Original Articles
A pilot study on the effects of playing background music in the workplace on overtime work
2021 Volume 3 Issue 1 Article ID: 2021-0005-OA
Details
2021 Volume 3 Issue 1 Article ID: 2021-0005-OA
Objectives: We investigated whether playing background music (BGM) in the workplace had any effects on overtime hours. Methods: In 15 workplaces, we used a crossover design and alternated between 2 months with and without BGM for 8 months. Using the attendance management records of the company, we documented overtime hours during the observation and the previous year. BGM was played at the end of working hours (A), from 15:00 to the end of work (B), and from beginning to the end (C) in each of the five arbitrarily selected offices. To evaluate the effect on overtime hours, we used a paired t-test to compare the mean overtime between periods with and without BGM and repeated analysis of variance to compare the changes in the trend of overtime in the same months between the observation year and the previous year by an interaction effect. Results: Patterns A, B, and C consisted of 625, 536, and 573 employees, respectively. The mean overtime hours for 4 months were significantly decreased in one (−4.3%), two (−19.2% and −10.7%), and three (−11.8%, −16.8%, and −4.4%) offices, respectively. Furthermore, the trend of overtime between the observation year and the previous year significantly changed in one (1/5), one (1/5), and three (3/5) offices, in patterns A, B, and C, respectively. Conclusions: There were offices that showed less overtime when playing BGM, particularly the entire day, than when BGM was not played. This pilot study suggests that conducting more extensive research in this area is worthwhile.
Health problems caused by long working hours have become a worldwide concern. The causal association between long working hours and the onset of cerebral and cardiovascular diseases has been extensively investigated1,2,3,4). A systematic review by Kivimäki et al.1) showed that the risk of developing cerebral infarction significantly increases with working hours per week. In a 20-year prospective cohort study by Hayashi et al.2), the risk for acute myocardial infarction was particularly high in men in their 50s who worked 11 hours or more per day.
Long working hours and mental illness are significantly associated5,6,7). A systematic review by Tsuno et al.5) revealed that long working hours are associated with depressive symptoms, particularly in women, young people, nonmanagers, neurotic disorders, and in an environment where subjects worked for >80 hours per month.
In terms of physical and mental aspects, long periods of overworking may be fatal; this is called “Karoshi”, or occupational sudden death, and is a health burden not only in Japan but also across the world8). Prevention of excessive overtime is an important occupational health issue. Shorter working hours may result in more regular exercise and higher smoking cessation rates9); whereas, longer working hours are directly linked to health problems10).
Recently, with the development of information technology (IT) equipments and changes in social life, new working styles have been widely practiced, including discretionary schedules, flextime, and working at home. Employees have been able to set the working place and working hours by themselves11). The objectives of flexible working styles tailored to the individual are to prevent long working hours and to increase employee motivation and performance. However, they are not always effective in reducing working hours12), and working at home may actually increase overtime hours13).
Currently, legal regulations and/or work-related rules are the most effective means to reduce overtime8). For medical nurses, the importance of reducing overtime hours has been frequently studied14,15). In one study, overtime hours were reduced before and after the introduction of an overtime limit14). In another study, the introduction of IT tools for nursing records effectively reduced overtime hours for medical nurses15). However, reports on specific methods to reduce overtime hours for other occupations are limited.
Playing background music (BGM) is a familiar stress reduction method to make an individual feel comfortable16). In health care, music therapy has been used in the field of psychiatry. BGM is used as a complementary treatment to increase the therapeutic responsiveness of patients with depression17), and it is used to change problem behavior and improve mood in elderly individuals with dementia18). BGM is also used as a rhythmic timekeeper or signal stimulus during exercise rehabilitation for patients with movement disorders19). Mainly, a faster tempo of music induces an arousal effect20). Music not only has a stress-reducing effect but also creates a motivational effect, depending on the type of music21). BGM has a positive effect on promoting communication22). According to Haake23), workplace music helps to increase work engagement and escape from work. Therefore, we assumed that music helps develop the conditioning and psychological effects necessary for work.
We hypothesized that playing BGM in the workplace may help to reduce overtime hours by conditioning the employees to recognize the end of work and by creating a natural awareness of the passage of time and/or psychological motivation to improve work efficiency. Thus, this study aimed to investigate whether playing BGM in a workplace was possibly associated with less work overtime, particularly in the office.
The Institutional Review Board of Clinical Research of Tokai University approved this study (No.16R-278). We recruited 15 business offices from 15 companies. The inclusion criteria were as follows: (1) had 50 or more employees, (2) was an IT-related company that had relative self-control of the working hour settings, and (3) could objectively calculate working hours using a time management system. Trial Registration: UMIN-CTR, UMIN000038158. Retrospectively registered on October 1, 2019.
