ARTICLE |
[Download PDF] |
|
| Year : 2012 | Volume
: 14
| Issue : 60 | Page : 260--272 |
Performance, fatigue and stress in open-plan offices: The effects of noise and restoration on hearing impaired and normal hearing individuals
Helena Jahncke, Niklas Halin
Department of Environmental Psychology, Faculty of Engineering and Sustainable Development, University of G�vle, SE-801 76 G�vle, Sweden
Correspondence Address:
Helena Jahncke
Department of Environmental Psychology, Faculty of Engineering and Sustainable Development, University of G�vle, SE-801 76 G�vle
Sweden
Abstract
Hearing impaired and normal hearing individuals were compared in two within-participant office noise conditions (high noise: 60 L Aeq and low noise: 30 L Aeq ). Performance, subjective fatigue, and physiological stress were tested during working on a simulated open-plan office. We also tested two between-participants restoration conditions following the work period with high noise (nature movie or continued office noise). Participants with a hearing impairment (N = 20) were matched with normal hearing participants (N = 18) and undertook one practice session and two counterbalanced experimental sessions. In each experimental session they worked for two hours with basic memory and attention tasks. We also measured physiological stress indicators (cortisol and catecholamines) and self-reports of mood and fatigue. The hearing impaired participants were more affected by high noise than the normal hearing participants, as shown by impaired performance for tasks that involve recall of semantic information. The hearing impaired participants were also more fatigued by high noise exposure than participants with normal hearing, and they tended to have higher stress hormone levels during the high noise compared to the low noise condition. Restoration with a movie increased performance and motivation for the normal hearing participants, while rest with continued noise did not. For the hearing impaired participants, continued noise during rest increased motivation and performance, while the movie did not. In summary, the impact of noise and restorative conditions varied with the hearing characteristics of the participants. The small sample size does however encourage caution when interpreting the results.
How to cite this article:
Jahncke H, Halin N. Performance, fatigue and stress in open-plan offices: The effects of noise and restoration on hearing impaired and normal hearing individuals.Noise Health 2012;14:260-272
|
How to cite this URL:
Jahncke H, Halin N. Performance, fatigue and stress in open-plan offices: The effects of noise and restoration on hearing impaired and normal hearing individuals. Noise Health [serial online] 2012 [cited 2023 Jun 4 ];14:260-272
Available from: https://www.noiseandhealth.org/text.asp?2012/14/60/260/102966 |
Full Text
Introduction
There are concerns about the effects of noise on cognitive performance and health in open-plan offices (see reviews). [1],[2],[3] These concerns are strengthened by experimental studies which indicate that noise (such as irrelevant speech) impairs performance [4],[5],[6] and increases stress. [7] In a previous study from our laboratory we also found effects of realistic office noise on cognitive performance. [8]
There are good reasons to expect that the impacts of noise in an open-plan office vary not only with the types of noise and the layout of the office, but also with the characteristics of the individuals exposed to noise. One important individual difference variable is hearing status. In Sweden more than 10% of the population have some hearing impairment and among those more than half are of working age. [9] No studies to our knowledge have systematically addressed how individuals with a hearing impairment fare (physiologically and psychologically) in open-plan offices. For example, it appears that no research has considered whether hearing impaired individuals' work performance is disrupted by background speech or how stressful they experience a noisy background to be.
If noise has a negative effect on office workers, one important strategy is to find ways to attenuate the negative outcomes. Jahncke et al. [8] showed that a 7 minute break from work in office noise promotes restorative experiences in university students with normal hearing. Restorative effects are typically seen after a break as less fatigue, reduced psychophysiological activity (e.g. blood pressure, stress hormones) and/or increased cognitive/attentional performance.
The aim of the present study was to investigate whether open-plan office noise affects performance and stress, and whether these effects differ between hearing impaired and normal hearing individuals. In addition we investigated the restorative effects of 14 minutes break with two different restorative conditions.
Hearing impairment and performance in Noise
Most studies concerning individuals with hearing impairment have focused on performance and listening effort in tasks involving speech processing. The results show that background noise has a larger effect for the hearing impaired than for normal hearing individuals in these tasks (for a review, see. [10] ) Furthermore, both subjective and objective measures indicate that background noise forces the hearing impaired to exert more effort in speech recognition tasks than is necessary for individuals with normal hearing. [11],[12] To the extent that the task involves speech perception (e.g. taking a phone call), background noise thus is therefore likely to be more detrimental for the hearing impaired than for individuals with normal hearing.
It is not apparent whether this stronger effect from background noise for the hearing impaired also applies to non-auditory tasks. By itself, a hearing loss may be favorable when performing visual tasks in office noise, as the noise is not as prominent or intelligible for the hearing impaired as it is for individuals with normal hearing. On the other hand, most individuals with a serious hearing impairment use a hearing aid, which may amplify the sound to a normal hearing level. However, the hearing aid may also distort the speech signal, [13] which may lead to more distraction for the hearing impaired. Recruitment, which refers to the finding that the same increase in signal strength gives a larger increase in perceived loudness in a non-normal ear than in the normal ear [14] may also play a part here by leading to the perception of sharp onsets of sound that capture attention away from focal task processing much like an auditory deviant. [15] Therefore, for the hearing impaired, as compared to the normal hearing individuals, there may be an additional demand of resisting attentional capture which may cause disruption of or less efficient, focal task processing. In the present study we included noise with temporal variations of speech, phones ringing, and office clatter etc. from working life (i.e. a recording from an actual open-plan office) to test whether the hearing impaired individuals are more distracted in a situation with louder noise than are individuals with normal hearing.
Hearing impairment, fatigue and stress
Babisch [16] reviewed findings concerning stress hormones in relation to research on the cardiovascular effects of noise. Interestingly there are some studies showing specific effects of noise on stress that are related to hearing and noise sensitivity. For example, Melamed and Bruhis [17] performed a field study on industrial workers and have shown that the workers had increased urinary cortisol levels after a day without wearing hearing protection compared with when using them (i.e. a difference of 30-33 dB).
