Effects of acoustic environment on work in private office rooms and open-plan offices longitudinalstudy during relocation
A. Kaarlela-Tuomaalaa, R. Heleniusa, E. Keskinenb and V. Hongistoa*
aFinnish Institute of Occupational Health, Indoor Environment Laboratory, Lemminkaisenkatu 1418 B, FIN-20520, Turku,Finland; bDepartment of Psychology, University of Turku, FIN-20014, Turku, Finland
The aim was to determine how the perceived work environment, especially acoustic environment, and its effectsdiffered in private office rooms and in open-plan offices. The subjects consisted of 31 workers who moved fromprivate office rooms to open-plan offices and who answered the questionnaire before and after the relocation. Privateoffice rooms were occupied only by one person while open-plan offices were occupied by more than 20 persons.Room acoustical descriptors showed a significant reduction in speech privacy after relocation. The noise levelaveraged over the whole work day did not change but the variability of noise level reduced significantly. Negativeeffects of acoustic environment increased significantly, including increased distraction, reduced privacy, increasedconcentration difficulties and increased use of coping strategies. Self-rated loss of work performance because of noisedoubled. Cognitively demanding work and phone conversations were most distracted by noise. The benefits that areoften associated with open-plan offices did not appear: cooperation became less pleasant and direct and informationflow did not change. Nowadays, most office workers, independent of job type, are located in open-plan officeswithout the individual needs of privacy, concentration and interaction being analysed. This intervention studyconsisted of professional workers. Their work tasks mainly required individual efforts, and interaction between otherworkers was not of primary concern, although necessary. The results suggest that the open-plan office is notrecommended for professional workers. Similar intervention studies should also be made for other job types.
Keywords: open-plan offices; landscaped offices; private office rooms; work performance; indoor environment;productivity; speech; office noise; acoustics; office satisfaction; speech intelligibility; irrelevant speech effect
There is a strong and increasing worldwide trend tobuild open-plan offices instead of small office rooms.Small office rooms can be occupied by a single personor can be shared by a couple of persons. In this study, aprivate office room is defined as a small office room,occupied only by one worker. Open-plan offices arefavoured because of, e.g. lower building expenses due tothe smaller number of partitions, lower rent costs dueto higher worker density, better adjustability and betteraccess of daylight especially in deep-framed buildings.Many architects prefer open-plan offices because of lessenclosure and spaciousness. Proponents of open-planoffices believe that they promote cooperation, socialrelations, communication, feedback, solidarity andknowledge-sharing between workers, as well as com-mitment to company values. However, there is norobust scientific evidence to support these beliefsespecially when different job types are considered.
From the building owners point of view, an open-plan office is easier to rent than a private office because
of its apparent flexibility regarding the degree ofenclosure and occupancy of space. Such premises leadto higher occupation and higher profit per investedcapital. It is common that a substantial proportion ofworkers resists open-plan offices. The lack of acousticand visual privacy, increased distraction by noise,reduced workstation size, uncontrollable socialcontacts and interruptions are the most feared featuresof open-plan offices. However, these facts do not playan important role when contacts are made becausethey are difficult to express using simple economicalmeasures, such as square metres per worker.
This study concentrates on the comparison ofprivate office rooms and open-plan offices, applying alongitudinal survey method during the office relocationof a company. The study utilises moremultidisciplinary research methods than any previousstudies in the area. Measurements of building acousticsand subjective responses were included before andafter the relocation. Before presenting the researchquestions, a relevant literature review is presented.
*Corresponding author. Email: email@example.com
Vol. 52, No. 11, November 2009, 14231444
ISSN 0014-0139 print/ISSN 1366-5847 online
2009 Taylor & FrancisDOI: 10.1080/00140130903154579
2. Literature review
There are five essentially different methodologicalapproaches to investigate the perception of workenvironments: cross-sectional survey; longitudinal sur-vey during relocation; retrospective survey; longitudi-nal field experiment; case study. The review focuses onthe first two methods because they give the mostreliable insight into the perceived differences betweenopen-plan offices and private office rooms.
A cross-sectional survey is the simplest method togather comparative information from different envir-onments. It consists of parallel and identical measure-ments in many office buildings. The measurement ismade only once and no experimental change in theworkplace is included. Therefore, the method cannotbe used to prove causal relationships. A summary ofspecial cross-sectional surveys, which have comparedopen-plan offices and private office rooms, is presentedin Table 1. It indicates that human-borne noise,especially co-workers speech, and lack of privacy arethe strongest environmental factors producing dissa-tisfaction in open-plan offices. The studies agree thatprivacy, self-rated performance and ability to concen-trate are lower and distractions stronger in an open-plan office. There is some disagreement about whetheropen-plan offices support communication. Unfortu-nately, there are a number of frequently cited cross-sectional surveys that could not be included in Table 1because the office type is not reported (Klitzman andStellman, 1989, Zweers et al. 1992), the samplecontains only buildings with indoor air problems(Lahtinen et al. 2004) or the study is restricted to acertain office type (Veitch et al. 2007).
A longitudinal study during relocation aims toshow the perceptions of the same occupants in twodifferent offices. These studies contain obvious experi-mental features and causality effects can be proved to alimited extent. The present study belongs to thiscategory, focusing on the comparison of the mostextreme office types, private office rooms and open-plan offices. Although there is permanent debate aboutthe benefits of both office types, there were only twoprevious papers where these two office types are ofprimary interest, keeping in mind that private officeroom is defined as a room for only one worker (Table2). Perception of noise was investigated in both studiesbut the results were not clearly reported. Sundstromet al. (1982) compared four office types. Unfortunately,noise ratings were reported only for the whole work-place but not for the special group experiencing therelocation from private rooms to open-plan offices.Acoustic conditions have been considered very little byZalesny and Farace (1987). It can be concluded thatthe experience of noise has been studied very little even
in these two relocation studies, which have a similarexperimental set-up as the present study.
Table 3 consists of miscellaneous longitudinalsurveys during relocations where private rooms andopen-plan offices have been compared. The workershave had private rooms, shared rooms and open-planoffices both before and after relocation. The relativeshare of the open-plan office has becomemore prevalentafter relocation. These do not present a scientificallyvalid analysis between private rooms and open-planoffices because of the small amount of occupantsexperiencing the relocation from private rooms to open-plan office. Three frequently cited relocation studies,Boyce (1974), Sundstrom et al. (1994) and Brill et al.(1984), have reported only the average result of alloccupants in both open-plan offices and small officerooms. Therefore, they were omitted from Table 3.
In a retrospective relocation survey, subjectsanswered the questionnaire only once (Table 4). Thesubjects were asked to remember the workingconditions before the relocation and compare that tothe present conditions in the new premises. Thereliability of retrospective surveys is questionablebecause the evaluation of prior working conditions isdifficult especially when the questionnaire is performed2 years after the relocation. The studies are inagreement that noise and lack of privacy became aproblem after relocation. Performance changes arecontradictory as is the quality of communication.Retrospective surveys of Boje (1971) and Oldham(1988) were not included in Table 4 because they lack acomparison of private rooms and open-plan offices.
Self-rated performance has been investigated inmany studies shown in Tables 14. However, it was notpossible to distinguish the role of noise in performancedecrements. In this study, the effect of noise onperformance in different office environments is ofprimary concern.
Nowadays, more and more work is informationand knowledge intensive. The performance of complexwork tasks often requires not only a successfulperformance of the individual task but also socialnetworking and teamwork. These set differentrequirements for the work environment and for thecognitive processes. Work tasks are often complex andthe analysis of required skills and process types isdifficult in field conditions. In psychological laboratorystudies, tasks are typically divided into verbal,mathematical, visual and motor tasks. Generalcognitive activities that are needed in human actionare, for example, perception, attention, memoryfunctions and decision making. These activities maysuffer differently from various environmental,situational and individual disturbances. Very fewprevious field studies have considered these questions
1424 A. Kaarlela-Tuomaala et al.
although it is of primary interest for space andworkstation design in workplaces.