Methods in playing BGMWe alternately played three BGM patterns in the office for 2 months and a “silence” period (no music) for another 2 months for a total of 8 months (eFigure 1). We used “Sound Design for OFFICE,” a background music service released in 2013 by USEN Corporation (Tokyo, Japan) that is ideal for work environment. In pattern A (offices 1–5), a slow-tempo healing music was played as BGM for 5 min, and a female announced the same sentence every day: “It’s time to finish work. Please enjoy your hobbies and leisure time. Thank you for your hard work all day today.” In pattern B (offices 6–10), BGM was played from 15:00 to the end of the day. In pattern C (offices 11–15), BGM was played from the start of work to the end. The genres of music included instrumental music selected from healing, jazz, and fusion styles. BGM was a mix of several genres of music according to workplace preferences. Essentially, the contents of music were planned to bring about a natural awareness of the time passage and of the need to work efficiently. For example, in an office with pattern C, healing music was played in the morning to soothe the employees’ body and mind after a long commute to work. As the work continues, a slightly faster tempo music was played. Loud music and music with vocals was played at lunchtime to make people aware of the lunch break and of the change in pace. Light tempo music was played post lunch, considering that workers tend to feel drowsy in the afternoon. Finally, faster tempo music was played again to motivate the employees to increase their efforts. In the office with a flextime system, BGM was played during the hours when most of the employees were working. These patterns were arbitrarily selected and acquired consent from each company. The BGM volume should have been set lower (6–10 db) than the office noise in the place closest to the speakers.
Schedule of BGM playingThe periods with and without BGM (silence) were set for 2 months each alternately. With regard to the periods of BGM or silence, the first 2 months from the start of the study were determined as period I, and every 2 months thereafter were defined as periods II, III, and IV sequentially. Offices 4, 9, 14, and 15 played BGM in periods I and III, and the others played in periods II and IV. The study was started in the month when the BGM apparatus was installed in the office. We examined offices 1, 2, and 4–15 before the Work Style Reform Act24) took effect, while office 3 was studied after the enforcement of the act.
Information on overtime hoursWe calculated the working hours daily from the entry time (when logged into the system) to the exit time (when logged out from the system) of employees according to the attendance management of each workplace. Overtime hours were defined as the hours spent working beyond the specified working hours per day at each business location. Then, the overtime hours per month were totaled for each month. In this study, we caluculated the overtime hours for 2 months because the periods with and without BGM were alternated every 2 months. Considering the change in monthly workload, the overtime hours for the same month in the previous year were calculated in the same manner.
Subjective evaluationUsing the website, we conducted self-administered questionnaire surveys four times in the last month of each period with or without BGM on a 10-point scale (high or good was 10) regarding the following items: (1) satisfaction with work (Q1: How is your satisfaction with job?), (2) satisfaction with private life (Q2: How is your personal life satisfaction?), (3) general physical condition (Q3: How is your health condition?), (4) sleep status (Q4: How is your sleep condition?), (5) awareness of work performance (Q5: When you consider the performance level in your workplace to be 5 for the past 1–2 years, how much do you estimate your work performance over the last 1 month?), (6) comfort of the workplace (Q6: When you consider the comfort level in your workplace to be 5 for the past 1–2 years, how comfortable is your workplace in the last month?), and (7) communication with other staff members (Q7: When you consider the communication volume in your workplace to be 5 for the past 1–2 years, how is the communication volume in your workplace in the most recent month?). This information was obtained anonymously.
Statistical analysisWe investigated employees who had been enrolled for 8 months during the study period. Meanwhile, we excluded the employees who worked outside the office for 5 days or more or who had no overtime work; they were deemed unsuitable because of the difference in work style and/or in the labor contract.
In evaluating the overtime hours, (1) the difference between the mean overtime hours during periods I–IV was examined using the repeated general linear model (GLM) and multiple comparisons with Bonferroni correction. In addition, the mean overtime hours for 4 months during the overall 8-month study period were compared by paired t-test. (2) To consider the changes in workload by business status during the various periods and the seasonal issues in the business, we first calculated the correlation in overtime hours in the same month between the present year (observation) and the previous year by using the Pearson’s correlation coefficient (eTable 1). Considering that the overtime hours of the relevant month strongly correlated with those of the earlier month in comparison with those of the same month in the previous year, we then compared the overtime hours of each employee in periods II, III, and IV with those of period I. The ratios of overtime in periods II, III, and IV to that in period I were calculated as “overtime ratio to period I.” This ratio was used as a parametric value that would act as an indicator for investigating worktime changes. Then, the significant difference in changes in “overtime ratio to period I” was evaluated by the significance of the interaction effect between the observation period and the previous year using a repeated GLM.
The score of each questionnaire item was significantly associated with the variables age, sex, overtime hours, and offices (data not shown), which were all regarded as covariates. Using the analysis of covariance (ANCOVA), we compared the difference among periods and between the presence and absence of BGM at every four periods. Moreover, to determine the characteristics of each office, we defined 7 points or more as a high score and compared the ratio of 7 points or more in the period with or without BGM using the chi-square test.
Table 1 summarizes the total number of employees at each business site and their age and gender distributions. Age and sex distributions were significantly different between patterns A, B, and C. Pattern A had the most number of employees in their 30s, whereas patterns B and C had the most number of employees in their 40s. With regard to the male-to-female ratio, pattern B had the highest female ratio, while pattern C had the highest male ratio.
