Acoustical Quality in Office Workstations, as Assessed by Occupant Surveys

We analyzed acoustic satisfaction in office environments in buildings surveyed by The Center For The Built Environment (CBE). A total of 23,450 respondents from 142 buildings were included in the analysis. Acoustic satisfaction in the CBE survey is a function of satisfaction with both noise and speech privacy. In the database people are significantly more dissatisfied with speech privacy than noise level (P < 0.01). Occupants in private offices are significantly more satisfied with the acoustics than occupants in cubicles (P<0.01). The results shows also that occupants in open office environments are significantly more satisfied with noise level and speech privacy than occupants working in cubicles (P<0.01). Among occupants dissatisfied with acoustics the most prevalent problems are: “People talking on the phone”, “People overhearing private conversations” and “People talking in surrounding offices”. Over 50% of cubicle occupants think acoustics interfere with their ability to get their job done. INDEX TERMS Noise, speech privacy, acoustic satisfaction, post occupancy evaluation (POE) INTRODUCTION Acoustics are an important attribute of commercial office building design. Noise is probably the most prevalent annoyance source in offices, and can lead to increased stress for occupants (Evans et al. 2000, Sundstrom 1994). Speech privacy may be an even more important effect than noise (Sundstrom 1994). Yet acoustics in most cases do not receive the same level of design attention as thermal, ventilation and other architectural and engineering considerations (Salter et al. 2003). The causes and consequences of poor acoustical performance are perhaps not adequately understood by designers and building owners. It would therefore be valuable to determine from a large population of office buildings how occupants perceive their acoustical environments, and what aspects of office building design are influencing these perceptions. The Center For The Built Environment (CBE) at UC Berkeley maintains an extensive post-occupancy-evaluation (POE) database for commercial buildings. The database contains results from CBE’s web-based Occupant Indoor Environmental Quality (IEQ) survey, which has been used since 1996 to measure occupant response to buildings, diagnose the cause of problems, evaluate new building technologies, identify trends in performance, and benchmark the quality of individual buildings against the population of similar buildings. At the time of this study, 142 buildings had been surveyed in the USA, involving 23,450 occupants. The survey includes modules for office layout, office furnishings, thermal comfort, air quality, lighting, acoustics and building cleanliness and maintenance. Each module is comprised of satisfaction-scale questions which branch to follow-up questions whenever dissatisfaction is indicated, to diagnose the causes of the dissatisfaction. The survey modules have been kept consistent over the years in order to accumulate the large, standardized database needed for benchmarking and trend analysis (Zagreus et al. 2003, 2004). The acoustics module of the database can be analyzed on its own, or examined for relationships to the building performance metrics in the other modules. RESEARCH METHODS When respondents fill out the web-based Occupant IEQ survey, the data are saved in a Microsoft SQL Server 2000 database (Zagreus et al. 2004). The data for this study were imported from the SQL database into Microsoft Excel. * Corresponding author email: [email protected] Proceedings: Indoor Air 2005 2402 Here the data were cleaned and prepared for statistical analysis by the computer program Statistica. The data can be divided up into subjective and objective variables. A subjective variable could be an occupant’s satisfaction with the noise level or satisfaction with the sound-privacy level. The occupants vote on a 7–point satisfaction scale ranging from –3 to 3. Objective variables are demographic or other background data such as gender and office type. Physical factors relevant to the acoustic analysis such as phones ringing, people talking, noise from HVAC system etc, are also found in the branching drill-down questions that follow dissatisfied votes. Subjective variable data analysis The subjective analysis correlates the distribution of noise level or speech privacy votes with a particular aspect of the office environment, like office type (private office, shared office, cubicle with high partitions, cubicle with low partitions, or open office environment without partitions), and gender. Statistical analyses of the subjective variables The P-values represent the probabilities that the Null Hypothesis is true; that there is no difference in the distribution of satisfaction with noise level or speech privacy caused by office types and gender. The P-values were calculated in Statistica using a non-parametric statistic, the Mann-Whitney U-test (Siegel et al. 1988). In this paper, we consider P-values less than 5% (P < 0.05) to be statistically significant. To establish whether there is a statistically significant difference between the two acoustic satisfaction variables, the Wilcoxon Matched-Pairs Signed-Ranks Test is used (Siegel et al. 1988). The correlation coefficient for the two acoustic satisfaction questions is then calculated to determine how important noise-level satisfaction is when people are rating their sound-privacy satisfaction and vice versa. Objective variable data analysis The Occupant IEQ survey asks those respondents who have indicated dissatisfaction with noise level or speech privacy to identify problems contributing to this dissatisfaction on a follow-up page. The percentage of respondents indicating that a particular factor contributes to acoustic dissatisfaction is calculated based upon the number of respondents who saw the acoustic dissatisfaction follow-up questions. RESULTS The following results are based on data from a total of 142 buildings (total of 23450 respondents) from the CBE database. We did not include any instances where the survey had been repeated in a building, or buildings located outside USA. The acoustic quality category consistently receives the lowest average satisfaction score of the nine core satisfaction categories in the Occupant IEQ survey (Figure 1). Figure 1. Average scores for the nine core categories and the two acoustic satisfaction questions in the Occupant IEQ survey based on 23450 respondents The acoustic category score is calculated as an average of the satisfaction scores of two acoustic questions: satisfaction with noise level and satisfaction with speech privacy, (also shown in Figure 1). We can see that the low level of satisfaction with speech privacy reduces the average acoustic category score, so speech privacy dissatisfaction is largely responsible for the low average acoustic ratings. Proceedings: Indoor Air 2005 2403 If the data are divided up by office type, the difference between the two acoustic satisfaction questions is even clearer. Figure 2 shows the average of the two acoustic satisfaction scores.