Workers may try to compensate for the disruptionand distraction created by different factors. When oneis highly motivated one may be more alert and makean even greater effort. Such an individual state cantherefore have an impact on work performance.Moreover, individual traits such as noise sensitivity(Weinstein 1974), extroversion (Eysenck and Eysenck1972), trait anxiety (Spielberger et al. 1970) and locusof control (Rotter 1966) have been associated withnoise effects on annoyance and performance indifferent laboratory studies. However, very few studieshave been conducted on the effect of individual factorson the perception of noise in offices.
Workers may not only change their strategy ofdoing a task but also behave differently at theworkplace to compensate for the environmentalfactors that decrease their ability to perform. Suchcoping methods are, for example, interruptingworking or taking work home. At work, one may alsotry to affect the factors that are detrimental by, forexample, making a proposal for an improvement tothe management. None of the previous studies hasproperly investigated the ways of coping in differentoffice environments although coping is expected tohave a strong connection with work performance.
Although many of the studies of Tables 14concentrated on noise effects, few of them haveincluded acoustical measurements along withquestionnaires (e.g. Nemecek and Grandjean 1973,Boyce 1974, Sundstrom et al. 1982, Veitch et al. 2007).In addition, most of the above-mentioned studies lacka description of the open-plan office although it playsa fundamental role in the perception of the acousticenvironment and privacy. According to Virjonen et al.(2009), acoustic privacy can vary significantly betweendifferent open-plan offices. Privacy in open-plan officeswith very low screens (bullpen) is significantly lowerthan in open-plan offices where workstations areenclosed by high screens (cubicles). Sound absorbersalso become more efficient in the latter type of spaces.Good speech privacy is easier to achieve in large andhigh open-plan offices than in small ones because oflower intensity of room reflections. In this study, thesefactors have been considered with sufficient accuracyso that the effect of the physical environment can bediscussed along with questionnaire data. This isimportant since the amount of noise variessignificantly between open-plan offices depending onthe amount of telephone work and conversations.
Virjonen et al. (2009) have introduced a newmeasurement method to describe the acousticconditions of open-plan offices. The measurementmethod has been designed to take into account theT
1426 A. Kaarlela-Tuomaala et al.
1428 A. Kaarlela-Tuomaala et al.
most distracting situation where a single speaker istalking and the others are listening. The new method isalso applied in a committee draft of a new measurementstandard (International Organization for Standardi-zation 2009). The measurement is made by placing aloudspeaker in one workstation. The sound pressurelevel (SPL) and speech transmission index (STI) aremeasured at different distances from the speaker.The method is fully described in section 3. STI is astandardised quantity, which is widely used amongacoustic designers. STI can also be predicted in open-plan offices with prediction models (Keranen et al.2007). The relation of STI and subjective speechintelligibility has been well documented by Hongisto(2005). For example, STI 0.50 means roughly that50% of syllables are correctly heard. Haka et al. (2009)have shown in a laboratory experiment that the STI ofspeech directly expresses the level of subjective speechdisturbance. Therefore, its use in this context is closelyargued. Virjonen et al. (2009) validated the measure-ment method in 16 acoustically different offices. Whenthe acoustics were well designed, speech was intelligibleonly within 5 m from the speaker. In the opposite case,speech remained intelligible 20 m from the speaker.Thus, the room acoustic design significantly affectedperceived distractions and privacy. This must also affectthe acoustic judgement of the open-plan office becauseHongisto (2005) has shown that STI explains well theeffects of speech on work performance and subjectivedisturbance.
In this context, it is useful to look at laboratoryexperiments regarding task performance during irrele-vant speech. Colle and Welsh (1976) were among thefirst to find the strong deteriorating effect of irrelevantspeech on the performance of serial memory. Sincethen, speech has been widely used among memoryresearchers. There is some disagreement amongresearchers as to what features of sounds, such asspeech, make them disruptive to performance andwhat psychological theory can explain it. Salame andBaddeley (1982) have argued that speech has obliga-tory access to the phonological part of the articulatoryloop system of working memory. Morris and Jones(1990) have suggested that when speech is irrelevant,its access is not necessarily obligatory. On the otherhand, Banbury et al. (2001) have argued that soundsthat markedly vary acoustically in pitch, timbre ortempo, such as speech, can also be disruptive toperformance although the effect is not as strong as withnormal speech. This is in accordance with thechanging-state hypothesis of Jones et al. (1992), exceptit gives no special role for speech. However, it cannotbe forgotten that speech in native language has astronger effect on work performance than syntheticsine speech having the same degree of variation and
modulation as speech but no verbal content (Tremblayet al. 2000). Later, Tremblay et al. (2000) suggestedthat the meaningfulness of sound makes a difference intasks that involve semantic processing. Also, recentwork by Marsh et al. (2008) shows that tasks thatrequire or encourage the use of semantic processes orstrategies are prone to distraction from the semanticattributes of speech.
Laboratory experiments that have attempted tosolve relevant questions in open-plan officeenvironments have been sparse. However, there aresome studies worth mentioning. Colle (1980) foundthat the SPL of speech did not affect task performance.It was suggested that speech intelligibility is moreimportant than loudness of speech. Thereafter, theeffect of speech intelligibility has been investigated,which created an important link to acoustic conditionsin open-plan offices. Ellermeier and Hellbruck (1998)proved that speech-to-noise-level ratio also explainedthe performance of serial memory in the range of lowspeech intelligibility.
However, subjective speech intelligibility in roomsalso depends on reverberation time, not only onspeech-to-noise-level ratio. Therefore, Hongisto (2005)created a model, based on previously publishedlaboratory experiments, where STI was suggested as apredictor of task performance on cognitivelydemanding tasks during normal speech found intypical office environments. Recent laboratoryexperiments of Venetjoki et al. (2006), Haka et al.(2009) and Kankkunen et al. (2009) confirmed thebasic features of the model. The task performance is atits best when speech does not exist, i.e. STI 0.00.The background noise level does not seem to influenceperformance when the A-weighted SPL is below 50dBA and the spectrum of background is smooth andcomfortable. Performance starts to reduce when STIexceeds 0.30. Performance decrement reaches itsmaximum level at about STI 0.60. In the STI range0.60 and 1.00, comparative experimental data are stilllacking but it is very probable that performance is nolonger affected because subjective speech intelligibilityis perfect in this range. It seems very probable that STIcan also predict the performance of cognitivelydemanding tasks in office environments. The presentstudy aims to show that STI explains well thesubjective performance decrements in real officeenvironments.
The primary aim of the study is to compare theacoustic environments of a private office room andopen-plan office using both occupant questionnairesand acoustic measurements. It was expected thatsatisfaction with office environment would besignificantly reduced after the relocation to theopen-plan office. To raise a wide but solid discussion
about this important topic, the methods of this studyincluded not only noise-related questions but alsoindoor environment, functional efficiency of officespace, psychosocial factors, coping and symptoms.Thus, the results can be utilised not only by acous-ticians but specialists among environmental psychol-ogy, ergonomics, architecture, space design, personnelmanagement, work hygiene, occupational health andindoor air.
3.1. Research motive
This longitudinal study was carried out in a companythat moved from an old office building to a modernnew office building. The old building consisted of onlysmall office rooms for one to four persons while thenew consisted of both small rooms and open-planoffices. The business area of the company wasengineering services and maintenance selling. Thecompany has a long history and a stable line ofbusiness, which reduced the risks of failure of thislongitudinal study. Contact with the company wasformed during the late design process of the new officewhen all decisions regarding acoustic environment hadalready been done. Instead, the company suggestedthat a questionnaire study was conducted because theycould utilise the results in the future when similarspaces are designed. The company did not present anyconditions for the experimental methods. The organi-sational structures and the job descriptions remainedthe same during the research period.