| Office Number | Group | n | Age, years | Male | ||||
|---|---|---|---|---|---|---|---|---|
| 20s | 30s | 40s | 50s | 60s | ||||
| 1 | A | 225 | 35.5% | 49.4% | 14.5% | 0.6% | 0.0% | 44.4% |
| 2 | A | 106 | 24.4% | 32.1% | 26.9% | 15.3% | 1.3% | 74.0% |
| 3 | A | 202 | 27.4% | 33.3% | 29.3% | 9.1% | 0.8% | 80.0% |
| 4 | A | 39 | 19.9% | 39.8% | 25.9% | 14.4% | 0.0% | 42.0% |
| 5 | A | 270 | 17.2% | 33.4% | 25.7% | 23.7% | 0.0% | 60.3% |
| Total | 842 | 25.6% | 37.8% | 23.8% | 12.5% | 0.3% | 61.7% | |
| 6 | B | 51 | 34.6% | 40.4% | 16.2% | 6.5% | 2.3% | 42.7% |
| 7 | B | 447 | 18.7% | 24.3% | 28.0% | 27.7% | 1.3% | 66.1% |
| 8 | B | 67 | 7.3% | 31.7% | 55.6% | 5.4% | 0.0% | 43.4% |
| 9 | B | 48 | 1.1% | 11.3% | 53.1% | 29.9% | 4.5% | 75.1% |
| 10 | B | 172 | 24.2% | 34.8% | 23.7% | 15.8% | 1.4% | 62.1% |
| Total | 785 | 18.9% | 27.5% | 30.2% | 22.0% | 1.5% | 58.1% | |
| 11 | C | 64 | 32.6% | 44.6% | 21.4% | 1.4% | 0.0% | 67.8% |
| 12 | C | 87 | 26.5% | 28.0% | 26.5% | 18.3% | 0.8% | 65.9% |
| 13 | C | 32 | 17.5% | 48.5% | 19.4% | 10.7% | 3.9% | 61.2% |
| 14 | C | 472 | 12.0% | 40.3% | 39.4% | 7.0% | 1.3% | 78.6% |
| 15 | C | 85 | 26.7% | 22.7% | 37.8% | 12.8% | 0.0% | 56.4% |
| Total | 740 | 17.5% | 37.7% | 35.4% | 8.7% | 1.1% | 64.3% | |
Table 2 shows the mean and standard deviation (SD) of overtime hours of all employees enrolled in this study. Working in five workplaces (offices) from five companies, 625, 536, and 577 employees in the observation (research) year and 490, 461, and 486 employees in the previous year were included, representing patterns A, B, and C, respectively. In the previous year, the overtime hours in offices 1, 4, 5, 9−12, 14, and 15 (pattern A: 3/5, pattern B: 2/5, pattern C: 4/5) were significantly different for each of the same periods; these offices could fluctuate with or without BGM intervention. In the observation year, offices 3−7, 9−14, and 15 (pattern A: 3/5, pattern B: 4/5, pattern C: 5/5) demonstrated a significant change in overtime hours.
| Number | type | n | Period I | Period II | Period III | Period IV | p-value | p-value of multiple comparisions | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Mean | SD | Mean | SD | Mean | SD | Mean | SD | I vs II | II vs III | III vs IV | ||||
| 1 | Observation year | 164 | 44.8 | 30.9 | 46.2 | 37.2 | 47.6 | 28.4 | 48.2 | 34.0 | .63 | .51 | .42 | .76 |
| Previous year | 130 | 41.2 | 29.1 | 46.8 | 29.2 | 43.5 | 27.3 | 40.7 | 26.3 | .03 | .01 | .80 | .18 | |
| 2 | Observation year | 87 | 21.6 | 12.5 | 19.4 | 10.9 | 19.2 | 10.8 | 20.1 | 11.5 | .19 | .06 | .87 | .42 |
| Previous year | 81 | 20.5 | 10.4 | 20.8 | 11.9 | 21.6 | 11.2 | 22.3 | 11.4 | .59 | .88 | .57 | .61 | |
| 3 | Observation year | 166 | 43.9 | 25.2 | 47.9 | 27.0 | 49.1 | 28.4 | 47.0 | 29.2 | .04 | .01 | .50 | .29 |
| Previous year | 150 | 44.1 | 25.8 | 44.5 | 28.4 | 47.0 | 27.5 | 46.5 | 27.5 | .36 | .88 | .18 | .81 | |
| 4 | Observation year | 37 | 59.4 | 28.7 | 58.2 | 33.2 | 66.8 | 42.6 | 49.8 | 47.0 | .01 | .72 | .03 | <.001 |
| Previous year | 28 | 51.1 | 30.7 | 48.2 | 28.6 | 64.6 | 38.9 | 68.3 | 34.2 | <.001 | .37 | .01 | .32 | |
| 5 | Observation year | 171 | 51.5 | 26.4 | 52.1 | 24.8 | 65.5 | 29.9 | 52.4 | 27.2 | <.001 | .65 | <.001 | <.001 |
| Previous year | 101 | ND | ND | 49.6 | 21.0 | 60.6 | 22.8 | 55.3 | 22.2 | <.001 | ND | <.001 | <.001 | |
| 6 | Observation year | 32 | 32.0 | 31.0 | 26.2 | 28.9 | 29.0 | 31.6 | 24.4 | 26.4 | .02 | .01 | .21 | .09 |
| Previous year | 27 | 34.8 | 32.7 | 32.7 | 40.2 | 36.1 | 40.0 | 27.9 | 27.9 | .10 | .65 | .14 | .02 | |
| 7 | Observation year | 285 | 44.8 | 30.6 | 46.5 | 29.7 | 53.9 | 32.3 | 47.0 | 31.3 | <.001 | .37 | <.001 | <.001 |
| Previous year | 209 | 52.8 | 33.2 | 52.9 | 33.4 | 54.9 | 31.9 | 56.4 | 39.4 | .30 | .93 | .34 | .46 | |
