Content uploaded by Esko Keskinen
Author content
All content in this area was uploaded by Esko Keskinen
Content may be subject to copyright.
Performance: 9th International Congress on Noise as a Public Health Problem (ICBEN) 2008, Foxwoods, CT
Perceived acoustic environment, work performance and well-being -
survey results from Finnish offices
Annu Haapakangas*
1
, Riikka Helenius
1
, Esko Keskinen
2
, Valtteri Hongisto
1
1 Finnish Institute of Occupational Health, Indoor environment laboratory, Lemminkäisenkatu 14-18
B, 20520 Turku, Finland, Europe.
2 University of Turku, Department of psychology, 20014 Turku, Finland
*corresponding author: e-mail: annu.haapakangas@ttl.fi
ABSTRACT
The aim of the study was to investigate workers' perceptions of the acoustic
environment of offices and study its relations to work performance and well-being.
Questionnaire results from 11 companies and 689 respondents were analyzed.
Occupants in private rooms and open offices were compared. Noise was the main
indoor environmental problem in open offices. Speech was the most distracting
source of noise in both office types but the degree of disturbance was lower in private
rooms. About half of open office occupants and 20 % of occupants in private rooms
were dissatisfied with acoustics. Office noise disturbed particularly conversations and
tasks relying on working memory and verbal processes, such as text comprehension
and creative thinking. Routine tasks were little disturbed by noise. In open offices,
attempts to cope with noise reflected risk factors to individual productivity and well-
being, such as taking extra breaks, compromising the quality of work, working
overtime and exerting oneself harder. Self-estimated waste of daily working time due
to noise was twofold in open offices. Open office workers experienced more stress
symptoms, particularly overstrain and difficulties in concentration, and attributed
these symptoms to office noise to a greater extent than workers in private rooms.
Possibilities to influence issues related to one's work and work-space privacy were
lower among open office occupants. The results suggests that private rooms are
superior to open offices in all respects.
INTRODUCTION
There is an increasing worldwide trend to build open offices instead of private room
offices. Open offices are preferred because of better space economy, spaciousness
and flexibility. Open offices are also assumed to facilitate teamwork and information
sharing. However, there is continuous debate on the positive and negative effects of
open offices on work performance and well-being. According to a meta-analysis car-
ried out by DeCroon et al. (2005), there is strong evidence that working in open of-
fices reduces worker’s psychological privacy and job satisfaction. Some evidence
exists that cognitive workload increases in open offices.
Cross-sectional office surveys that have compared different office lay-outs, e.g.
Becker et al. (1983), Danielsson (2005), Pejtersen et al. (2006) and Jensen et al.
(2005), have shown that the most severe factor causing office dissatisfaction is noise.
Danielsson (2005) compared several office lay-outs and concluded that dissatisfac-
tion with noise and privacy was highest in large open offices and lowest in cellular
offices. Pejtersen et al. (2006) found that the percentage of occupants complaining
about noise was ten-fold in large open offices compared to cellular offices. The same
study demonstrated an association between office size and several symptoms, in-
cluding fatigue, headache and difficulties in concentration. Open office occupants
Performance: 9th International Congress on Noise as a Public Health Problem (ICBEN) 2008, Foxwoods, CT
have also reported more subjective performance loss due to noise than cellular office
workers, e.g. in the amount of accomplished work (Becker et al. 1983).
The present study aims to improve the general understanding of open office condi-
tions and its effects on work performance and worker well-being. This is done by us-
ing a questionnaire method that addresses a wide range of issues related to noise
disturbance, its effects on work and workers and the functional performance of office
lay-out. Open office conditions are compared to conditions in private offices. This pa-
per continues the work of Helenius et al. (2007) using to large extent the same mate-
rial.
METHODS
Subjects
A total of 689 subjects from 11 office buildings took part in the study. In addition,
there were 60 respondents occupying shared offices of 2 to 4 people but their results
are not reported in this paper. Data was gathered between 2002 and 2008. Back-
ground information of the data is presented in Table 1. The acoustical conditions of
the office buildings represented typical Finnish offices built after 1990. Seven of the
studied companies had a combination of private rooms and open offices while three
companies had mainly open offices and one had only private rooms. The number of
respondents varied between 13 and 196 in different companies. Different lines of
business were included in the sample. The survey always targeted all workers of a
department participating in the study so the workers represented a wide range of pro-
fessions not enumerated here.