The subjects of this study consisted of 31 workers whomoved from private office rooms to open-plan officesand filled in the questionnaire before and after therelocation. Before the relocation, the personnel of thecompany numbered 121, with 85 working in privaterooms. The rest worked in shared office rooms for twoto four persons. In total, 69 persons out of 85responded before relocation. However, only 31 (45%)of them responded after the relocation. Private officerooms were occupied by one person only. After therelocation, the personnel of the company numbered129, with 88 working in open-plan offices. The workershad fixed workstations both in the old and new officefacility. The principal layouts of offices are presented inFigure 1.
The mean age of subjects was 35 years, rangingfrom 26 to 56 years. Most workers were working full-time. The workers evaluated that they were at theirworkstations most of their working time, on average,6.4 h daily. Of the subjects, 70% were male. The most
typical education was college or technical university.Two worked as supervisors. The subjects had donetheir present work on average for 2.5 years and theyhad worked in the company on average for 8 years.Subjects spent on average 38% (from 4% to 80%) oftheir work time on a typical work day processing textsand about 15% (from 2% to 50%) planning andengaging in creative work, work discussions, phoneconversations and mathematical tasks: 8% of theirtime was spent on practical tasks, such as using acopying machine. Subjects self-rated hearing abilitywas normal. Only two out of 31 reported that theirhearing ability was slightly weaker than average.
3.3. Data collection methods
Data were gathered by acoustical measurements andquestionnaires 2 months before (October 2002) and 4months after (April 2003) the relocation. The subjectswere informed on the first occasion that there would bea second measurement after the relocation. They wereasked to write the first five letters of their mothersmaiden name on the form so that the responses beforeand after the relocation could be matched individually.The questionnaires were mainly given personally to theworkers and they were asked to fill them in at theirworkstation. The questionnaire took about 15 min tocomplete. Subjects were able to ask questions ifneeded. The completed forms were returned to adesignated location or to the researchers during the 2 dwhen the researchers were present. For absent subjects,the questionnaire was left at their desk with a returnenvelope. A third questionnaire was carried out 1 yearafter the relocation (June 2004). Unfortunately, only12 subjects of 31 responded and the results are notpresented.
3.4. Acoustic measurements
The aim was to describe both the noise exposure ofworkers and the spatial attenuation of speech betweenoffice workstations. In this way, the questions can beanswered as to how much noise there is and how itmight affect workers at different distances from thesound source. The main source of noise was expectedto be speech and the measurement methods werechosen correspondingly.
Long-term SPL measurements were carried outwith logging noise dosimeters (CEL-460 and CEL-360;Casella, NH, USA). The aim was to record theambient noise levels in the room. Personal noiseexposure measurements were not made because it isimpossible to separate workers own speech fromexternal speech. The microphone was located on atripod at a height of 1.2 m. The distance
1430 A. Kaarlela-Tuomaala et al.
to the nearest worker was approximately 12 m toavoid the direct sound of a speaking person. Theequivalent A-weighted SPL, LA,eq,7h, expresses theaverage SPL over the 7-h working day. Using the samemicrophones, 1-min equivalent SPL samples, LA,eq,60s,were stored to obtain sound profiles, as shown inFigure 2. From these, A-weighted percentile SPLsLA1% and LA99% were calculated to determine thevariability of noise. LA1% expresses the A-weightedSPL, which was exceeded for 1% of the 7-h workday. It represents the highest SPL in the room.Respectively, LA99% represents the A-weighted SPL,which was exceeded for 99% of the work day. Itrepresents the lowest SPL in the room, being close tothe ventilation noise level. In this study, the variabilityof noise was defined as the difference of LA1% andLA99%.
The speech propagation measurements were car-ried out using the new measurement method described
by Virjonen et al. (2009). The method is very similar tothe committee draft ISO/CD 33823 (InternationalOrganization for Standardization 2009). Themeasurement principle reflects the typical distractingsituation in an office where one worker is speaking andthe other workers must listen to the speech. Listenersat different distances are considered. The measurementresults describe, using only two single-numberquantities, how far from the speaker the speech causesa strong distraction. The quantities are distractiondistance, rD, which is the distance where STI fallsbelow 0.50, and the spatial decay rate of speech, DL2,which is the reduction of A-weighted SPL of speechwhen the distance to the speaker is doubled. Theresults are compared to target values presented byVirjonen et al. (2009).
An omni-directional loudspeaker (Bruel & Kjr4296; Naerum, Denmark) was located in oneworkstation. Locations of measurement points are
Figure 1. Layout descriptions of the offices a) before and b) after relocation. The locations of acoustic measurements areindicated. c) Photograph of one of the four open-plan offices (8 6 24 6 3 m). Here, the screen height was 165 cm. Theceiling was made of perforated plasterboard with weak sound-absorbing properties.
shown in the office layouts of Figure 1. Bothloudspeaker centre and microphone were placed at aheight of 1.2 m, which corresponds to the location ofthe ears and mouth of a seated person. Normal speecheffort was used. The corresponding SPLs at a distanceof 1 m from the mouth in free field are 57, 59, 58, 54,49, 42 and 36 dB in octave bands 1258000 Hz. Thiscorresponds to an A-weighted SPL of 59 dB. Inmeasurement of SPL, pink noise is used. In measure-ment of STI, a sine sweep signal was used to determinethe impulse response between source and receiver. Amicrophone was placed on several workstations atdifferent distances from the loudspeaker, successively.The STI and the equivalent SPL of speech, Lp,S, wasmeasured at several workstations at different distancesfrom the loudspeaker. SPL was measured using a real-time analyser (Bruel & Kjr 2260A) in the listenerslocation. STI was determined according to the originalmethod described by Houtgast and Steeneken (1985),according to which STI is calculated according to themeasured SPL of speech, early decay time and back-ground noise level.
The airborne sound insulation between officerooms before relocation was measured according toISO 1404 (International Organization for Standardi-zation 1995). ISO 7171 (International Organizationfor Standardization 1996) was used to determine theweighted sound reduction index, Rw.
3.5. Questionnaire and statistical analyses
The background of the questionnaire has beenpublished in Finnish by Kaarlela et al. (2003). Themain part of the questionnaire is described in Table 5.Background questions concerned age, gender, level ofeducation, job title, subordinates, time doing the
present work and working for the company, workinghours at the office per week, time working at ones deskon a typical day and level of hearing ability. The resultsof the background information questions werereported in section 3.2. Also, questions regardingtypical stress-inducing factors were asked, e.g.possibilities of influence, support from others, mentalstrain, haste and satisfaction with work. SPSS 11.5 and17 (SPSS Inc., Chicago, IL, USA) were used in thestatistical analysis of questionnaires. The significanceof change was tested by Wilcoxon two-related-samplestest. Paired-samples T-test was used when analysingthe sum variables. The results are presented in means.The significance of group differences in backgroundand individual traits was analysed by w2 test (n 31).The group differences were studied using sum variables(questions 1, 4, 9 and 11 in Table 5) and specificquestions (questions 2, 3 and 10 in Table 5).
The change of equivalent SPL during the workingday, LA,eq,7h, and the variation of SPL, LA1%-LA99%,during the relocation was tested by Wilcoxon two-related-samples test. The associations between themand the subjective data (question 5 in Table 5) weredetermined by correlation coefficient. Before therelocation, analysis with the subjective data was doneusing six subjects and with 31 subjects after therelocation. Significant results are reported whenp 0.05.
3.6. Description of the office spaces
Prior to the move, the company worked in a two-storey cellular office with a thin building frame andsingle corridor plan (Figure 1a). Thus, all rooms weresmall but equipped with a window. The rooms wereoccupied by one to four workers. The rooms were
Figure 2. An example of a sound level measurement covering the whole working day in a private office room (workstation4 in Figure 3).
1432 A. Kaarlela-Tuomaala et al.
equipped with a door but they were usually open to thecorridor. The drywall partitions between the roomsextended from floor to ceiling to provide proper soundinsulation.