| 8 | Observation year | 67 | 48.9 | 24.0 | 50.1 | 27.9 | 51.9 | 32.0 | 47.2 | 32.4 | .75 | .80 | .61 | .32 |
| Previous year | 67 | 64.5 | 32.6 | 65.1 | 30.3 | 55.4 | 21.5 | 54.8 | 28.3 | .07 | .93 | .03 | .86 | |
| 9 | Observation year | 37 | 70.9 | 41.1 | 89.7 | 52.0 | 58.1 | 39.7 | 56.5 | 33.7 | <.001 | <.001 | <.001 | .66 |
| Previous year | 36 | 93.1 | 59.3 | 119.7 | 77.8 | 61.6 | 44.1 | 71.3 | 39.8 | <.001 | <.001 | <.001 | .04 | |
| 10 | Observation year | 147 | 66.0 | 30.6 | 81.1 | 35.6 | 65.9 | 33.1 | 69.8 | 32.3 | <.001 | <.001 | <.001 | .03 |
| Previous year | 122 | 75.7 | 34.2 | 75.9 | 33.0 | 72.2 | 31.5 | 81.2 | 36.8 | <.001 | .92 | .03 | <.001 | |
| 11 | Observation year | 48 | 60.5 | 27.4 | 46.4 | 26.4 | 53.0 | 32.2 | 56.2 | 30.4 | <.001 | <.001 | .01 | .37 |
| Previous year | 42 | 71.7 | 31.6 | 53.4 | 23.2 | 56.4 | 23.7 | 56.0 | 25.8 | .01 | <.001 | .19 | .90 | |
| 12 | Observation year | 74 | 39.0 | 21.2 | 34.7 | 17.6 | 35.1 | 19.4 | 40.7 | 25.1 | .02 | .05 | .81 | .01 |
| Previous year | 62 | 38.0 | 22.7 | 32.9 | 17.9 | 33.7 | 20.5 | 41.2 | 20.4 | <.001 | .03 | .73 | <.001 | |
| 13 | Observation year | 20 | 83.8 | 37.0 | 60.8 | 32.6 | 62.1 | 31.1 | 68.0 | 55.2 | .01 | <.001 | .82 | .41 |
| Previous year | 8 | 82.3 | 34.2 | 89.6 | 24.5 | 92.4 | 19.8 | 96.7 | 38.8 | .67 | .60 | .63 | .64 | |
| 14 | Observation year | 357 | 67.0 | 33.9 | 71.3 | 30.9 | 68.5 | 31.1 | 65.1 | 32.8 | <.001 | <.001 | .007 | <.001 |
| Previous year | 306 | 79.5 | 36.7 | 84.9 | 37.2 | 77.3 | 33.2 | 76.1 | 35.9 | <.001 | <.001 | .001 | .34 | |
| 15 | Observation year | 78 | 52.7 | 29.0 | 50.6 | 31.8 | 38.7 | 26.3 | 46.2 | 28.0 | <.001 | .42 | <.001 | .12 |
| Previous year | 66 | 65.2 | 34.6 | 57.7 | 28.5 | 65.3 | 28.5 | 65.7 | 27.4 | .04 | .02 | .04 | .88 | |
SD, standard deviation.
Among periods I–IV in the previous year, offices 1, 4, 5, 9−14 and 15 revealed significant differences in overtime, but by multiple comparisons, only office 9 revealed significantly different overtime. Similarly, among offices 3−7, 9−14, and 15 that were significantly different in the entire observation period, only office 10 in pattern B and office 14 in pattern C clearly revealed significant differences for the entire period by multiple comparisons for all periods. In offices 3−7, 9, 11−13, and 15, changes were only partly significant. The mean overtime hours for 4 months with or without BGM were significantly decreased in offices 5, 6, 9, 11, 13, and 14, with reduction rates of −4.3%, −19.2%, −10.7%, −11.8%, −16.8%, and −4.4%, respectively (Table 3). Thus, the significant reductions were observed on 1/5, 2/5, and 3/5 in patterns A, B, and C, respectively. Conversely, only office 10 in pattern B had significantly increased overtime when played with BGM compared with that without BGM (+13.9%).
| Office Number | Pattern | n | BGM (−) | BGM (+) | Reduction (%) | p-value | ||
|---|---|---|---|---|---|---|---|---|
| Mean | SD | Mean | SD | |||||
| 1 | A | 164 | 23.1 | 13.0 | 23.9 | 14.8 | 3.5% | .18 |
| 2 | A | 87 | 10.2 | 4.8 | 9.8 | 4.6 | −3.9% | .25 |
| 3 | A | 166 | 23.3 | 12.0 | 23.3 | 12.3 | 0.1% | .99 |
| 4 | A | 37 | 29.9 | 18.9 | 31.6 | 16.3 | 5.5% | .13 |
| 5 | A | 171 | 29.3 | 5.1 | 28.0 | 12.3 | −4.3% | <.001 |
| 6 | B | 32 | 15.3 | 15.3 | 12.3 | 13.2 | −19.2% | <.001 |
| 7 | B | 285 | 24.7 | 13.5 | 24.5 | 13.8 | −0.8% | .71 |
| 8 | B | 67 | 25.2 | 12.5 | 25.4 | 13.2 | 0.6% | .90 |
| 9 | B | 37 | 36.1 | 20.7 | 32.3 | 19.2 | −10.7% | <.001 |
| 10 | B | 147 | 33.0 | 14.9 | 37.6 | 15.6 | 13.9% | <.001 |
| 11 | C | 48 | 28.4 | 14.0 | 25.1 | 13.1 | −11.8% | <.001 |
| 12 | C | 74 | 18.5 | 9.2 | 18.4 | 9.0 | −0.7% | .82 |
| 13 | C | 20 | 36.5 | 16.2 | 30.4 | 15.9 | −16.8% | <.001 |
| 14 | C | 357 | 35.4 | 14.9 | 33.9 | 15.3 | −4.4% | <.001 |
| 15 | C | 78 | 22.6 | 13.3 | 22.9 | 12.5 | 1.3% | .68 |
BGM, background music; SD, standard deviation.