Persson-Waye et al. [18] have also showed that after two hours of cognitive work in low-frequency noise (i.e., ventilation noise), normal hearing individuals that reported themselves as highly sensitive to noise maintain higher cortisol levels than individuals that report less sensitivity to noise. Therefore, higher cortisol levels could be expected in individuals who are highly sensitive to sounds because of a hearing impairment. However, to the authors' knowledge no studies have been reported that demonstrate such a relationship.
Restoration
The study of restoration in different contexts has been guided by different theoretical approaches that entail different outcomes over time. [19],[20],[21] There are lack of studies that investigate low-cost ways to provide well-being and efficiency at work. [22] One way to promote restoration has been to use film-clips or pictures of nature [23],[24] that can easily be used during a break in an office environment. [8] For a review of the psychological benefits of nature experiences see Bowler et al. [25] However, from reported research on restorative environments, it cannot be well specified how long the restorative period should be to improve performance and how varying sound conditions might influence the effectiveness of the restorative period. We attempted to address this in the present study. In a previous study Jahncke et al. [8] tested two cognitive tasks in an attempt to tap fatigue effects: the response inhibition task (SART) and a cognitive inhibition task (Proactive interference), but failed to show any effects of a seven minute rest on task performance. The authors argued that this may be due to ceiling effects. In the present study we attempt to make the response inhibition task more difficult. We also include another task, mental arithmetic, which has been shown to be cognitively fatiguing to perform. [26] Performance on this task may therefore be more sensitive to the benefits of a restorative break. In addition, we double the rest period to 14 minutes.
Hypotheses
Our hypotheses are
Hypothesis 1: Within each of the two hearing groups, there will be an effect of noise on (a) cognitive processing resulting in depreciation in cognitive performance with high noise, (b) stress hormones resulting an increase in hormones associated with stress in the context of high noise, and (c) self ratings of sleepiness and lack of motivation, whereby sleepiness and lack of motivation will increase with work in the high noise compared to the low noise condition.
Hypothesis 2: There will be an interaction between noise and hearing level on the dependent measures. The hearing impaired individuals are expected to have more difficulties (i.e. task performance will be reduced, stress hormone levels will increase and they will be more fatigued) in high noise compared to normal hearing individuals because they are possibly more sensitive and distracted by high sounds. In low noise however, normal hearing individuals may have more difficulties than the hearing impaired because the noise might be more prominent for them at this level.
Hypothesis 3: Participants will restore from fatigue (measured as performance on cognitive tasks, self-ratings and cortisol levels) to differing degrees as a function of restoration conditions. More precisely, we expect a nature movie without sound (positive stimuli) to be more restorative than noise (negative stimuli).
Methods
Design
We used a 2 × 2 mixed factorial experiments, with one between-participants factor (hearing level) and one within-participant factor (noise condition). We also had two different between-participants restorative conditions at the end of the high noise sessions. The participants either watched a nature movie with no sound or listened to office noise without a movie (between-participants manipulations). During the low noise session all participants sat in quiet during the restorative period, to assure the same session length in both noise conditions.
Participants
The participants in the study were 20 hearing impaired (nine females; median age = 53) and 18 normal hearing individuals (eight females; median age = 48). The participants with hearing impairment were recruited from the Swedish association of the Hard of Hearing Persons (HRF) and we only included those in the age span 20-65 years old. Those who reported having severe tinnitus and/or Menières disease were excluded. In the next step we recruited matched normal hearing participants following the criteria of the same gender, age (accepted with small age variation), occupation and education as the hearing impaired individuals. Out of the 38 participants, 20 had earlier worked with office tasks, and 10 had some experience with open-plan offices. Participants' hearing abilities were screened [see [Figure 1]a and b for the means of the audiograms for the hearing impaired and normal hearing group]. The standard deviation across all values was 24.83 for the hearing impaired and 5.28 for the normal hearing participants. The criterion to be included in the hearing impaired group was a mean hearing loss of 28 dB or more over the frequencies 500, 1000, 2000 and 4000 Hz, which did not give any overlap between the two groups. The participants with hearing-aids were told to have them turned on during the experiments. Participants were informed of the nature of the study before participating and they were randomly assigned to the two sequence orders of the two noise conditions. Participation in all sessions of the experiment was compensated with 990 Swedish crowns.{Figure 1}
Research setting
The research was carried out in an office laboratory (63 m 2 ) at the University of Gävle. The room was designed to simulate a neutral open-plan office including windows to a white room with simulated outdoor lighting. Ambient conditions in the indoor environment were kept approximately constant throughout the experiments. The room temperature at the work place varies in-between 20.7-21.6°C during the experimental sessions. The airflow was 20 l/s for each person in the experiment, and the CO 2 concentration was always well below 1000 ppm. Luminance was set to 480-520 lx incident on the height of the seating in each workstation and met the current recommendations in Sweden for work with computers. [27] The luminance in the room outside the window was set to 1950 lx measured at the middle of the window.
The workstations were separated by 1.43 m high screens. Each cubicle was 1.24 m wide. At each end of the set of cubicles a pane of glass was installed to better avoid direct sounds from the loudspeakers to the participants sitting close by. Behind the participants on each side were also 3.60 m wide × 1.80 m high screens to better avoid direct sounds from the loudspeakers behind.
Noise conditions
We recorded the noise used in the experiment in an actual open-plan office in Sweden. From the multi-channel recording, one hour of office noise was extracted, edited and reproduced in the test room with eight loudspeakers, two at each wall, and one subwoofer. In the high noise condition, additional phone and sound signals were added to the office noise. The phone signals consisted of different mobile tunes and stationary telephones ringing. The speech signals were conversations recorded from radio, from which we cut away one of the voices to simulate telephone conversations. The high noise was reproduced with an equivalent A-weighted sound level of 60 L Aeq in the room. In the low noise condition the added phone and sound signals were excluded and the noise was low pass filtered, which reduced L Aeq in the room by 12 dBA. The sound level was further attenuated to reach 30 L Aeq . The one hour office noise tape was looped once, and the sound signals in the latter hour were rotated 180 degrees to equalize noise exposure throughout the room.
