions of noise in experience-based decision-making. Whereas noise may distract people’s attention by undermining their ability to detect discrepancies between provided information and experienced reality, if the discrepancies are sufficiently salient, people are still able to detect them and make a corresponding change in their behaviors. However, even when this happens, people in the noise condition still perform significantly worse than their counterparts in the non-noise condition (see
Figure 3). This suggests that the combination of noise and erroneous information has a very problematic effect that is difficult to eliminate.
This study is perhaps the first to demonstrate noise’s negative effect on experience-based decision-making. We believe that this influence was caused by its negative influence on attention. The distraction of noise may undermine the information updating process, which is very important for experience-based decision-making but maybe less so for description-based decision-making. This is also one reason why there is no significant negative effect of noise in previous studies of description decision-making [
16,
17,
18]. We also observed that this effect could worsen over time and with hard-to-detect erroneous information, which can also be explained by the framework of attentional reduction and the maximum adaptation theory of noise. Future studies are needed to replicate these findings and expand them to more comprehensive types of noise (e.g., different intensities, frequencies, and durations) and experience-based decision-making tasks. It would also be interesting to test whether such an effect could be attenuated when participants were more experienced and/or had a higher level of motivation.
This study has several important practical implications. First, it again shows the importance of reducing noise in work settings, especially while learning the patterns of a phenomenon and making repeated decisions. For example, it is worth attempting to reduce environmental noise when maintenance workers fail to sharpen their skills effectively. Second, the noise’s accumulation effect suggests that if noise exposure is unavoidable, the exposure time for certain tasks might need to be reduced. Third, providing task performers with information is essential under noisy conditions. However, sometimes, frontline operators are given incorrect instructions or rely on unsuitable plans, maps, blueprints, or handbooks. In such situations, they must be aware that noise might hinder their ability to identify possible problems. Information tools designed to mark and visualize discrepancies between the provided information and experienced reality might be useful in these situations.
This study has several limitations. First, we only examined the effect of a specific type of noise on one experience-based decision-making task among college students. Younger generations are proficient multitaskers (thus supporting higher numbers of distractors like noise). In our study, the noise’s negative effect is still significant. Therefore, the noise’s negative effect might be more significant in other participant groups who are not proficient multitaskers. Second, although we believe that the noise effect may result from undermined attention, another possible explanation for undermined working memory cannot be ruled out.