Table 1: Background information
Number of respondents
Private room Open office
Age in years
Range (mean)
Female
%
Male
%
Sample A 93 260 19-65 (44,4) 36,8 63,2
Sample B 88 248 20-65 (40,9) 63,4 36,6
Full sample 181 508 19-65 (42,7) 49,7 50,3
Questionnaire
An office acoustics questionnaire was developed on the basis of a literature review
and a pilot study. The questionnaire had several sections. Indoor environment and
Noise sources covered the disturbance of indoor environmental factors in general,
the disturbance of specific noise sources, satisfaction with work environment and
acoustic satisfaction. Noise effects covered the disturbance of different work tasks,
behavioral efforts to cope with noise and self-estimated waste of daily working time
due to noise. Well-being covered general stress symptoms and symptoms attributed
to office noise. Psychosocial environment covered psychosocial stress factors, e.g.
job satisfaction and hurry at work. Office lay-out performance assessed quality of
teamwork and communication, privacy, comfort and availability of practical resources
in the office area. Work space preference was also inquired with one question. Most
questions were answered on a 5-point Likert scale. Individual factors, e.g. noise sen-
sitivity, were also assessed but are not reported in this paper.
Some modifications were made to the questionnaire during the research period and
some companies did not allow all sections to be included, e.g. questions about the
Performance: 9th International Congress on Noise as a Public Health Problem (ICBEN) 2008, Foxwoods, CT
psychosocial issues. Therefore, the number of respondents varies in different ques-
tions and is reported separately for each analysis. In the sections Indoor Environment
and Noise Sources the data from all companies could not be combined because of a
change in the phrasing of the question: in half of the offices (Sample A), respondents
were asked to rate how often they were disturbed whereas the other half (Sample B)
rated how much they were disturbed. Asking about frequency instead of degree of
disturbance resulted in higher estimates of distraction and some of the differences
were statistically significant. Data was therefore analyzed separately for the two sub-
samples in these specific sections.
RESULTS
The data was analyzed with SPSS 16.0 statistical program. The comparisons be-
tween open offices and private rooms were performed using Mann-Whitney U-test.
Indoor environment. Noise was the main indoor environmental problem in open of-
fices in both samples. Open office occupants were significantly more disturbed by
noise than workers in private rooms. Open office occupants also complained more
about other indoor environment factors than did workers in private rooms. The distur-
bance of indoor environment factors was similar in both subsamples and only the
subsample focusing on the frequency of disturbance is reported in Table 2. Distur-
bance caused by noise is reported for both samples.
Table 2: The average disturbance of indoor environment factors. Scale 1-5, with 1 indicating no dis-
turbance and 5 indicating highest level of disturbance. Subsample A rated the frequency of distur-
bance while subsample B rated the degree of disturbance.
Mean (SD)
Sub-
sample
N Items
Cronbach's
alpha
Private room Open office
P-value
Thermal conditions
A 344 2 0,683 2,11 (0,89) 2,40 (0,98) 0,016
Air quality
A 346
4 0,800
1,93 (0,79) 2,19 (0,80) 0,007
A 346 1 - 2,45 (0,97) 3,55 (1,16) 0,000
Noise
B 335 1 - 2,50 (0,97) 3,29 (1,05) 0,000
Lighting (amount of
light and glare)
A 352 1 - 2,12 (0,98) 2,51 (1,12) 0,004
Noise sources. Results for the Sample A focusing on the frequency of disturbance
of sounds are shown in Table 3. The most distracting sound sources in open offices
were speech near one's work station and sounds of phones ringing. Speech also dis-
turbed private room occupants the most but not to the same extent. The pattern of
disturbance from different sounds was similar in Samples A and B, except that in
Sample B open office occupants were less disturbed by ventilation noise (p < .01)
than private room occupants.
Satisfaction with the indoor environment as a whole and acoustic satisfaction were
lower among open office occupants (Table 4). Fifty percent of open office occupants
were dissatisfied with acoustics at their work station while only 21 percent of private
rooms occupants were dissatisfied.
Performance: 9th International Congress on Noise as a Public Health Problem (ICBEN) 2008, Foxwoods, CT
Table 3: The disturbance to concentration caused by different noise sources. The table shows mean
values for Sample A on the scale from 1 (never disturbs) to 5 (disturbs very often); standard deviations
in brackets.