The sound level measurements showed the same L Aeq in the first and second hour of the reproduced office noise. The statistical distribution of sound pressure levels was similar in the high and low noise conditions except that the equivalent levels differed by 30 dB. The Speech Transmission Index (STI) was the same for the high noise condition (60 dB) in this study, as for the high noise condition (51 dB) in the earlier study by Jahncke et al. [8] The STI values in the low noise condition (30 dBA) were lower than in the low noise condition (39 dB) by Jahncke and colleagues. The speech noise in the low noise condition was now so close to the hearing threshold that it was not possible to hear the weakest speech sounds.
Restoration conditions
After two hours of work in office noise the participants went through a restoration period for 14 minutes. Two different restorative conditions were tested after the high noise condition: half of the participants saw a movie with clips of nature environments in quiet and half of the participants were exposed to continued office noise. In the low noise condition, all participants sat in quiet during the restorative period. In the analysis of restoration effects only the movie and office noise conditions were compared, as the quiet condition differed from these to conditions also with respect to the preceding noise condition.
Measures
Cognitive performance
We used several different cognitive tasks to measure the effects of the noise and restoration manipulations. A response inhibition- and an arithmetic test (further described below) were included to measure fatigue/restoration and were performed three times: in the beginning (pretest), after 2 hours of work (Post work), and after the restoration period at the end of the experimental sessions (Post rest). The other cognitive tasks were included twice to measure change across time in noise, except for the reading comprehension task that was included once per session because of the time it took to perform [see [Figure 2] for an overview of the task order]. The same presentation order of dependent measures was used for all participants to control for the timing between task and noise. As we were not trying to make an evaluation of each specific test, we were ready to take the confounding between when and which task was performed in return for a reduced error term.{Figure 2}
Response inhibition
Response inhibition was measured with the Sustained Attention to Response Test (SART), which taps the ability to inhibit a physical response and/or sustain attention. [28] Digits from one to nine were presented repetitively and the participants were told to respond with a key press to all numbers except three target numbers (i.e., 2, 5 and 9). A digit was presented on the computer screen once every 1000 ms and remained for 1000 ms. Each trial consisted of 140 digits where 15% were targets (i.e., number 2, 5, 9) to which the response was to be inhibited. The following scores were considered: (1) the sum of errors of commission (i.e., the number of times the response to the targets was not successfully inhibited); (2) the sum of errors of omission (i.e., the number of times a response was not made to a number which should have been responded to), and (3) reaction time (i.e., the mean response time in milliseconds for all numbers except the target numbers).
Arithmetic's
This task was a computational task with ten single-digit numbers, which were presented one by one on the screen. The operation to perform (+ or -) was presented between the numbers. The participants controlled the presentation speed. When all the ten numbers were presented the participants typed in the answer. [29] In total there were 16 sets and the scores were the sum of correct answers and the time it took to complete the arithmetic task.
Math
This was a simple addition task whereby participants were required to add double-digit (e.g. easy task; 44+38) and triple-digit numbers (e.g. difficult task; 122+435). Each experimental block consisted of 15 double-digit and 15 triple-digit expressions. The answer time was set at 20 seconds for each math expression. The task always started with a difficult expression that was followed by an easy expression and so on. The scores were the sum of correct answers (divided in the easy task and the difficult task) and the mean response time for the expressions (also divided in the easy task and the difficult task).
Word memory
Memory for words was measured with a Proactive interference (PI) task, which taps the ability to inhibit information that once was relevant for recall but has since become irrelevant for the task. [30] The participants learned words from eight lists with ten words each. The separate lists were composed of words drawn from the same category. After a given list of ten words had been presented, participants were required to free recall the words from the most recent list and ignore (suppress) words from earlier lists. We used the following categories: "fish and other water living creatures", "fruits and vegetables", "animals", "cloths and accessories", "kitchen- and gardening tools", "countries" and "cities in Sweden". The words were presented on the computer screen every 1000 ms and remained for 1000 ms. The orders of the categories within each experimental session were counterbalanced between participants. The following four scores were considered: (1) PI points: the sum of words recalled from the correct list; (2) PI pro: the sum of error words (i.e. from earlier lists); (3) PI lacking answers: the sum of words not remembered; and (4) PI other errors: the sum of error words (i.e. new words not coming from earlier lists).
Reading comprehension
The texts for the reading task were taken from the Swedish National University Aptitude Test. One sentence at a time was presented and the participant decided when the next sentence should be presented by clicking the space button. In 25 of the 43 sentences one word was missing and the participants were given a choice between four alternatives. Sometimes the information for the correct choice was in the text that no longer was on the screen, sometimes the information appeared on the screen simultaneously. The scores were the sum of correct answers and the time it took to complete the reading task.
When all the text was read, two variations of memory questions were posed. In the first set the participant was presented with a sentence and was asked to judge whether the sentence was taken from the text read or not. In some of these sentences an important word from the original text had been changed and in other sentences there was no similarity with the original text. Scores were the sum of correctly marked sentences. In the second set there were questions about the information stated explicitly in the text, (e.g. years, book titles, places, names). Scores were the sum of correctly marked details.
Search task
In this task participants had to locate a specific object among different information channels. The information was organized in a table. The seven columns contained information about price, location, area, year etc. The twenty rows kept together the information about a given object (e.g. a person, a house or a country). The participants were asked to find the object that met a set of criteria, either by using two columns (easy) or four columns (difficult). Therefore this task taps processes required to search through and understand the contents of a table with information, while successively updating and memorizing which information is most correct according to the target criterion. Each experimental block consisted of the twelve questions (e.g. six easy questions and six difficult questions) and time was limited to one minute per question before a new question was presented. The scores were the sum of correct answers (divided into the easy task and the difficult task) and the time it took to complete the search task (also divided into the easy task and the difficult task).
Serial recall
In this task with numbers, the participants were told to remember a string of eight one digit numbers (i.e. 1-8) and then recall them in the correct order. Each number was presented for 400 milliseconds followed by a pause of 350 milliseconds before the next number was presented. After the eight digits were presented an answer box appeared on the computer screen with the numbers 1-8 in random order. The participants then had to organize the numbers in the correct order. Each block consisted of 27 lists and only numbers written at their correct positions were scored correct. We also scored lists that were complete (e.g. when all the eight digits were assigned their correct positions).