The company moved to a new two-storey buildingon the neighbouring site. The new office was acombination of small office rooms and open-planoffices (Figure 1b). The wall between the small roomsand open-plan office area was transparent to permitnatural light penetration. The open-plan offices werenarrow, only 8 m wide and approximately 25 m long.The workstations were enclosed on two or three sideswith screens of a height of either 127 cm or 165 cm.The ceiling was covered by perforated plasterboardwith sound absorption class C (EN 11654; Interna-tional Organization for Standardization 1997), whichcorresponds to moderate noise reduction coefficient 0.50. All walls and the floor were acousticallyreflecting. Strong sound reflections from them
combined with a low level of ventilation noise (lowspeech masking) meant that the office screens becameineffective regarding attenuation of speech. The spatialattenuation of speech was almost equal with the twodifferent screen heights. Therefore, the open-planoffices with low and high screen heights wereconsidered acoustically equivalent in this study.
4.1. Noise and acoustics
The equivalent A-weighted SPLs, LA,eq,7h, did notdiffer significantly from each other in the open-planand private office rooms (Figure 3). Instead, thevariability of noise, LA,1% LA,99%, was higher in theprivate rooms than in the open-plan office (p 5 0.001,Figure 3). The average A-weighted SPL and thevariability of SPL were not associated with the self-rated disturbance caused by noise.
Table 5. Main content of the questionnaire including the values of Cronbachs alpha of sum variables.
Questions (explanation of the items)
Work environment Private room Open-plan
1 How much have the following indoor environmental factorsdisturbed you at your workstation during the last 3 months? (9 items)
A 0.79 0.74
2 How satisfied are you with the work environment as a whole? B3 How satisfied are you with the acoustical conditions as a whole at
4 How much do the following sounds disturb your concentration onyour work at your workstation? (12 sound sources)
A 0.88 0.82
5 The number (112) of the most disturbing sound (selection from 12alternatives as in item 4)
Effects of sounds A6 How much do the sounds disturb the following types of work? (six
work types)7 How often do you act in the following way to cope with your work
because of the sounds in your work environment? (12 methods)D 0.81 0.80
8 When you think about the effects of the sounds in your workenvironment, how many min are wasted per day?
9 How well can you carry on a discussion at your work station withoutraising your voice?
10 How well can you hear speech that you dont need to hear for yourwork?
Well-being11 How much of the following symptoms or feelings have you had
during the last 3 months? (eight symptoms)A 0.87 0.87
12 How well do the following statements describe you? (four items fornoise sensitivity)
E 0.85 0.89
13 How often do you feel the following? (four items for trait anxiety) F 0.90 0.8714 How well do the following statements describe you? (three items for
extroversion)E 0.80 0.74
15 How well do the following statements describe you? (two items forinner locus of control)
E * *
Work space16 How well do the following statements apply to your work space? (11
Scale: A 1 not at all 5 very much; B 1 very dissatisfied 5 very satisfied; C open question; D 1 never 5 very often; E 1very poorly 5 very well; F 1 almost never 5 almost always; G 1 not at all 5 very well.*Private office: r 0.480; open office: r 0.737.
The STI values were higher between adjacent open-plan workstations than between adjacent office rooms,which predicts higher disturbance by speech. Inaddition, the spatial attenuation of STI was signifi-cantly weaker in the open-plan office (Figure 4a).While the radius of distraction, rD, was less than 3 m in
the private office, it was 9 m in the open-plan office.The spatial decay rate of speech was DL2 8 dB inthe open-plan office (Figure 4b).
According to Figure 4a, STI between private officerooms was 0.42 when doors were open, while it was0.76 between two neighbouring workstations in the
Figure 3. Average noise level and the variation of noise at several workstations: A-weighted equivalent sound pressure level,LA,eq,7h and variability of noise, LA1% LA99%. Workstation averaged values are given in parentheses. The recorded noiseincludes all kinds of sounds in the office.
Figure 4. The spatial decay of speech between workstations in the open-plan office and private room office: (a) spatial decay ofspeech transmission index (STI); (b) spatial decay of sound pressure level of speech. The faster STI and sound pressure level ofspeech are reduced, the better speech privacy is achieved. Letters A, B and C refer to the nearest workstation to the speaker.Corresponding sound spectra are presented in Figure 5.
1434 A. Kaarlela-Tuomaala et al.
open-plan office. When the doors are closed, STI fallsto 0.20. The frequency content of speech heard fromneighbouring workstations is illustrated in Figure 5 inboth office types.
In the open-plan office, the A-weighted SPL ofspeech was 52 dB in the next workstation (point A inFigure 4b). In the private office room, the value was41 dB in the next room (point B) when doors wereopen. Closing the doors reduced the SPL down to35 dB (point C), which is lower than the backgroundnoise level of the room, 39 dB. Thus, it was signifi-cantly more difficult to hear speech in the private roomoffice.
Figure 6. How much have the following indoor environmental factors disturbed you at your workstation duringthe last 3 months? Mean values and the significance of change (p-value).
Figure 5. Frequency analysis of speech at the nearest workstation to the speaker. The letters A, B and C refer toindicated measurement points in Figure 4.
4.2. Subjective ratings
The satisfaction with the work environment as a wholedecreased after relocation to open-plan offices(p 5 0.05). Workers were more satisfied with theacoustic environment in the private office rooms(p 5 0.01). The disturbance caused by indoor envir-onment factors is presented in Figure 6. Noisebothered subjects more after relocation (p 5 0.001),while the other factors bothered less. The disturbanceby all the indoor environmental factors together wasgreater in the open-plan than in the private office (sumvariable, p 5 0.01). The self-estimated distractioncaused by different sounds is presented in Figure 7. In
the open-plan office, unnecessary speech could beheard easier, which was obvious also from acousticaldata (p 5 0.01). It was more difficult to conversewithout raising ones voice in the open-plan offices(p 5 0.01). Speech and laughter were most oftenchosen as the most disturbing sounds in both officetypes.
Noise disturbed the most work, which consisted ofreading, writing or processing (Table 6). In the open-plan office, mathematical tasks and phone conversa-tions were disturbed significantly more. Difficultieswith concentration increased slightly after relocation(Table 7). According to Figure 8, the use of copingstrategies was more common in the open-plan offices.The use of all coping strategies increased (sumvariable, p 5 0.001). The self-estimated waste ofworking time due to noise doubled after relocation(p 5 0.05) (Table 8). After relocation, subjects experi-enced significantly less privacy and peace for workingand cooperation became significantly less pleasant anddirect (Figure 9).
Subjects who experienced their work as moredemanding used more coping strategies in the open-plan offices (p 5 0.05). Women were more noise-sensitive and more disturbed by noise (p 5 0.05,open-plan). They were also more disturbed by theindoor environmental factors and used copingmethods more than men (p 5 0.05, private). The
more noise-sensitive subjects more often used copingstrategies because of sounds and more of their workingtime was wasted due to noise compared with the lessnoise-sensitive subjects (p 5 0.05). Noise sensitivityincreased after moving to the open-plan office(p 5 0.05). Anxiousness predicted higher prevalenceof symptoms (p 5 0.01). There was no significantdifference in the level of the experienced common stressfactors measured before and after the relocation.
Table 6. How much do the sounds disturb the followingtypes of work? (Mean values and the significance of thedifference (pvalue)).
Text processing,writing, reading
3.0 3.1 NS
Mathematical tasks,billing, statistics
2.2 2.9 50.01
Planning or creativework, programming
2.6 2.7 NS
Discussions connectingwith work
2.3 2.3 NS
Telephone discussions 2.5 3.0 50.05Practical organisation(e.g. copying)
1.6 1.6 NS
Scale: 1 not at all; 5 very much.
Figure 7. How much do the following sounds disturb your concentration on your work at your workstation? Mean valuesand the significance of change (p-value).