Figure 1 shows a comparison of changes in overtime hours in the observation (research) year and changes in overtime hours in the same period of the previous year. Furthermore, eTable 2 presents detailed data. Offices 4, 6, 13, 14, and 15 at 1/5, 1/5, and 3/5 offices in patterns A, B, and C, respectively, showed a significant interaction effect and reduced overtime hours. On GLM, office 2 in pattern A and office 8 in pattern B seemingly had an interaction by BGM, but no apparent significant difference was observed. The mean ratio between the observation year and the previous year was significantly different in offices 4, 13, 14, and 15 at 1/5, 0/5, and 3/5 offices in patterns A, B, and C, respectively (eTable 2).
Comparison of changes in overtime hours between observation year and same period of the previous year. Data are presented as means of the ratio to each overtime in the period I. Blue line represents the mean ratio of the observation year. Red line represents the mean ratio of the previous year. (A) pattern A, a)–e) represent offices 1–5 sequentially. (B) pattern B, a)–e) represent offices 6–10 sequentially. (C) pattern C, a)–e) represent offices 11–15 sequentially. P indicates the significance of the interaction effect between the variables of the period I–IV and with/without BGM.
Figure 2 illustrates the changes in the subjective survey during periods, and Table 4 presents the detailed data. Furthermore, eFigure 2 shows the percentage of high scores for each office. Any of the items shown in Figure 2 and Table 4 had no statistically significant differences; however, pattern B revealed almost no change, pattern A showed a decreasing tendency, and pattern C displayed an increasing tendency in the latter half period with BGM. As a result of examining the rate of showing a high score related to the presence or absence of BGM by offices, we divided the offices into two according to higher and lower scores, but no certain tendency was observed among all three patterns.
Changes in the subjective survey during periods. Self-administered questionnaire surveys were performed four times in the last month of each period with or without BGM on a 10-point scale (high or good was 10). Data are represented as the mean value. Blue, red, and gray lines represent the mean of five offices with the patterns A, B, and C.
| Period I | Period II | Period III | Period IV | p-value | BGM (−) | BGM (+) | p-value | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Pattern | Mean | 95% CI | Mean | 95% CI | Mean | 95% CI | Mean | 95% CI | Mean | 95% CI | Mean | 95% CI | |||||||||
| Satisfaction with work | A | 5.79 | 5.46 | 6.13 | 5.80 | 5.46 | 6.13 | 5.51 | 5.19 | 5.82 | 5.36 | 5.02 | 5.70 | .23 | 5.67 | 5.43 | 5.91 | 5.54 | 5.31 | 5.76 | .44 |
| B | 5.75 | 5.57 | 5.94 | 5.95 | 5.76 | 6.15 | 5.91 | 5.71 | 6.10 | 5.71 | 5.50 | 5.92 | .16 | 5.87 | 5.74 | 6.01 | 5.79 | 5.65 | 5.93 | .40 | |
| C | 5.75 | 5.46 | 6.04 | 5.64 | 5.34 | 5.93 | 5.44 | 5.13 | 5.74 | 5.76 | 5.46 | 6.05 | .63 | 5.84 | 5.63 | 6.06 | 5.47 | 5.28 | 5.66 | .00 | |
| Satisfaction with private | A | 5.79 | 5.46 | 6.13 | 5.80 | 5.46 | 6.13 | 5.51 | 5.19 | 5.82 | 5.36 | 5.02 | 5.70 | .05 | 6.12 | 5.91 | 6.34 | 6.07 | 5.87 | 6.27 | .70 |
| B | 5.75 | 5.57 | 5.94 | 5.95 | 5.76 | 6.15 | 5.91 | 5.71 | 6.10 | 5.71 | 5.50 | 5.92 | .74 | 6.64 | 6.49 | 6.78 | 6.47 | 6.32 | 6.61 | .10 | |
| C | 5.75 | 5.46 | 6.04 | 5.64 | 5.34 | 5.93 | 5.44 | 5.13 | 5.74 | 5.76 | 5.46 | 6.05 | .79 | 6.17 | 5.94 | 6.40 | 5.90 | 5.69 | 6.10 | .08 | |
| General physical condition | A | 5.51 | 5.16 | 5.85 | 5.56 | 5.22 | 5.90 | 5.25 | 4.93 | 5.57 | 5.22 | 4.86 | 5.57 | .29 | 5.46 | 5.21 | 5.70 | 5.29 | 5.06 | 5.52 | .33 |
| B | 6.02 | 5.82 | 6.21 | 6.10 | 5.89 | 6.31 | 6.00 | 5.79 | 6.20 | 5.89 | 5.67 | 6.11 | .64 | 6.07 | 5.92 | 6.22 | 5.94 | 5.79 | 6.08 | .21 | |
| C | 5.47 | 5.14 | 5.80 | 5.55 | 5.22 | 5.89 | 5.61 | 5.27 | 5.96 | 5.48 | 5.15 | 5.81 | .74 | 5.57 | 5.33 | 5.81 | 5.46 | 5.25 | 5.68 | .50 | |
| Sleep status | A | 5.41 | 5.07 | 5.76 | 5.48 | 5.14 | 5.82 | 5.12 | 4.80 | 5.44 | 4.98 | 4.63 | 5.33 | .26 | 5.26 | 5.02 | 5.51 | 5.22 | 4.99 | 5.45 | .80 |
| B | 5.67 | 5.47 | 5.86 | 5.95 | 5.74 | 6.16 | 5.92 | 5.71 | 6.12 | 5.71 | 5.49 | 5.94 | .16 | 5.86 | 5.71 | 6.01 | 5.75 | 5.61 | 5.90 | .32 | |