Self ratings of fatigue
Perceived sleepiness and motivation were measured with the Swedish Occupational Fatigue Inventory (SOFI) developed by Åhsberg, Gamberale and Kjellberg. [31] The questions are based on a factor analysis where five components were established for 25 items describing feelings of cognitive- and physical fatigue. Only the three components describing cognitive fatigue were selected for this study. Lack of energy was measured with the words "worn out" and "exhausted", Lack of motivation with the words "passive" and "uninterested", and Sleepiness with the words "sleepy" and "the number of yawns for the last ten minutes" on a scale ranged from 1 = not at all to 4 = a lot.
Stress hormones
Urinary catecholamine's
The urine samples were collected directly when all participants arrived to the laboratory (Pre work), in the middle of the work period (Mid work; after 1 hour 30 minutes) and at the end of the experimental session (Post rest; after 3 hours). For each participant, the exact time after the bladder had been emptied by voluntary voiding was noted, the total urine volume was measured, and 40 ml of urine was mixed in 100 ml glass bottles containing 0.5 ml 2 M HCL as a preservative. Samples were stored in a freezer at - 20°C until subsequent preparation. The samples were thawed within 2 months and purified according to the BIO-RAD method (Urinary Catecholamine's by HPLC, Reagent Kit, Catalogue Number 195-5841/N) and assayed following standard methods for high performance liquid chromatography. The amounts of the catecholamines Norepinephrine and Epinephrine were expressed as ng/h/kg body weight. The analyses were carried out by a professional blind to the experimental conditions at the University of Gävle.
Salivary cortisol
Saliva samples were obtained with Salivette tubes (Sarstedt, Landskrona, Sweden) at four times during the experimental session: Pre work, after approximately 1 hour of work, Post work and Post rest [[Figure 2] for an overview of the experimental session]. It is however, unclear how to handle the timing of the measurements of the peak response in cortisol as there might be a time lag before the reaction is measurable. [32] The radio immunoassays (RIA) of the samples were carried out in the Stress Research Institute at Stockholm University by a professional blind to the experimental conditions.
Procedure
Data collection took place at a simulated office at the University of Gävle. Participants went through a practice session for one hour, several days in advance of the first experimental session. The practice session was meant to reduce possible training effects on the cognitive tasks. The experimental sessions only took place on Tuesdays and Wednesdays. The experimental sessions were run between four and seven pm and for each participant, one day separated the two sessions. The procedure for one experimental session is presented in [Figure 2].
After the first urine and saliva samples, the participants completed the SOFI and performed the first block of fatigue tasks in quiet. After the fatigue tasks they preceded through the first block of cognitive tasks while one of the two different noise conditions (low noise, or high noise) was played back. The session order of the noise conditions was counterbalanced but not the cognitive tasks between the experimental sessions. The Search task and Word memory task were counterbalanced between participants within one session to reduce the risk of an order effect because of the different task categories. After 1.5 hours of work the participants left the simulated office for the second urine sample. Back in the laboratory the second block of cognitive tasks started. Next, the second fatigue block started and the noise was turned off. In the beginning of this block a saliva sample was collected and the SOFI was again completed.
The next phase was the restoration period wherein the participants were instructed to put their headphones on. In the high noise condition the participants either watched a nature movie with no sound or listened to office noise (without any film) during the restoration period. The participants were randomly assigned to one of the two restoration conditions. In the low noise condition all participants sat in quiet for 14 minutes to keep the same session length. After the restoration period the participants put off their headphones and went through the last fatigue block in silence. The whole procedure took about three hours to complete and one to eight participants were tested at each occasion.
Statistical analyses
Because of failures to follow instructions 3 participants were excluded from the analyses of the noise effects, leaving 34 participants for further analysis. One participant was excluded from the analyses of the restoration effects due to equipment malfunctioning during the restoration period. There are further data missing for the following number of participants in these measures: SART (1), Reading comprehension (2), Search task (9), SOFI (4), Norepinephrine (3), and Epinephrine (3) due to lack of answers and equipment malfunctioning. A General Linear Model (SPSS version 18) with repeated measures (mixed) design was used for the analysis of noise and fatigue effects. When Mauchly's test indicated significant non-sphericity in the variance-covariance matrix for the within-subject analyses, the Greenhouse-Geisser adjusted degrees of freedom and corresponding P-values are reported for the F-tests.
Results
Results are reported first for effects during work, and then for the effects of restoration. For both sections the presentation order begins with the effects on cognitive performance, then on psychophysiological stress and then finally on self-ratings of fatigue.
Noise effects on cognitive performance during work
In the design, noise conditions (high and low) was a within-participant variable. However, given evidence that the performance measures showed strong transfer effects (i.e. interactions of performance and session order), we defaulted to treat noise conditions as a between-persons variable with the high and low noise levels in the first experimental session as a between-participants variable.
Math
There were no significant effects of noise or hearing on the number of correct answers in the math tasks. This might be due to a ceiling effect since few participants had more than three errors. Further, the analysis showed no significant main effect of noise on speed. Therefore, there is no support for Hypothesis 1 (a) - that mathematical performance was impaired by noise level.
However, there was an interaction between noise and hearing on performance speed. As shown in [Figure 3]a both the hearing impaired- and normal hearing groups appeared to work faster in the easy tasks in high noise than in low noise. In the difficult tasks [Figure 3]b the hearing impaired appeared to work faster in high noise than in low noise, whereas the normal hearing groups performed at about the same speed in both high and low noise. An ANOVA with noise condition and hearing (2 × 2) as between-participants factors and time and difficulty (2 × 2) as within-participant factors showed the Noise × Hearing × Difficulty interaction to be significant (F (1, 30) = 4.62, P < 0.05, partial ŋ 2 = 0.13). The interaction contradicts Hypothesis 2 because the hearing impaired worked faster in high noise than the normal hearing individuals. Also for the easy tasks the results were not in line with Hypothesis 2, as both hearing groups were faster in high noise.{Figure 3}
Word memory
The analysis showed no significant main effect of Noise on the measures of word memory. This contradicts Hypothesis 1 (a) which predicts impaired performance during the higher noise level. Further, as shown in [Figure 4], the hearing impaired participants appeared to do worse (had more omissions) in the word memory task (PI) during high noise, than in the low noise condition, whereas the opposite was true for the group with normal hearing. However, the ANOVA only showed a trend towards Hearing × Noise interaction F (1, 30) = 3.30, P = 0.079, partial ŋ2 = 0.10. The results thus showed a trend in line with Hypothesis 2, indicating that the hearing impaired had more difficulties in high noise compared to normal hearing individuals.{Figure 4}
Reading comprehension
In contradiction to Hypothesis 1 (a) there were no main effects of noise on any of the reading scores. Again, only when hearing status was considered, an effect of noise level was seen.