1436 A. Kaarlela-Tuomaala et al.
5.1. Noise and acoustic measurements
Time-averaged SPLs over the working day werepractically the same in both office types. This disagreeswith general judgements, according to which open-plan offices are noisier than private offices. Noisiness isa subjective descriptor but a noise level meter does notseem to be the most appropriate tool for its assessmentin different office environments. This finding can beexplained by room acoustical facts and normal humanbehaviour. First, the private office rooms were small(10 m2) compared to the open-plan offices (at least
200 m2). Even though the distance from the noisemeter to the next worker was the same in both offices,the SPL caused by a conversation was amplified by thenearby hard walls in private office rooms. In open-planoffices, the reflecting surfaces are more distant, whichreduced the SPL of identical conversation. Typically,the difference is 2 to 5 dB. Second, a stronger voiceeffort is easily used in a private room because of betteracoustic privacy.
According to Figure 3, the average variation ofnoise levels was larger in private office rooms. Inprivate rooms, there are clear periods of silence,dominated by ventilation noise, and periods ofconversation inside the room. Speech coming from thecorridor and other rooms does not increase the averageSPL because their SPL is much less than 50 dB (seeFigure 5). In the open-plan office, conversations fromnearby or distant workstations can be heard most ofthe time and silent periods seldom occur. This explainsthe higher variability of noise in the private room.
The variability of SPL was not associated with thesubjective disturbance of noise in either the privaterooms or open-plan offices. The reason for this isobvious: the noise level meter does not indicate whichsounds are relevant to the worker and which are not.When speech is present in a private room, it is typicallyuseful for the worker; it is caused by a colleaguevisiting the room. In the open-plan office, most of thespeech will be experienced as noise because speechsounds are heard from a larger area.
Table 7. How much of the following symptoms or feelingshave you had during the last three months? (Mean values andthe significance of change (pvalue)).
Prevalence of symptoms
Concentration problems 2.4 2.8 0.05Memory problems 2.0 2.1 NSMotivation problems 2.3 2.2 NSTiredness and overstrain 3.0 2.8 NSNegative feelings such
as feeling displeased2.5 2.6 NS
Headache, neck andshoulders pain
3.1 2.8 NS
Scale: 1 not at all; 5 very much.
Figure 8. How often do you act in the following way to cope with your work because of the sounds in your workenvironment? Mean values and the significance of change (p-value).
Keighley (1970) suggested that the variability ofoffice noise described the perceived noise annoyance inoffices better than average noise level. The presentresults support this suggestion if one looks at noisemeasurements only, but it is not suggested thatvariability of noise could be used alone as an indicatorof acoustical conditions in private rooms. The reason isthat high variability of SPL predicts high amounts ofuseful speech for the occupant in the private room,whereas in open-plan offices, high variability predictshigh amount of useless speech and risk of disturbance.Therefore, the use of noise variability as a predictor ofdisturbance in a workstation is not justified. Properanalysis of variability would require post processing,where useful and useless speech is separated. This ispractically impossible. Therefore, STI is a significantly
more appropriate quantity for the prediction of speechdisturbance because the discussion about useless anduseful speech is not needed.
According to Figure 4a, speech could be clearlyunderstood at much longer distances from the speakerin the open-plan office compared to the private office.In the open-plan office, the radius of distraction, i.e.the distance at which STI drops below 0.50, wasrD 9 m. According to Figure 4b, the spatial decayrate of speech with distance doubling DL2 was 8 dB.The stronger speech attenuates with distance, thebetter speech privacy is achieved. The open-plan officereaches an acoustic class, CC, which is a very typicallevel (Virjonen et al. 2009). The best possible class, AA,requires rD 4 5 m and DL2 4 11 dB. The roomacoustical quality of the open-plan office fulfils theminimum requirements for teamwork, wherecommunication between workers is continuous and animportant part of work. Speech privacy was notperceived as being sufficiently high in this workplacebecause STI was very high between neighbouringworkstations. The acoustic results of the open-planoffice fall within typical results obtained in open-planoffices (Hongisto et al. 2004, Virjonen et al. 2009).
There is a significant difference regarding speechintelligibility between the studied office types. Whenthe doors were closed/open, STI was 0.22/0.42 betweenneighbouring office rooms while it was 0.76 betweenneighbouring open-plan workstations. The reason forthe high STI value was low background noise level andlow amount of room absorption.
Table 8. When you think about the effects of the sounds inyour work environment, how many min are wasted per day?(n 21).
Mean 18 36Median 15 20Minimum 0 0Maximum 60 120Higher quartile 30 60Lower quartile 6 8Percentage of daily work time 7 h 4.3 8.6
Figure 9. How well do the following statements describe your work space? Mean values and the significance of change(p-value).
1438 A. Kaarlela-Tuomaala et al.
The sound insulation between private office roomswas low. The weighted sound reduction index betweennearby office rooms was Rw 27 dB when both doorswere closed. The recommended minimum value is35 dB, which is usually easily achieved. As a result,speech could still be understood over the partition wall(Figure 5). The acoustic conditions of private officerooms were poorer than usually obtained. Typically,private rooms have better sound insulation. It can besuggested that the subjective perception of the studiedprivate room office was poorer than could be generallyfound in private office rooms. Respectively, theacoustic difference between a private room office andopen-plan office became especially small. On thesegrounds, the large differences in subjective ratingsbetween the two experiment office types were notexaggerated by large acoustic differences but vice versa.The subjective differences between private room officesand open-plan offices are expected to be larger ingeneral when the acoustic differences are larger.
The simultaneous use of absorptionmaterials, sound-absorbing screens and speech masking sound wouldproduce the lowest STI and, thus, best speech privacy inopen-plan offices (Virjonen et al. 2009). The open-planoffice lacked sufficient masking sound and absorptionmaterials. The background noise level of ventilation wasLA,eq 39 dB while the recommended noise level is42 dB. Absorption materials were placed only to theceiling and their absorption efficiency was too low. Thewalls and furniture were strongly sound reflecting.
5.2. Indoor environment and noise sources
Subjectswere less satisfiedwith the open-plan offices thanthe private office rooms as a whole, despite the fact thatthe new office fulfilled the requirements for a modernhead office. Moreover, the satisfaction with the acousticenvironment was lower. The disturbance caused by officenoise had increased andnoise became themostdisturbingindoor environmental factor. This result is in agreementwithprevious studies,whereopen-planandprivateofficeshave been directly compared (Tables 14).On the basis ofmeasurement results, it is plausible to suggest thatinadequate room acoustic design was one of the mainreasons for lowoffice satisfaction in open-plan offices.Allother indoor environment factors, except for noise,improved after the relocation (Figure 6). In the open-plan office, perception of thermal conditions and indoorair (temperature, draught, smells, tobacco smoke) wereimprovedbecause ofmodern and efficient ventilation andcooling systems.
According to Figure 7, speech and laughter werethe most disturbing sounds in both offices but theywere more distracting in the open-plan offices. Thetelephone ringing was the second most distracting
sound in the open-plan office. Constant andpredictable noises, e.g. ventilation and computers,caused very little distraction in the open-plan office. Inthe private office, traffic and construction soundsoriginated from the construction of their new officebuilding nearby. Because the new building wasequipped with a centralised outlet ventilation system,fresh air needed to be taken in through the facade airsupplies, thus causing the easy transmission ofconstruction and traffic noises to the building interior.
In the private office rooms, vibrations were alsoexceptionally distracting. They were caused by thelightweight intermediate floor, which is not typicallyused in Finland. In the open-plan office, the floor wasmade of concrete and the construction work wasalmost completed. Therefore, the distraction of thesefour sound sources was reduced dramatically.
The average background SPL was LA,eq 39 dBin both private and open-plan offices, but the variationwas larger in the private office rooms, 3644 dB, andnegligible in the open-plan office, 3741 dB. Therefore,the distraction caused by ventilation was dramaticallyreduced after relocation.