| C | 5.19 | 4.87 | 5.52 | 5.47 | 5.13 | 5.80 | 5.19 | 4.85 | 5.54 | 5.12 | 4.79 | 5.45 | .09 | 5.20 | 4.96 | 5.44 | 5.29 | 5.07 | 5.50 | .59 | |
| Work performance | A | 5.49 | 5.20 | 5.78 | 5.66 | 5.38 | 5.95 | 5.41 | 5.14 | 5.68 | 5.10 | 4.81 | 5.40 | .10 | 5.48 | 5.27 | 5.68 | 5.37 | 5.18 | 5.56 | .45 |
| B | 5.40 | 5.24 | 5.57 | 5.46 | 5.29 | 5.64 | 5.53 | 5.35 | 5.70 | 5.47 | 5.29 | 5.66 | .67 | 5.50 | 5.38 | 5.62 | 5.42 | 5.30 | 5.54 | .39 | |
| C | 5.34 | 5.09 | 5.59 | 5.41 | 5.16 | 5.67 | 5.30 | 5.04 | 5.57 | 5.32 | 5.07 | 5.57 | .95 | 5.42 | 5.24 | 5.61 | 5.28 | 5.12 | 5.45 | .26 | |
| Comfort of the workplace | A | 5.45 | 5.11 | 5.78 | 5.45 | 5.11 | 5.78 | 5.29 | 4.98 | 5.60 | 5.36 | 5.02 | 5.71 | .86 | 5.37 | 5.13 | 5.61 | 5.39 | 5.16 | 5.61 | .93 |
| B | 5.48 | 5.30 | 5.66 | 5.65 | 5.45 | 5.84 | 5.59 | 5.40 | 5.79 | 5.61 | 5.41 | 5.82 | .35 | 5.62 | 5.49 | 5.76 | 5.54 | 5.40 | 5.67 | .38 | |
| C | 5.60 | 5.32 | 5.89 | 5.39 | 5.10 | 5.68 | 5.15 | 4.85 | 5.45 | 5.50 | 5.21 | 5.79 | .27 | 5.48 | 5.27 | 5.69 | 5.39 | 5.20 | 5.58 | .54 | |
| Communication | A | 5.51 | 5.28 | 5.74 | 5.57 | 5.35 | 5.80 | 5.44 | 5.23 | 5.65 | 5.26 | 5.03 | 5.50 | .42 | 5.40 | 5.24 | 5.57 | 5.48 | 5.32 | 5.63 | .53 |
| B | 5.62 | 5.45 | 5.79 | 5.62 | 5.44 | 5.79 | 5.64 | 5.47 | 5.82 | 5.68 | 5.49 | 5.87 | .91 | 5.64 | 5.52 | 5.77 | 5.65 | 5.52 | 5.77 | .97 | |
| C | 5.57 | 5.32 | 5.82 | 5.56 | 5.30 | 5.82 | 5.31 | 5.04 | 5.58 | 5.42 | 5.16 | 5.67 | .34 | 5.46 | 5.27 | 5.64 | 5.45 | 5.28 | 5.62 | .96 | |
BGM, background music; CI, confidence interval.
As shown in eFigure 2, the BGM+ score in office 11 was significantly higher compared with silence for 6/7 items (satisfaction with work, general physical condition, sleep status, awareness of work performance, comfort of the workplace, and communication with other staff members). However, the percentage of high scores was clearly lower in the environment with BGM than silence in some offices (satisfaction with work, office 15; sleep status, office 12; awareness of work performance, office 15; communication with other staff members, office 6).
In this study, we investigated whether playing BGM in the workplace has any effects on reducing overtime hours according to the expected effectiveness in the following ways: (1) a conditioning effect that uses music to communicate “goodbye” to mark the end of work hours and (2) a natural awareness of the passage of time caused by changes in music. Regarding (1), patterns A and B were set up to notify the end of working hours by playing BGM at the end of working hours. For (2), to provide psychological motivation to work efficiently, BGM was played from 15:00 to the end of work in pattern B and for the whole day in pattern C.
To the best of our knowledge, this study is the first to evaluate the association between playing BGM and overtime hours other than using the regulation of setting a limit of working overtime as a company rule. However, evaluating the effects on changing overtime with only one index was difficult. One major factor is assumed to be the seasonal effect in the amount of work according to business type, and the second one is a change in the amount of work at that time by chance in a year. Generally, considering the seasonal factor, the increase or decrease of overtime hours in terms of economic evaluation is often compared within the same month or between periods of the previous year. Therefore, we first examined the correlations in overtime hours in the same month between the observation year and the previous year. Results showed that most of the offices strongly correlated with those in the earlier month in the observation year than those in the same month of the previous year. To minimize the impact of the second factor as a fortuitous change, we conducted alterations in the situation with and without playing BGM every 2 months and examined the changes in overtime hours. If the BGM had some effects on reducing overtime hours, the trend of overtime hours in the previous year would be different from that in the observation year. In addition, the trend of the overtime hours during the period with BGM would decrease, whereas that of the overtime hours during the period without BGM would increase.