As shown in [Figure 5] the participants with a hearing impairment appeared to remember more sentences in low noise compared to those in high noise, while the normal hearing group showed the opposite effect. This was confirmed by the ANOVA, which showed a significant interaction between Hearing and Noise on memory for the contents (points for correctly marked sentences), F (1, 28) = 4.70, P < 0.05, partial ŋ 2 = 0.14. Further, memory for specific details in the text (points of correctly marked details) showed a similar pattern, but in this case there was only a tendency toward a Hearing × Noise interaction, F (1, 28) = 3.28, P = 0.081, partial ŋ 2 = 0.11. These results were in line with Hypothesis 2 showing that the hearing impaired had more difficulties in high noise than the normal hearing individuals.{Figure 5}
Search task and serial recall
There were no statistically significant effects of Noise on the Search task and for Serial recall performance. The participants had only a few correct answers in the difficult part of the Search task which suggests a floor effect.
Noise effects on acute psychophysiological stress during work
Since there are large individual variations in stress hormone levels, the analyses of the physiological measures were only conducted with noise as a within-participant comparison. The mean cortisol level was gradually lowered between the three measurement occasions both in the low and the high noise conditions, F (1.31, 39.27) = 33.82, P < 0.001, partial ŋ2 = 0.53, [Figure 6]a and b. There was no main effect of noise, but in low noise the hearing impaired and normal hearing participants had approximately the same cortisol levels at all measurement occasions, whereas in high noise the hearing impaired group had higher levels than the group with normal hearing. This was reflected in the ANOVA as a tendency towards a Time × Noise × Hearing interaction, F (1.37, 41.23) = 3.06, P = 0.075, partial ŋ 2 = 0.093. In summary, we found no significant support for Hypothesis 1 (b) but a tendency towards higher stress levels during high noise exposure for the hearing impaired participants in line with Hypothesis 2.{Figure 6}
Norepinephrine levels declined over time, F (1.28, 33.19) = 10.18, P < 0.002, partial ŋ 2 = 0.28. However, there was no significant main effect of noise on the level of catecholamines over time and no interaction between noise and hearing. No support thus was found for Hypotheses 1 (b).
Noise effects on self ratings of fatigue during work
In this section the focus is only on the noise effects which were measured Pre work and Post work. In the section concerning restoration we also include measures from Post rest (after the period of restoration).
In the first step we analyzed the SOFI ratings with session order and hearing (2 × 2) as between-participant factors and noise and time (2 × 2) as within-subjects factors. The analysis revealed a significant main effect of time on all the scales, sleepy, F (1, 29) = 43.67, P < 0.001, partial ŋ 2 = 0.60; the amount of yawning, F (1, 29) = 11.48, P < 0.01, partial ŋ 2 = 0.28; worn out, F (1, 29) = 52.83, P < 0.001, partial ŋ 2 = 0.65; exhausted, F (1, 29) = 42.38, P < 0.001, partial ŋ 2 = 0.59; passive, F (1, 29) = 54.72, P < 0.001, partial ŋ 2 = 0.65; and uninterested, F (1, 29) = 62.26, P < 0.001, partial ŋ 2 = 0.68; indicating an increase of sleepiness, lack of energy and loss of motivation over time [Table 1]. However, the participants were more sleepy, tired and were less motivated over time during both high and low noise exposure, which contradicts Hypothesis 1 a of a main effect of noise level.{Table 1}
Further, the hearing impaired participants and the normal hearing participants rated themselves at about the same level of sleepiness during the low noise exposure. During the high noise exposure the hearing impaired were more sleepy than the normal hearing group [Figure 7], yielding a significant Noise × Hearing interaction on the rating of the amount of yawning, F (1, 29) = 5.56, P < 0.05, partial ŋ 2 = 0.16.{Figure 7}
Unexpectedly, participants became more uninterested (unmotivated) over time in low noise than when they were in high noise, which resulted in a significant Time × Noise interaction on the ratings of being uninterested, F (1, 29) = 4.43, P < 0.05, partial ŋ 2 = 0.13. Note also that the mean values indicated that the normal hearing participants were less motivated than the hearing impaired during both noise conditions [Table 2]; however, this Time × Noise × Hearing interaction was not significant.{Table 2}
In summary, the hearing impaired participants were more tired during high noise exposure than the normal hearing participants, which is in line with Hypothesis 2.
Effects of restoration on cognitive performance
As a first step we analyzed if the participants had a drop in arithmetic performance from Pre work to Post work during high noise to establish that the participants were in a state of need for cognitive restoration. Although discrepancies were found in performance between session orders, the overall pattern showed decreased quality and increased speed during work in high noise exposure. The ANOVA showed that the Time × Session order interaction was significant for both correct answers; F (1, 29) = 5.05, P < 0.05, partial ŋ 2= 0.15, and performance speed; F (1, 29) = 9.23, P < 0.01, partial ŋ 2= 0.24.
As a second step we analyzed whether there was a difference in performance between a restorative period with movie or noise from Post work to Post rest. As shown in [Figure 8]a the normal hearing participants performed better (correct answers) after restoration with a movie, while their performance declined after continued noise exposure. As shown in [Figure 8]b, the opposite pattern emerged for the hearing impaired participants with decreased performance with movie and improved performance with noise. An ANOVA with Restorative condition, Hearing and Session order (2 × 2 × 2) as between-participants factors and Time (2) as a within-participant factor, showed that the Restorative condition × Time × Hearing interaction was significant; F (1, 25) = 7.64, P < 0.02, partial ŋ 2= 0.23. In summary, only the normal hearing participants followed Hypothesis 3, with increased performance after restoration with the movie and decreased performance with noise.{Figure 8}
Effects of restoration on acute psychophysiological stress
No effect of restoration was found on the urinary catecholamines and salivary cortisol, yielding no support for Hypotheses 3. However, see the effects of time on the urinary catecholamines and salivary cortisol under the section "Noise effects on acute psychophysiological stress" above.