5.3. Noise and work performance
According to Table 6, noise disturbed most theprocessing of text, reading and writing in both offices.Counting and billing tasks were more disturbed bynoise in the open-plan office than in the private office.A work task such as billing also involves mathematicalskills. Office noise has been shown to impairperformance on prose and mental arithmetic tasks inlaboratory experiments (Banbury and Berry 1997,1998). Logie and Baddeley (1987) and Banbury andBerry (1998) have also found that the performance ofmathematical tasks suffers from speech disturbance.Banbury et al. (2001) argued that the obligatory accessof sound to memory results in cognitive organisationalactivities such as seriation. This again, conflicts withthe deliberate process of seriating the task-relevantmaterial. Jones et al. (2008) state that the performanceof both serial recall and semantic memory tasks mayalso suffer from inhibiting task inappropriateinformation that comes from the obligatory analysis ofsounds.
Marsh et al. (2008) have shown that the semanticsimilarity between the relevant and irrelevant materialdictates the degree of disruption by speech to theongoing semantic task. In office environments, thesurrounding speech often consists of semanticallysimilar material to the material processed in differentwork tasks. Surprisingly, noise disturbance did notreduce the performance of cognitively demanding taskssuch as planning, creative work and compiling after
relocation. The possible deterioration of performancemay be difficult to estimate in the aforementionedtasks. The result that tasks of practical organisationseemed to be left unaffected by noise was no surprise.In future studies, a more detailed analysis of worktasks is needed in order to better associate the findingsof cognitive laboratory studies to work tasks in thefield.
Phone conversations were more disturbed in theopen-plan office. One could more easily hear irrelevantspeech and it was more difficult to converse withoutraising ones voice. This suggests that it is not onlyspeech intelligibility, i.e. the STI value that predicts thedistraction; too high SPL of surrounding speech canalso affect tasks involving conversation. According tothe Lombard effect, people naturally raise their voicelevel by 0.6 dB per 1.0 dB increase in background SPLwhen the ambient SPL exceeds 45 dB (Pick et al. 1989).This supports the fact that the SPL of speech needs tobe considered along with STI when room acousticalconditions are evaluated in open-plan offices, as shownin Figure 4.
In the cross-sectional survey of Pejtersen et al.(2006), the prevalence of symptoms was at a higherlevel in open-plan offices than in private office rooms.In this study, only concentration problems wereexperienced slightly more after the relocation. Therewas no difference in the amount of memory problems(Table 7). This result of concentration difficulties is inaccordance with other field studies (e.g. Pejtersen et al.2006, Haapakangas et al. 2008). According to the fieldstudy of Banbury and Berry (2005), it is difficult tohabituate to varying and unpredictable sounds withhigh information content. According to the literaturereview of Banbury et al. (2001), irrelevant sounds tendto impair cognitive performance by breaking throughselective attention. Because of the stimulus in theenvironment, people may feel more alert during theday. If one compensates for the detrimental effects ofnoise with greater effort, it may result in fatigue. In thepresent study, fatigue was experienced the most,followed by headache and neck and shoulder pain,but the strength of these symptoms did not changeafter the relocation. Inquiries about subjects feelingsafter the work day might be considered in futurestudies.
Work motivation did not change after the reloca-tion. This is in agreement with Pejtersen et al. (2006),whose cross-sectional survey included both privateoffice rooms and open-plan offices of different sizes.Open-plan office noise has been shown to result inmotivational deficits in laboratory conditions, e.g.Evans and Johnson (2000). It is suggested thatmotivation is significantly lower in laboratory condi-tions than during the persons actual office work. Work
motivation seems to depend more strongly on otherfactors than physical environment.
The level of negative feelings did not change eventhough there was a strong increase in the use of copingstrategies. According to a cross-sectional survey byKlitzman and Stellman (1989), noise was one of thestrongest physical and psychological predictors ofpsychological well-being at offices. Noise explained therating of office satisfaction but only modestly forfatigue, irritation, generalised distress and jobsatisfaction. Unfortunately, Klitzman and Stellmandid not report the office type.
According to Figure 8, subjects systematically usedmore coping strategies in the open-plan office. Previousstudies have not included the measurement of copingstrategies. Some coping strategies have either a director an indirect effect on work performance. Subjectsmade an even greater effort after the relocation becauseof noise, which was evident from the increased use ofcoping methods, e.g. working more slowly andworking more over time. This agrees with Becker et al.(1983). Subjects organised their work differently in theopen-plan than in the private rooms due to noise: theyput their work off till another time and interruptedwork. Subjects also left their desk, changedworkstation or did their work at home because of noisemore often in the open-plan compared to the privateoffice. These kinds of interruptions result in a directloss of work performance. Workers discussed the noiseproblem with their colleagues and managers moreoften after moving to the open-plan office. Behaviouralrules or office etiquette are recommended in the open-plan offices to minimise the detrimental effects ofinterruptions and distractions.
Even though subjects tried to be quieter at theirworkstation in the open-plan office than in the privaterooms, it was more difficult because at the same timethey found it more difficult to converse without raisingtheir voice. This suggests that office etiquette will notprovide a sufficiently acceptable acoustic workenvironment. The studied open-plan office sufferedfrom poor or moderate room acoustic design. Properroom acoustic design might have helped the situationsignificantly.
Noise caused the waste of working time significantlymore in the open-plan office than in the private roomoffice (Table 8). Self-rated overall performance hasbeen reported to decrease in several previous studiesafter moving to an open-plan office (see Tables 14).However, there are only a few previous studies thathave addressed the effect of noise alone onperformance. For example, Haapakangas et al. (2008)used nearly the same measures in their cross-sectionaloffice survey. The average waste of working time was20 min in open-plan offices and 9 min in private offices
1440 A. Kaarlela-Tuomaala et al.
(p 5 0.05). The results were in agreement with thepresent study. These two results give strong support tothe previous suggestions made by numerous laboratoryexperiments that work performance will deteriorate dueto speech in open-plan offices.
5.4. Functional efficiency of office layout
According to Figure 9, people experienced less privacyand peace for work in the open-plan office. This is inagreement with most previous studies presented inTables 14. Privacy means the possibility to retreatfrom people, to control information flow and toregulate interaction (Sundstrom and Sundstrom1986). Acoustic privacy is an essential part of it inthe open-plan office. Privacy is one of the factors thatexplain psychological well-being at work (Klitzmanand Stellman 1989). Whether workers can get used toreduced privacy is also a topic for future study.
According to Figure 9, cooperation was experi-enced significantly less pleasant and direct in the open-plan office than in the private office, althoughinformation flow and approachability of colleaguesdid not change. Yet, it has been suggested by severalauthors that open-plan offices facilitate communica-tion, cooperation and approachability (e.g. Hundertand Greenfield 1969, Allen and Gerstberger 1973).However, published studies where cooperation orcommunication was found to improve in open-planoffices represent workplaces where continuous coop-eration and teamwork was necessary. Thus, the open-plan office was not the best office solution for theworkers of the current experimental group because thesubjects were professionals doing mainly individualwork. The office layout had no effect on the work flowin the company of the present study. The private officeclearly supported their work better than the open-planoffice. The results might have been different in anorganisation mainly involved in teamwork. Thesefindings together support the suggestion that the jobdemands of each worker or team should be carefullyanalysed before new office premises are designed.
5.5. Individual and background factors
Coping strategies were used more by those whose workwas more demanding. This was an expected result.Complex tasks that demand intensive concentrationand cognitive capacity suffer more than simple,well-rehearsed motoristic tasks. Zalesny and Farace(1987), for example, have shown that job rankinfluences employee reactions to the office environ-ment. The influence of job complexity should beinvestigated in future studies to find scientific evidencefor the suggestion that cognitively demanding
professional and managerial workers should not beplaced in open-plan offices if their daily work consistsmainly of individual work or confidential matters.
Women reported more noise disturbance. In futurestudies, the possible difference between the genders inwork performance and well-being in the open-planoffice should not be neglected. Extroversion, traitanxiety and locus of control were not associated withperceptions of noise and its adverse effects in this studyeven though such associations have been found inlaboratory conditions.