Comprehensively, the positive effect was observed on 20%, 20%, and 60% of five offices in patterns A, B, and C, respectively. We found that chimes and/or music encourage “goodbye” to inform the end of working hours. However, the actual effect of these conditionings was observed in only one of the five offices. Rather, in three of the five offices, playing BGM during all working hours appeared to reduce overtime hours (pattern C), which suggests that music might be more effective in helping workers recognize the natural passage of time. However, we were unable to conduct any statistical analysis because of the small sample size.
We hypothesized that playing pleasant BGM could improve working performance, comfort, and communication21,22,23), resulting in a reduction in overtime. In fact, in the subjective survey, workers in one office (office 11) showed significantly positive results in almost all survey items. In addition, we observed a significant reduction in overtime in this office. However, among the other offices demonstrating an association between playing BGM and reduced overtime in pattern C, the subjective survey results revealed no significant effects of BGM on working performance, comfort, or communication.
There are several possible reasons why not all offices showed significant positive results. First, the subjective surveys were performed anonymously; thus, we could not analyze individual changes. In addition, a relatively low participation rate might have caused selection bias. Second, the employees might not have perceived the music played as comfortable, particularly with regard to the rhythm and volume25). In a previous study, the effects of music differed among introverted and extroverted individuals26), and listeners’ musical preferences influenced the effects of the music27). According to Lehmann and Seufert28), a part of working memory is used by BGM while playing BGM; therefore, individuals with a limited working memory are unable to learn while BGM is playing. A meta-analysis by Kämpfe et al.29) reported that music may have a positive impact on emotional responses because of its influence via mood and arousal, whereas music may cause a decrease in comprehension and attention because of right-brain activation. Thus, for some of the workers, BGM might be an uncomfortable experience that adversely affected their work environment. In this study, the BGM volume should have been set lower (6–10 dB) than the office noise in the place closest to the speakers. Thus, BGM should theoretically have been playing at a volume that is lower (6–10 dB) than office noise for all workers. In the offices, speakers were not arranged to have a uniform volume, and there was a difference in the BGM volume heard between workers who were working at the place closest to the speaker and those who were working far from it. Whether each different volume matched the worker’s tastes may have affected the results. These complexities indicate that the selected music and volume does not always suit the tastes of all employees. The expected results observed in office 11 might be because the content and volume of the music was more preferable and suitable for a large number of workers.
We cannot conclude that BGM in the office was effective in reducing overtime hours because there are many biases, such as patterns and work contents of each company, and the interaction is complicated. In this study, the type of business in all 15 offices was associated with information technology. The office members consisted of a large number of system engineers, and others were personnel or involved in labor and general affairs. The main characteristics of the business were data analysis, data maintenance, programming, system development, Internet support service, and data entry. In this study, among the C groups, the job types that showed a decline in overtime were mainly system engineers (offices 13 and 14) and Internet support service (office 11), We think that some types of system enginees spend most of time processing bugs of programming. Internet support service includes instruction on how to use PC and correspondence to malfunction. These might be relatively monotonous jobs. Moreover, the offices that appeared to show a decline in overtime had a relatively higher workload (offices 11, 13, and 14; mean 70.4 hr/month in period I) compared with other offices (offices 12 and 15; mean 45.8 hr/month in period I). From these results, we speculated that these biases may be the reason why BGM appeared to work or BGM is suitable for jobs that are monotonous and have long working time.
With regard to the study’s strength, this study was well designed to evaluate the effect on overtime hours by changing the schedule and using different patterns of playing BGM. Details on overtime hours from individual employees were obtained from both the observation year and the previous year. In addition, the patterns of BGM were arbitrarily assigned to reduce bias. Historical comparison and alteration were also investigated, but not parallel comparison because of large variations, such as differences in the attributes and systems of each office.
In contrast, this study has several limitations. First, the influence of social impact on working hours is unavoidable, and consequently, the effectiveness of BGM might be under- or overestimated. Second, we could not evaluate the statistical significance because of the small sample size among the three groups. Third, participating in business offices that had a high awareness of managing working hours may yield some bias. Social pressure might make the overtime reduction effect difficult to achieve in the observation year compared with that in the previous year because of factors other than BGM, such as the revision of Labor Standards Law for overtime reduction. Fourth, the period of 2 months with or without playing BGM was considerably short to get accustomed to BGM, leading to discomfort because employees remained unaccustomed to BGM. Employees in these offices may have conducted the subjective survey based on the knowledge that they were participating in this study. Long-term effects of BGM on overtime should also be evaluated. Despite having several limitations, this pilot study suggests that playing BGM throughout the entire day in the workplace is associated with less work overtime in some offices.
In this pilot study, there were offices that showed less overtime among employees when BGM was played, particularly throughout the entire day, than when BGM was not played. The possible association between playing BGM throughout the entire day and less overtime might be seen in some suitable situations of BGM and working conditions such as business type, content, and styles. Further detailed, large-scale, and long-term empirical studies seem to be worth conducting to investigate the effects of playing BGM on reducing overtime, stratified with BGM situations and working conditions.
We would like to acknowledge all the study staff for their commitment to data collection. This study received financial and technical support by USEN Corporation (a subsidiary of USEN-NEXT HOLDINGS) to use the BGM “Sound Design for OFFICE.” The funder did not play a role in the design of the study, collection, analysis, and interpretation of data, or in writing the manuscript.