Effects of restoration on self ratings of fatigue
As stated previously the participants were sleepy prior to the restorative period, as the participants showed a significant increase in sleepiness, a significant decrease in energy, and a significant loss of motivation from Pre work to Post work during both high and low noise exposure.
As shown in [Figure 9]a, the normal hearing participants had a loss in motivation (i.e. passiveness) after a restorative period in noise and an increase in motivation after they had watched the movie. However, the hearing impaired showed the opposite pattern with an increase in motivation after a restorative period in noise and a decrease with movie [Figure 9]b. An ANOVA with Restorative condition, Hearing and Session order (2 × 2 × 2) as between-participants factors and time (2) as a within-participant factor showed that the Time × Hearing × Restorative condition interaction was significant for lack of motivation (passiveness); F (1, 24) = 5.50, P < 0.05, partial ŋ 2 = 0.19.{Figure 9}
For the other SOFI-variables there was a session-order interaction, which made us leave those analyses from further consideration.
In summary, only the normal hearing participants followed Hypothesis 3, with increased motivation after restoration with the movie and loss in motivation with noise.
Discussion
Effects of noise on cognitive performance
There were no main effects of noise level on any of the tasks (i.e. Math, Memory for words, Reading, Search task and Serial recall), contradicting Hypothesis 1 a. The high noise level tested was 60 L Aeq , which is in the region of a normal speech level. However, when exploring the effects of hearing status, we found an interaction between hearing and noise level in line with Hypothesis 2 for two of the tasks. The hearing impaired participants were more affected by high noise than the normal hearing, as shown by decreased performance in memory for the contents of a text, and a tendency for decreased performance in the memory for words test. Possibly, noise at 60 L Aeq was distracting for the hearing impaired as they presumably are more sensitive for noise.
An interesting result was also that the normal hearing participants worked better overall during high noise than low noise. One plausible explanation may be their motivational and arousal level. For example, Yerkes and Dodson's [33] theory predicts a need of a higher level of arousal for optimal performance on easy tasks than more difficult tasks (i.e. the "task difficulty hypothesis"). If most of the tasks we employed were perceived as quite easy to the participants they most probably needed arousing stimuli (i.e. noise) to perform well. In line with this, Suedfeld and Loewen [34] have showed that feelings of arousal can increase with noise. The subjective ratings also showed that the participants had significantly higher motivation in high noise than low noise, although sleepiness increased over time. However, there are some difficulties with these theories, particularly in specifying the optimum level of arousal.
One notable result was that the easy math tasks benefited from a higher noise level. In this task (i.e. adding two digit numbers) both hearing impaired and normal hearing participants worked faster in high noise than in low noise. In the more difficult math task (i.e. adding three digit numbers) the hearing impaired participants worked faster in high noise, while the normal hearing participants worked at about the same speed in high and low noise. This benefit is in line with a previous study by Loed, Holding and Baker [35] which showed that background noise can increase performance of simple arithmetic's (i.e. speed in work completion) when performed during the afternoon. The present experiment was also conducted during the afternoon, therefore, the results appear to be consistent with an arousal explanation.
Another question that arises is why there was no interaction between noise level and hearing status for two of the tasks (i.e. Serial recall and Search task). The lack of an interaction with noise level may be explained by earlier studies showing that some tasks are not vulnerable to a variation in noise intensity; rather performance is impaired by the changing properties of the sound. [36] However, the tasks shown to be impaired by noise intensity when hearing status was considered (i.e. memory for contents of a text, memory for words), need more complex semantic processing and involvement of long term memory (e.g. processing of word meaning, rehearsal of solutions). This is in line with an earlier study [8] that also showed word memory (i.e. semantic memory) to be impaired by office noise level but not, for instance, serial recall. The present results therefore support the finding that only some tasks are impaired by office noise. [37] One important objective for future studies will be to decompose tasks into their component processes and then test the impact of different noise sources on those constituent processes.
Effects of noise on acute stress
The participants showed a significant decline in cortisol in both noise conditions over time, consistent with the decline that occurs as a part of normal circadian variation. However, there was a tendency for the hearing impaired participants to show higher stress levels over time in high noise compared to low noise. The same pattern of decline over the experimental session in both noise conditions was also evident in the norepinephrine, measured from urine. In summary, we found no main effect of noise on stress hormones, contradicting Hypothesis 1b. The results merely showed a weak tendency in line with Hypothesis 2 demonstrating that the hearing impaired participants were more stressed by the high noise condition compared to the normal hearing participants.
Effects of noise on self-ratings of fatigue
In both noise conditions there was an increase of sleepiness and loss of motivation over time. However, there was no main effect of noise level on fatigue, contradicting Hypothesis 1b. Instead there was an interaction between noise and hearing status in line with Hypothesis 2. During low noise both hearing groups were at about the same level of sleepiness. However, during high noise the hearing impaired were expected to be more tired than the normal hearing participants.
Restoration
The participants felt sleepier, less motivated and less energetic from Pre work to Post work, and their performance quality decreased in the arithmetic task during the work period. Thus, the results meet the criteria for analyzing cognitive restoration (i.e. an antecedent condition indicating fatigue).
It was shown that the normal hearing participants΄ performance and motivation improved with the nature movie, and declined after continued noise exposure. This supports earlier findings [8] that normal hearing participants restore better with a nature movie than with noise. However, for the hearing impaired participants of the present study, the opposite pattern emerged. Both their performance and motivation decreased after restoration with the nature movie, and improved after noise exposure. This is difficult to explain, but one possible explanation is that continued noise exposure might have counteracted their possibilities to slow down, relax, and feel how sleepy they were.
Further, no restorative effects were found on the salivary cortisol and urinary catecholamines. The stress hormones showed a decline over time in accordance with the normal circadian rhythm.