Measuring individual noise sensitivity, however, isrecommended in future studies. In the present study,noise sensitivity increased after moving to the open-plan office. The noise sensitivity items were: I am verysensitive to sounds: they disturb me, I can concentrateirrespective of what is happening around me, I easilyget used to most sounds and Sometimes the soundsget on my nerves and make me irritable. Although theselected questions were general in nature, they seem tohave reflected the change in the acoustic environment.An alternative noise sensitivity questionnaire has beendeveloped recently by Griefahn (2008), which isexpected to be more appropriate for workplaces.
The relocation did not result in a significant changein work satisfaction nor in the measured stress-inducingfactors. According to the meta-analysis of de Croonet al. (2005), there is strong evidence that working inopen-plan offices decreases job satisfaction. Therelationship between physical office satisfaction(lighting, privacy, acoustics, air quality, thermalconditions) and cost-related measures such as workefficiency, sick leave and turnover intentions should befurther investigated to promote investments in a betterwork environment. Also, the positive effect of involvingworkers in office design and the relocation processshould be remembered. Allen and Gerstberger (1973)and Sprecklmeyer (1993) reported longitudinalrelocation studies (Table 3) where office satisfactionafter the relocation was enhanced by managementtechniques such as participation and involvement inoffice design.
6. Practical implications
It seems that the open-plan office was not the ideal typeof office environment for the workers of this company,where individual cognitive work was done instead ofdynamic interactive work. Therefore, none of theindicators improved after relocation from the privateoffice rooms to the open-plan office. It is possible thatdifferent results would have been obtained ifcontinuous interaction between workers had beenmore important in this workplace, where individualwork was emphasised.
Significantly greater attention should be paid toindividual job demands and individual needs forspeech privacy when office space is designed. Privateoffice rooms are recommended for workers performingmainly individual and cognitively demanding workand needing confidential conversations. Small inter-active teams should be located in shared rooms and,preferably, not to large open-plan offices because theteam is difficult to isolate from other parts of the openspace. Open-plan offices can be used for non-intensiveand dynamic work or travelling workers. In all cases,anonymous private rooms should be provided for theopen-plan office workers because of incidental periodsof concentration-demanding work, paired work orprivate conversations. The use of alternative work-station types in the same office building should beencouraged because this facilitates the selection ofworkstation according to the current work task andthis improves the feeling of control over the workenvironment. Common rules for behaviour should becreated together in each department. Workmatesshould be encouraged to continuously take fellowworkers into consideration, e.g. by having privatephone calls and short conversations in anonymousrooms, using low speech levels whenever possible andnot leaving their mobile phones ringing on the workdesks. It is also recommended that workers areinvolved with workplace design and that the relocationprocess is managed by using professionals.
Globally, most office workstations are alreadylocated in open-plan offices. On the grounds of thisstudy, the improvement of their acoustic environmentwould be of primary importance if higher officesatisfaction is desired. However, there is very littlescientific evidence that better acoustic design wouldlead to better subjective ratings. In the future,subjective ratings of acoustically well-designed open-plan offices should be compared to the poorly designedopen-plan office to answer this important question asto whether workers could be significantly moresatisfied with the office when acoustics are correctlyimplemented. This kind of investigation should bedone as a longitudinal field experiment (intervention),where the same subjects experience the same open-planoffice before and after acoustic renovation. Attemptsof such field experiments are presented by Heleniusand Hongisto (2004) and Hongisto (2008).
The room acoustic failure of open-plan offices isvery usual because only a few countries have publishedguidelines for their acoustic design. The committeedraft ISO/CD 33823 (International Organization forStandardization 2009) has been written to moderatethis problem. Proper acoustic design of open-planoffices presupposes consideration of several factorssimultaneously, e.g. screen height, degree of enclosure,
absorption material in room surfaces and furniture,distance between workstations and speech masking. Tofacilitate all-inclusive acoustic design, a free Internet-based design tool has been developed by Keranen et al.(2007).
Thanks are due to the company who put their premises andworkers at our disposal. The work was started in theProductive Office 2005 project, which was financed by 27companies and Tekes. The work was completed during theMAKSI project (Modelled and perceived indoorenvironment), which was financed by 11 companies andTekes. Thanks to Ms Annu Haapakangas for givingcomments on this manuscript.
Allen, T.J. and Gerstberger, P.G., 1973. A field experiment toimprove communication in a product engineeringdepartment: The non-territorial office. Human Factors,15, 487498.
Banbury, S. and Berry, D.C., 1997. Habituation anddishabituation to speech and office noise. Journal ofExperimental Psychology: Applied, 3, 181195.
Banbury, S. and Berry, D.C., 1998. Disruption of office-related tasks by speech and office noise. British Journal ofPsychology, 89, 499517.
Banbury, S.P. and Berry, D.C., 2005. Office noise andemployee concentration: Identifying causes ofdisruption and potential improvements. Ergonomics,48 (1), 2537.
Banbury, S.P., et al., 2001. Auditory distraction and short-term memory: phenomena and practical implications.Human Factors, 43 (1), 1229.
Becker, F.D., et al., 1983. Office design in a communitycollege. Environment and Behavior, 15, 699726.
Boje, A., 1971. Open-plan offices. In: B.H. Walley, ed.London: Business Books.
Boyce, P.R., 1974. Users assessments of a landscaped office.Journal of Architectural Research, 3, 4462.
Brennan, A., Chugh, J.S., and Kline, T., 2002. Traditionalversus open office design A longitudinal field study.Environment and Behavior, 34 (3), 279299.
Brill, M., et al., 1984. Using office design to increaseproductivity, volume 1. Buffalo, NY: Workplace Design& Productivity.
Brookes, M.J. and Kaplan, A., 1972. The office environment:Space planning and affective behavior. Human Factors,14 (5), 373391.
Colle, H.A., 1980. Auditory encoding in visual short-termrecall: effects of noise intensity and spatial location.Journal of Verbal Learning and Verbal Behavior, 19, 722735.
Colle, H.A. and Welsh, A., 1976. Acoustic masking inprimary memory. Journal of Verbal Learning and VerbalBehavior, 15, 1731.
Danielsson, C., 2005. Office environment, health and jobsatisfaction. Licentiate Thesis. Stockholm, Sweden: KTHTechnology and Health.
de Croon, E.M., et al., 2005. The effect of office conceptson worker health and performance: a systematicreview of the literature. Ergonomics, 48 (2),119134.
1442 A. Kaarlela-Tuomaala et al.
Ellermeier, W. and Hellbruck, J., 1998. Is level irrelevant inirrelevant speech? Effects of loudness, signal-to-noiseratio, and binaural unmasking. Journal of ExperimentalPsychology: Human Perception and Performance, 24 (5),14061414.
Evans, G.W. and Johnson, D., 2000. Stress and open-officenoise. Journal of Applied Psychology, 85 (5), 779783.
Eysenck, H.J. and Eysenck, S.B.G., 1972. Eysenck Person-ality Inventory, Form A. 2nd ed. London: University ofLondon Press, Hazell Watson & Viney, Bucks.
Griefahn, B., 2008. Determination of noise sensitivity withinan internet survey using a reduced version of the NoiseSensitivity Questionnaire. Acoustics, 2008/3460.
Haapakangas, A., et al., 2008. Perceived acoustic environ-ment, work performance and well-being survey resultsfrom Finnish offices. In: 9th international congress onnoise as a public health problem (ICBEN) 2008, 2125July, Mashantucket, Connecticut, USA. Dortmund,Germany: Leibniz Research Centre for Working Envir-onment and Human Factors, 434441.
Haka, M., et al., 2009. Performance effects and subjec-tive disturbance of speech in acoustically differentoffice types laboratory experiment. Indoor Air. Avail-able from: http://www3.interscience.wiley.com/journal/122386634/abstract?CRETRY1&SRETRY0.