Mean (SD)
N
Private room Open office
P-value
Speech in open office (near one's desk) 244 not relevant 3,40 (1,24)
Speech from adjacent rooms 221 2,33 (1,08) 2,02 (1,16) 0,009
Speech from common facilities, e.g. coffee
rooms
335 1,83 (1,04) 2,33 (1,27) 0,001
Ventilation noise 335 1,56 (0,85) 2,03 (1,15) 0,000
Own pc 335 1,44 (0,74) 1,57 (0,78) ns
Office equipment 337 1,45 (0,77) 2,15 (1,07) 0,000
Phones ringing 335 1,97 (0,85) 3,05 (1,13) 0,000
Radio, music 337 1,30 (0,64) 1,54 (0,72) 0,002
Traffic on corridors, doors, elevator 336 2,03 (0,95) 2,58 (1,24) 0,000
Construction work, reparations 336 1,63 (0,76) 1,74 (0,71) ns
Sounds made by others working 333 1,25 (0,46) 2,08 (1,13) 0,000
Environmental noise from outside 335 1,44 (0,67) 1,34 (0,53) ns
Table 4: Satisfaction with work environment as a whole (n=422) and satisfaction with acoustics at
one's work station (n=464). Outermost classes are combined in the table but statistical tests were
conducted using original distributions. Percentages within office types are shown.
Satisfaction with work envi-
ronment (p <.001)
Satisfaction with acoustics
(p <.001)
Private room Open office Private room Open office
very or somewhat dissatisfied 9,3 30,4 21,1 50,0
neutral 14,7 20,8 21,1 21,4
very or somewhat satisfied 76,0 48,8 57,9 28,6
Table 5: Disturbance of different types of tasks due to workplace noise. The table shows mean values
for disturbance on the scale from 1 (not at all disturbed) to 5 (very much disturbed).
Noise effects. Conversations and complex verbal tasks, such as text processing and
planning, were more disturbed by noise in open offices than in private rooms (Table
5). Routine work and arithmetic tasks were less affected and the degree of distur-
bance did not differ between the office types. Self-estimated waste of daily working
time due to noise was higher in open offices (Table 6).
Behavioral efforts to cope with noise took place more often in open offices than in
private rooms (Table 7). These included taking extra breaks, exerting one-self harder,
N Items
Cronbach's
alpha
Private room
Mean (SD)
Open office
Mean (SD)
P-value
Conversations
653 2 0,834 2,12 (1,01) 2,78 (1,15) 0,000
Complex verbal tasks
622 2 0,754 2,46 (1,02) 2,78 (1,19) 0,003
Routine work
593 1 - 1,35 (0,69) 1,45 (0,76) ns
Arithmetic tasks
560 1 - 2,02 (1,02) 2,23 (1,28) ns
Performance: 9th International Congress on Noise as a Public Health Problem (ICBEN) 2008, Foxwoods, CT
working overtime and doing remote work. Quality of work was also more often com-
promised in open offices in order to cope with noise. Compromising quality of work
correlated with overall coping (r = .702, p < .01) but it was left out of the sum variable
because of a lower number of respondents.
Table 6: Self-estimated waste of daily working time due to noise in minutes
N
Private room
Mean (SD)
Open office
Mean (SD)
P-value
Wasted working time 615 12,00 (15,00) 21,48 (20,11) 0,000
Table 7: Behavioral coping efforts and stress symptoms on a scale from 1 to 5 (1 indicating no symp-
toms/coping and 5 indicating very much symptoms/coping)
N Items
Cronbach's
alpha
Private room
Mean (SD)
Open office
Mean (SD)
P-value
Coping efforts
581 4 0,840 1,76 (0,64) 2,14 (0,81) 0,000
Compromising quality of work
444 1 - 1,43 (0,69) 1,85 (1,02) 0,000
Stress symptoms
380 4 0,890 2,27 (0,75) 2,64 (0,95) 0,000
Well-being. Overall stress was higher among open office workers (Table 7). Sepa-
rate analyses of each symptom showed that particularly difficulties in concentration
were more prevalent among open office occupants (Table 8). Workers in open offices
also experienced more tiredness and exhaustion. Irritation and motivational difficul-
ties seemed to be more prevalent among open office occupants but the difference
failed to reach statistical significance.
Table 8: Prevalence of stress symptoms and the percentage of occupants attributing symptoms to
noise. Prevalence of symptoms was evaluated on a scale from 1 (not at all) to 5 (very much). Percent-
age of noise-related symptoms is calculated for the population expressing little or more symptoms
(values 2-5 in the symptom prevalence question).