In summary, Hypothesis 3 only held for the normal hearing participants' performance and motivation. They showed significantly increased performance and motivation after restoration with the movie and decreased performance and motivation with noise. The hearing impaired contradicted Hypothesis 3 with the opposite pattern.
A comparison with our previous experiment
Jahncke et al, [8] reported an experiment with open-plan office noise, cognitive performance, restoration and subjective ratings that in many respects is comparable to the experiment reported here. However, there are some discrepancies in the results, which may or may not be due to differences in the sounds employed, in the cognitive tests employed and/or the characteristics of the participants. For instance the participants in the previous study were young (mean age = 26 years) students at the University of Gävle, while the participants in the study reported here were older (mean age close to 50 years) and had less formal schooling and had different occupations. To probe whether this difference had any impact on the difference in results between the two experiments, a follow up statistical comparison was made between the results on a comparable test (i.e. serial recall) from participants in Jahncke et al,[8] and participants from the present study. This analysis showed that normal hearing participants in the study of Jahncke et al,[8] that were students, performed better in low noise than the normal hearing participants in the present study, t (53) = 2.27, P < 0.05, although the low noise in Jahncke et al, [8] was 39 L Aeq as compared to 30 L Aeq in the present study. It should be noted that serial recall is rather insensitive to sound level. [36]
The baseline difference in performance level between the normal hearing groups, in the two experiments, may be one of the explanations as to why the results of the present study differ from the results of the earlier study by Jahncke and colleagues. It is however, difficult to further explain the discrepancies in our findings.
Limitations
It should be acknowledged that this study was small in scale and that we had additional problems with missing data. There is therefore a risk that the conclusions are underpowered. The lack of interactions especially may have been seriously influenced by the small N in our samples. It is possible that different results may have been found in a larger sample, especially when considering how much variation there may be in the hearing impaired group.
Further, we cannot tell whether we would have observed an increase in fatigue if we had not conducted our study from 4-7 pm. This for most office work would be towards the end of the working day and another choice of time frame might have changed the results. However, we choose to conduct these studies during the afternoon as this is more appropriate for measuring the stress hormones, as we expect a decrease in stress hormone levels in the afternoon. If the stress hormones are measured when a daily increase is expected we could not rule out whether the increase is due to noise or the circadian rhythm.
According to the stress hormones, most of the tasks employed were most probably perceived as quite easy to the participants. This could explain why we did not see any main effect of noise on the stress hormones in our study.
It can also be questioned if it is of real-life relevance to watch a nature movie with no sound as a tool to be used in real office life. Outside the experimental studies it is probably of better practical use to take a walk in a real nature environment during the break. Further, we are not able to say whether it is the silence, the nature, or both that are important as our stimuli contained both these features. This is something for future research to examine further.
Conclusion and Practical Implications
The results of this study showed that the hearing impaired were more distracted by high noise than the normal hearing, as indicated by decreased performance in the tasks requiring recall of semantic information. However, they were not distracted when they performed the two tasks (Serial recall and Search task) that are underpinned by processes, which previously have shown to be insensitive to noise level.
The hearing impaired individuals were also more fatigued by a higher noise exposure than people with normal hearing and they tended to have higher stress hormone levels during high noise compared to the low noise. The present study however, cannot disentangle whether the effects reported are a result of hearing impairment or a result of the use of hearing aids. The participants might also have used different types of hearing aids, which can differ in performance and suitability for our test conditions. [38]
An important finding in this study is that it is not just the types of noise and layout of the office that affect employees. The impact also varies with the hearing characteristics of the persons exposed to noise and the tasks they perform. Therefore, special consideration needs to be made to the individual prerequisites. For example, it is rarely possible for hearing impaired persons to decrease the negative impact from surrounding noise by turning off the hearing aids, when the work requires a readiness for phone calls and communication with colleagues. Therefore other solutions might be needed (e.g. quiet rooms) to be able to enhance performance and wellbeing for these individuals.
There should also be further concerns for designing sound environments that can promote daily restoration at workplaces, as indicated by an earlier study of Jahncke et al. [8] The present study showed that a break with a nature movie improved the normal hearing participants΄ performance in arithmetic's, and that continued noise exposure during the break decreased their performance. These results underscore the need to address the issue of how to enhance restorative qualities within the noisy environments of our daily surroundings.
Acknowledgments
This research was funded by AFA Insurance and by the University of Gävle and was conducted in partial fulfillment of the requirements for a doctoral degree. We would like to thank Staffan Hygge and Terry Hartig for helpful advice in planning and conducting this study. We are also grateful to Kenth Dimberg at the University of Gävle for analyzing the urine samples and Lars Holmberg at the Stress Research Institute, Stockholm University for analyzing the saliva samples. We also thank Örjan Johansson and Johan Odelius at Luleå University of Technology for their help with the recordings and arrangements of the office noise used in this experiment and to Valtteri Hongisto and David Oliva for their help with analyzing the STI of speech in the office noise recordings. We also thank Anders Kjellberg, John Marsh, Patrik Sörqvist and the anonymous reviewers for helpful comments on this manuscript, the Swedish Association of Hard of Hearing People (HRF) for generous support, and Ecophon and S-line office for their donations to furnish the simulated open- plan office.