Helenius, R. and Hongisto, V., 2004. The effect of acousticalimprovement of an open-plan office on workers. In:Proceedings of Inter-Noise 2004, paper 674, 2125August, Prague, Czech Republic.
Hongisto, V., 2005. A model predicting the effect of speech ofvarying intelligibility on work performance. Indoor Air,15, 458468.
Hongisto, V., 2008. Effects of sound masking on workers acase study in a landscaped office. In: 9th internationalcongress on noise as a public health problem (ICBEN)2008, 2125 July, Mashantucket, Connecticut, USA.Dortmund, Germany: Leibniz Research Centre forWorking Environment and Human Factors, 442449.
Hongisto, V., Keranen, J., and Larm, P., 2004. Simple modelfor the acoustical design of open-plan offices. ActaAcustica United with Acustica, 90, 481495.
Houtgast, T. and Steeneken, H.J.M., 1985. A review of theMTF concept in room acoustics and its use forestimating speech intelligibility in auditoria. Journal ofthe Acoustical Society of America, 77 (3), 10691077.
Hundert, A.T. and Greenfield, N., 1969. Physical space andorganizational behavior: a study of office landscape. In:Proceedings of the 77th annual convention of the AmericanPsychological Association, 601602.
International Organization for Standardization, 1995. ISO1404:1995 (E). Acoustics -measurement of sound insula-tion in buildings and of building elements, part 4: Field mea-surements of airborne sound insulation of between rooms.
International Organization for Standardization, 1996. ISO7171:1996 (E). Acoustics rating of sound insulation inbuildings and of building elements, part 1: airborne soundinsulation.
International Organization for Standardization, 1997. ENISO 11654:1997. Acoustics Sound absorbers for use inbuildings Rating of sound absorption.
International Organization for Standardization, 2009. Com-mittee draft ISO/CD 33823. Acoustics Measurement ofroom acoustic parameters Part 3: Open plan spaces (ISOTC 43 SC 2, N969, 20090414).
Ives, R.S. and Ferdinands, R., 1974. Working in a land-scaped office. Personnel Practice Bulletin, 30 (2), 126141.
Jensen, K.L., Arens, E., and Zagreus, L., 2005. Acousticalquality in office workstations as assessed by occupantsurveys. In: Proceedings of 10th International Conferenceon Indoor Air Quality and Climate, 49 September,Beijing, China, CD-ROM, 24012405.
Jones, D., et al., 2008. Varieties of auditory distraction. In:9th international congress on noise as a public healthproblem (ICBEN), Foxwoods. Dortmund, Germany:Leibniz Research Centre for Working Environment andHuman Factors.
Jones, D.M., Madden, C., and Miles, C., 1992. Privilegedaccess by irrelevant speech to short-term memory: Therole of changing state. Quarterly Journal of ExperimentalPsychology: Human Experimental Psychology, 44A, 645659.
Kaarlela, A., et al., 2003. Acoustical environmentquestionnaire for offices method development. WorkEnvironment Research Report Series 4 (In Finnish,abstract in English). Helsinki, Finland: Finnish Instituteof Occupational Health.
Kankkunen, E., et al., 2009. Comparison of five speechmasking sounds for open-plan offices Laboratoryexperiment measuring performance and acousticsatisfaction. The Journal of the Acoustical Society ofAmerica. April. Submitted for publication.
Keighley, E.C., 1970. Acceptability criteria for noise in largeoffices. Journal of Sound Vibration, 11 (1), 8393.
Keranen, J., Virjonen, P., and Hongisto, V., 2007. A newmodel for acoustical design of open-plan offices. In:International congress on acoustics, paper RBA-10006-IP, 27 September, Madrid, Spain. Madrid: SpanishAcoustical Society.
Klitzman, S. and Stellman, J.M., 1989. The impact of thephysical environment on the psychological well-being ofoffice workers. Social Science Medicine, 29 (6), 733742.
Kraemer, Sieverts & Partner, 1977. Open-plan offices: Newideas, experience and improvements. London, UK:McGraw-Hill Book Company.
Lahtinen, M., Sundman-Digert, C., and Reijula, K., 2004.Psychosocial work environment and indoor air problems:a questionnaire as a means of problem diagnosis.Occupational and Environmental Medicine, 61, 143149.
Logie, R.H. and Baddeley, A.D., 1987. Cognitive processesin counting. Journal of Experimental Psychology:Learning, Memory and Cognition, 13, 310326.
Marsh, J.E., Hughes, R.W., and Jones, D.M., 2008.Auditory distraction in semantic memory: A process-based approach. Journal of Memory and Language, 58,682700.
Morris, N. and Jones, D.M., 1990. Habituation to irrelevantspeech: Effects on a visual short-term memory task.Perception and Psychophysics, 47 (3), 291297.
Nemecek, J. and Grandjean, E., 1973. Results of anergonomic investigation of large-space offices. HumanFactors, 15 (2), 111124.
Oldham, G.R., 1988. Effects of changes in workspacepartitions and spatial density on employee reactions: Aquasi experiment. Journal of Applied Psychology, 73,253258.
Oldham, G.R. and Brass, D.J., 1979. Employee reactions toan open-plan office: A naturally-occurring quasi-experiment. Administrative Science Quarterly, 24, 267284.
Pejtersen, J., et al., 2006. Indoor climate, psychosocial workenvironment and symptoms in open-plan offices. IndoorAir, 16 (5), 392401.
Pick, H.L., et al., 1989. Inhibiting the Lombard effect.Journal of the Acoustical Society of America, 85 (2), 894900.
Rotter, J.B., 1966. Generalized expectancies for internalversus external control of reinforcement. PsychologicalMonographs: General and Applied, 80, 128.
Salame, P. and Baddeley, A.D., 1982. Disruption of short-term memory by unattended speech: Implications for thestructure of working memory. Journal of Verbal Learningand Verbal Behavior, 21, 150164.
Spielberger, C.D., Gorsuch, R.L., and Lushene, R.E., 1970.STAI-manual for the State-Trait Anxiety Inventory.California, USA: Consulting Psychologists Press.
Sprecklmeyer, K.F., 1993. Office relocation and environ-mental change: a case study. Environment and Behavior,25, 181204.
Sundstrom, E., Herbert, R.K., and Brown, D.W., 1982.Privacy and communication in an open-plan office: a casestudy. Environment and Behavior, 14 (3), 379392.
Sundstrom, E. and Sundstrom, M.G., 1986. Work places.The psychology of the physical environment in offices andfactories. New York, USA: Cambridge University Press.
Sundstrom, E., et al., 1994. Office noise, satisfaction, andperformance. Environment and Behavior, 26 (2),195222.
Tremblay, S., et al., 2000. The irrelevant sound effect: Doesspeech play a special role? Journal of ExperimentalPsychology: Learning, Memory and Cognition, 26 (6),17501754.
Veitch, J.A., et al., 2007. A model of satisfaction with open-plan office conditions: COPE field findings. Journal ofEnvironmental Psychology, 27 (3), 177189.
Venetjoki, N., et al., 2006. The effect of speech and speechintelligibility on task performance. Ergonomics, 49 (11),10681091.
Virjonen, P., Keranen, J., and Hongisto, V., 2009.Determination of acoustical conditions in open-planoffices Proposal for new measurement method andtarget values. Acta Acustica United with Acustica, 95,279290.
Weinstein, N.D., 1974. Effect of noise on intellectualperformance. Journal of Applied Psychology, 59 (5), 548554.
Zalesny, M.D. and Farace, R.V., 1987. Traditional versusopen offices: a comparison of sociotechnical, socialrelations, and symbolic meaning perspectives. Academyof Management Journal, 30 (2), 240259.
Zweers, T., et al., 1992. Health and indoor climatecomplaints of 7043 office workers in 61 buildings in theNetherlands. Indoor Air, 2, 127136.
1444 A. Kaarlela-Tuomaala et al.