Prevalence of symptom,
mean (SD)
Occupants attributing
symptom to noise, %
N
Private Open
P-value
Private Open
P-value
Irritation 476 2,36 (0,88) 2,58 (1,10) 0,058 16,3 48,6 0,000
Tiredness or
exhaustion
450 2,60 (0,93) 2,91 (1,05) 0,008 9,3 42,1 0,000
Difficulties in
concentration
475 2,24 (0,94) 2,69 (1,16) 0,000 23,0 56,3 0,000
Motivational
difficulties
407 2,19 (1,01) 2,41 (1,12) 0,064 10,7 28,2 0,000
Percentage of occupants attributing symptoms to office noise was calculated for the
population that indicated having symptoms (values 2 'little' to 5 'very much'). About 87
percent of private room occupants and 82 percent of open office occupants belonged
to this group. Those who indicated that their symptoms might be due to office noise
'to some degree' or more (values 3 to 5) were considered to attribute the symptom to
noise. Those respondents whose symptoms were 'little' or 'not at all' due to noise
(values 1 to 2) were considered not to have noise-related symptoms. The results
show that open office occupants attributed symptoms to office noise to a greater ex-
tent than private office occupants.
Performance: 9th International Congress on Noise as a Public Health Problem (ICBEN) 2008, Foxwoods, CT
Psychosocial environment. Most psychosocial stress factors did not differ between
the office types (Table 9). Open office occupants received more support from co-
workers or managers than did workers in private offices. This may also reflect the
content of work as it is likely that workers with private offices have more independent
job descriptions, and therefore, less need for support. Possibilities to influence issues
related to one's work were perceived lower among open office occupants.
Table 9: Psychosocial stress factors. Scale 1= not at all, 5= very much
N
Private room
Mean (SD)
Open office
Mean (SD) P-value
Hurry at work 392 3,48 (0,85) 3,43 (0,90) ns
Work feels interesting and inspiring 336 3,66 (0,90) 3,61 (0,93) ns
Mental strain experienced at work 444 3,33 (0,80) 3,25 (0,90) ns
Support received from co-workers or manager 248 3,25 (0,94) 3,54 (0,81) 0,015
Possibilities to influence one's work 303 3,26 (0,95) 2,88 (0,96) 0,003
Job satisfaction 380 3,82 (0,68) 3,71 (0,82) ns
Functional performance of the office lay-out. Open office occupants experienced
lower privacy in their work area than private room occupants (Table 10). Less practi-
cal resources, such as work space and meeting facilities, were perceived to be avail-
able in open offices. Comfort was assessed more negatively in open offices than in
private offices. Contrary to expectations, the quality of teamwork and communication
did not differ between open offices and private offices. In fact, the mean values for
the quality of communication are nearly identical. The sum variable for teamwork in-
cluded statements such as, 'colleagues are within easy reach', 'information is shared
well between colleagues' and 'collaboration is effective'.
Table 10: Functional performance of the office lay-out. Factors have been assessed on a scale from 1
to 5, with '1' indicating most negative assessment and '5' most positive
N Items
Cronbach's
alpha
Private room
Mean (SD)
Open office
Mean (SD)
P-value
Privacy 492 2 0,888 3,83 (0,78) 2,34 (1,04) 0,000
Teamwork and communication 490 4 0,845 3,75 (0,62) 3,74 (0,73) ns
Availability of practical resources 489 3 0,714 3,77 (0,74) 3,37 (0,84) 0,000
Comfort 489 2 0,776 3,33 (0,74) 2,66 (0,95) 0,000
Work space preference. Results for work space preference are shown in Table 11.
The results show that 21 percent of open office occupants prefer working in open
offices and 33 percent would choose a shared office.
Table 11: Workspace preference in percentages for open office and private room occupants (N=569)
Prefered work space
Work space at present
Private room
Shared office of 2 to
4 persons Open office
Private room
97,6 0,6 1,8
Open office
46,9 32,6 20,5
Performance: 9th International Congress on Noise as a Public Health Problem (ICBEN) 2008, Foxwoods, CT
DISCUSSION
The results suggest that open offices have versatile acoustic problems in terms of
subjective disturbance, performance effects and worker well-being. The expected
benefits of open offices regarding functional efficiency were not supported by the re-
sults. Modern office work is increasingly characterized by cognitively demanding
tasks in which background noise is perceived as particularly disturbing. The study
gives no support to an extensive preference of open offices when workers' well-being
and efficiency are of main concern.