References
| 1 | Navai M, Veitch JA. Acoustic satisfaction in open-plan offices: Review and recommendations (IRC-RR-151). Canada: National research council Canada, Institute for research in construction; 2003. |
| 2 | Oommen VG, Knowles MK, Zhao IZ. Should health service managers embrace open plan work environments? A review. Asia Pac J Health Manag 2008;3:37-43. |
| 3 | Rashid M, Zimring C. A review of the empirical literature on the relationships between indoor environment and stress in health care and office settings: Problems and prospects of sharing evidence. Environ Behav 2008;40:151-90. |
| 4 | Haka M, Haapakangas A, Keränen J, Hakala J, Keskinen E, Hongisto V. Performance effects and subjective disturbance of speech in acoustically different office types - A laboratory experiment. Indoor Air 2009;19:454- 67. |
| 5 | Schlittmeier S, Hellbrück J, Thaden R, Vorländer M. The impact of background speech varying in intelligibility: Effects on cognitive performance and perceived disturbance. Ergonomics 2008;51:719-36. |
| 6 | Sörqvist P, Halin N, Hygge S. Individual differences in susceptibility to the effects of speech on reading comprehension. Appl Cogn Psychol 2009;24:67-76. |
| 7 | Evans GW, Johnson D. Stress and open-office noise. J Appl Psychol 2000;85:779-83. |
| 8 | Jahncke H, Hygge S, Halin N, Green AM, Dimberg K. Open-plan office noise: Cognitive performance and restoration. J Environ Psychol 2011;31:373-82. |
| 9 | HRF: S årsrapport 2008: "Adjö yxskaft" [HRF´s annual report 2008: "What's that got to do with the price of fish?"]. HRF - the Swedish association of hard of hearing people. c2008 -. Available from: http://www.hrf.se. [Last cited on 2012 Jan]. |
| 10 | Moore BC. Speech processing for the hearing-impaired: Success, failures, and implications for speech mechanisms. Speech Commun 2003;41:81-91. |
| 11 | Larsby B, Hällgren M, Lyxell B, Arlinger S. Cognitive performance and perceived effort in speech processing tasks: Effects of different noise backgrounds in normal-hearing and hearing-impaired subjects. Int J Audiol 2005;44:131-43. |
| 12 | Rakerd B, Seitz PF, Whearty M. Assessing the cognitive demands of speech listening for people with hearing losses. Ear Hear 1996;17:97- 106. |
| 13 | Arlinger S. Störning av talkommunikation [Disturbance of spoken communication]. In: Arbete och hälsa 1999:27 Störande buller - Kunskapsöversikt för kriteriedokumentation [Work and health 1999:27 Disturbing noise - A knowledge survey for criteria documentation, not available in English]. Stockholm: Arbetslivsinstitutet; 1999. |
| 14 | Harris JD. A brief critical review of loudness recruitment. Psychol Bull 1953;50:190-203. |
| 15 | Hughes RW, Vachon F, Jones DM. Disruption of short-term memory by changing and deviant sounds: Support for a duplex-mechanism account of auditory distraction. J Exp Psychol Learn Mem Cogn 2007;33:1050- 61. |
| 16 | Babisch W. Stress hormones in the research on cardiovascular effects of noise. Noise Health 2003;5:1-11. |
| 17 | Melamed S, Bruhis S. Effects of chronic industrial noise exposure on urinary. Cortisol, fatigue, and irritability: A controlled field experiment. J Occup Environ Med 1996;38:252-6. |
| 18 | Persson-Waye K, Bengtsson J, Rylander R, Hucklebridge F, Evans P, Clow A. Low frequency noise enhances cortisol among noise sensitive subjects during work performance. Life Sci 2002;70:745-58. |
| 19 | Ulrich RS. Aesthetic and affective response to natural environment. In: Altman I, Wohlwill JF, editors. Behavior and the natural environment. New York: Plenum Press; 1983. p. 85-125. |
| 20 | Kaplan R, Kaplan S. The experience of nature: A psychological perspective. New York: Cambridge University Press; 1989. |
| 21 | Hartig T. Three steps to understanding restorative environments as health resources. In: Ward Thompson C, Travlou P, editors. Open space: People space. New York: Taylor and Francis Inc; 2007. |
| 22 | Kaplan R. The role of nature in the context of the workplace. Landsc Urban plan 1993;26:193-201. |
| 23 | Berman MG, Jonides J, Kaplan S. The cognitive benefits of interacting with nature. Psychol Sci 2008;19:1207-11. |
| 24 | Laumann K, Gärling T, Stormark KM. Selective attention and heart rate responses to natural and urban environments. J Environ Psychol 2003;23:125-34. |
| 25 | Bowler DE, Buyung-Ali LM, Knight TM, Pulin AS. A systematic review of evidence for the added benefits to health of exposure to natural environments. BMC Public Health 2010;10:456. |
| 26 | Holding DH, Baker MA. Toward meaningful noise research. J Gen Psychol 1987;114:395-410. |
| 27 | Swedish standard, designation SS-EN 12464-1. Ljus och belysning - Belysning av arbetsplatser - Del 1: Arbetsplatsens utformning [Light and lighting - Lighting of work places - Part 1: Workplace Design]. ICS code 91.160.10; Stockholm: Swedish standards institute 2003. |
| 28 | Manly T, Robertson IH, Galloway M, Hawkins K. The absent mind: Further investigations of sustained attention to response. Neuropsychologia 1999;37:661-70. |
| 29 | Banbury S, Berry DC. Disruption of office-related tasks by speech and office noise. Br J Psychol 1998;89:499-517. |
| 30 | Craik FI, Birtwistle J. Proactive inhibition in free recall. J Exp Psychol 1971;91:120-3. |
| 31 | Åhsberg E, Gamberale F, Kjellberg A. Upplevd trötthetskvalitet vid olika arbetsuppgifter. Utveckling av ett mätinstrument [Perceived fatigue quality during different work tasks. Development of a measure]. Stockholm: Arbetsmiljöinstitutet; 1995. No. 20. |
| 32 | Dickerson SS, Kemeny ME. Acute stressors and cortisol responses: A theoretical integration and synthesis of laboratory research. Psychol Bull 2004;130:355-91. |
| 33 | Yerkes RM, Dodson JD. The relation of strength of stimulus to rapidity of habit-formation. J Comp Neurol Psychol 1908;18:459-82. |
| 34 | Suedfeld P, Loewen LJ. Cognitive and arousal effects of masking office noise. Environ Behav 1992;24:381-95. |
| 35 | Loeb M, Holding DH, Baker MA. Noise stress and circadian arousal in self-paced computation. Motiv Emot 1982;6:43-8. |
| 36 | Tremblay S, Jones DM. Change of intensity fails to produce an irrelevant sound effect: Implications for the representation of unattended sound. J Exp Psychol Hum Percept Perform 1999;25:1005-15. |
| 37 | Hongisto V. A model predicting the effect of speech of varying intelligibility on work performance. Indoor Air 2005;15:458-68. |
| 38 | Gatehouse S, Naylor G, Elberling C. Benefits from hearing aids in relation to the interaction between the user and the environment. Int J Audiol 2003;42:77-85. |
|