Objective measurement of performance effects of noise is very difficult in real offices.
In this study, subjective evaluations of wasted working time due to noise were higher
in open offices than in private rooms, providing one measure for the performance
effects of noise. The estimates of lost minutes cannot be regarded as exact in objec-
tive terms but the finding that workers change their behavior to cope with noise sup-
ports the conclusion that working time is indeed wasted because of noise. For exam-
ple, open office occupants reported taking extra breaks and rescheduling work due to
workplace noise.
Open office occupants suffered more from difficulties in concentration and tiredness.
In open offices, a greater percentage of those suffering from symptoms attributed
symptoms to office noise. It seems unlikely that the higher stress levels among open
office occupants were due to differences in psychosocial work environments as these
factors were mostly assessed similarly in open and private offices. Our results are in
line with the view that open office conditions and the accompanying lack of privacy
form an extra stress factor to an individual worker. Further analyses will be conducted
with the data to address the relations between office type, stress, indoor environment
and psychosocial environment in more depth.
The results contradict the most common assumption of the benefits of open office
layouts, that is, facilitation of communication and co-operation. The quality of team-
work and communication did not differ between open offices and private rooms at all.
However, it is likely that in most of the studied open offices the respondents' work
was characterized mainly by individual performance in which constant availability of
colleagues and information exchange is not necessary. Open offices may be suitable
for specific jobs that are mainly comprised of teamwork.
The study does not suggest that open offices should not be used. Twenty-one per-
cent of open office workers preferred open offices to other office types. Although evi-
dence could not be presented in this study, it is probable that many of these persons
have a continuous need for communication with colleagues. The main problem
seems to be that the selection of occupant's workstation is not based on the analysis
of job demands. As periods of individual work and telephone conversations are still
predominant in most office professions, open offices do not provide sufficient acous-
tic, visual and psychological privacy for typical office work.
Future studies should include a more detailed analysis of the job type of open office
workers and a more detailed analysis of the open office type. There are very large
differences in the size of open offices, in the facilities available to workers and also in
the flexibility of workstations which were not considered in this study. It is very impor-
tant to be able to develop instructions for designing most appropriate work environ-
ments.
Acoustic conditions in the open offices can vary significantly. Hongisto et al. (2007)
have shown that in good offices, the distraction of speech restricts to 5 meters from
Performance: 9th International Congress on Noise as a Public Health Problem (ICBEN) 2008, Foxwoods, CT
the speaker while in worst offices, the speech distracts up to 20 meters from the
speaker. The acoustic quality of open offices should also be measured in future sur-
veys to show how acoustic problems depend on acoustic design.
ACKNOWLEDGEMENTS
This study was carried out as a part of a national research programme MAKSI (Per-
ceived and Modelled Indoor Environment) funded by Tekes, National Technology
Agency, Finnish Institute of Occupational Health and several participating companies.
Thanks are due to the companies and office workers who participated the study. The
authors thank Henri Riuttala for valuable advice on the statistical analyses and Anu
Kaarlela-Tuomaala and Jaana Jokitulppo for co-operation during early years.
REFERENCES
Becker FD, Gield B, Gaylin K, Sayer S (1983). Office design in a community college. Environ Behav 15: 699-726.
Danielsson C (2005). Office environment, health and job satisfaction. Stockholm, Sweden: KTH Technology and Health,
Licentiate Thesis.
DeCroon EM, Sluiter JK, Kuijer PPFM, Frings-Dresen MHW (2005). The effect of office concepts on worker health and per-
formance: as systematic review of the literature. Ergonomics 48: 119-134.
Helenius R, Keskinen E, Haapakangas A, Hongisto V (2007). Acoustic environment in Finnish offices - the summary of ques-
tionnaire studies. In: International Congress on Acoustics, paper RBA-10-001, 2-7 September 2007, Madrid, Spain.
Hongisto V, Virjonen P, Keränen J (2007). Determination of acoustic conditions in open offices and suggestions for acoustic
classification. In: 19th International Congress on Acoustics, Madrid, Spain, Sept 2-7, 2007.
Jensen KL, Arens E, Zagreus L (2005). Acoustical quality in office workstations as assessed by occupant surveys. In: Pro-
ceedings of Indoor Air 2005, pp 2401-2405.
Pejtersen J, Allermann L, Kristensen TS, Poulsen OM (2006). Indoor climate, psychosocial work environment and symptoms
in open-plan offices. Indoor Air 16: 392-401.