This study was approved by the Institutional Review Board for Clinical Research at Tokai University School of Medicine (IRB-Tokai) (Protocol Number 16R-278).
Written informed consent was obtained from the representative of each office, according to the instruction of the IRB-Tokai and based on Ethical Guidelines for Medical and Health Research Involving Human Subjects published by the Japan Minister of Health, Labour and Welfare. Overtime data were used secondarily. Regarding data from questionnaires, individual consent was considered to be obtained by providing responses to questions.
Trial Registration: UMIN-CTR, UMIN000038158. Retrospectively registered on October 1, 2019. https://upload.umin.ac.jp/cgi-open-bin/ctr_e/ctr_view.cgi?recptno=R000043482
MT received financial support from USEN Corporation (Tokyo, Japan). Other authors declare no competing interests.
The datasets generated and/or analyzed during the current study are not publicly available because of the applicable restrictions to the availability of these data, which were used under license for the current study. However, the data are available from the corresponding author on reasonable request.
MT and HF conceived and designed the study, performed the research, and collected the data. YF, KF, and MT analyzed the data. YF and MT wrote the manuscript. HF and KF supervised the project and provided critical comments. All authors reviewed the manuscript.
Study methods (A) and schedule (B). (A) Three patterns of background music (BGM) playing in the office. Pattern A (offices 1–5): BGM was played for 5 min and a female announced the same sentence every day. Pattern B (offices 6–10): BGM was played from 15:00 to the end of the day. Pattern C (offices 11–15): BGM selected by each company was played from the start of work to the end. (B) The period with and without BGM (silence) was set to 2 months, and the periods were repeated every 2 months alternately for 8 months.
| Observation year | Observation year vs. Previous year | |||||||
|---|---|---|---|---|---|---|---|---|
| Number | n | Period I vs.II | Period II vs.III | Period III vs.IV | Peirod I | Period II | Period III | Period IV |
| 1 | 81 | .62 | .70 | .68 | .32 | .34 | .49 | .42 |
| 2 | 150 | .56 | .38 | .55 | .45 | .28 | .36 | .32 |
| 3 | 28 | .73 | .66 | .60 | .38 | .43 | .35 | .51 |
| 4 | 13 | .79 | .82 | .87 | .35 | .31 | .56 | .44 |
| 5 | 101 | .70 | .72 | .75 | .59 | .49 | .50 | .50 |
| 6 | 27 | .91 | .92 | .89 | .78 | .83 | .86 | .80 |
| 7 | 209 | .47 | .54 | .62 | .28 | .32 | .29 | .31 |
| 8 | 21 | .51 | .84 | .60 | .38 | .80 | .85 | .49 |
| 9 | 36 | .85 | .82 | .83 | .77 | .82 | .78 | .81 |
| 10 | 122 | .82 | .81 | .80 | .63 | .57 | .64 | .67 |
| 11 | 42 | .84 | .80 | .72 | .53 | .73 | .67 | .58 |
| 12 | 62 | .59 | .71 | .69 | .48 | .22 | .35 | .42 |
| 13 | 8 | .59 | .62 | .32 | .59 | .68 | −.05 | .36 |
| 14 | 306 | .79 | .80 | .85 | .60 | .55 | .20 | .24 |
| 15 | 66 | .70 | .81 | .78 | .62 | .43 | .33 | .44 |
| Period I | Period II | Period III | Period IV | Total | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| No | n | Mean | 95% CI | Mean | 95% CI | Mean | 95% CI | Mean | 95% CI | p | ||||||
| 1 | Observation year | 108 | 1 | 1.23 | 1.06 | 1.41 | 1.29 | 1.09 | 1.48 | 1.39 | 1.17 | 1.60 | 1.23 | 1.10 | 1.36 | |
| Previous year | 108 | 1 | 1.39 | 1.22 | 1.57 | 1.35 | 1.16 | 1.55 | 1.22 | 1.00 | 1.43 | 1.24 | 1.11 | 1.37 | .90 | |
| 2 | Observation year | 58 | 1 | .92 | .73 | 1.11 | .93 | .74 | 1.12 | 1.03 | .84 | 1.22 | .97 | .85 | 1.09 | |
| Previous year | 58 | 1 | 1.21 | 1.02 | 1.40 | 1.14 | .95 | 1.33 | 1.20 | 1.01 | 1.40 | 1.14 | 1.02 | 1.26 | .05 | |
| 3 | Observation year | 139 | 1 | 1.26 | 1.12 | 1.40 | 1.25 | 1.06 | 1.44 | 1.21 | 1.05 | 1.36 | 1.18 | 1.08 | 1.28 | |
| Previous year | 139 | 1 | 1.05 | .91 | 1.19 | 1.33 | 1.15 | 1.52 | 1.20 | 1.04 | 1.35 | 1.14 | 1.04 | 1.25 | .64 | |
| 4 | Observation year | 25 | 1 | .90 | .78 | 1.02 | 1.09 | .82 | 1.35 | .78 | .53 | 1.04 | .94 | .81 | 1.08 | |
| Previous year | 25 | 1 | .94 | .81 | 1.06 | 1.29 | 1.03 | 1.56 | 1.42 | 1.17 | 1.68 | 1.16 | 1.03 | 1.30 | .02 | |
| 5 | Observation year | 86 | 1 | 1.08 | 1.01 | 1.15 | 1.39 | .83 | 1.94 | 1.07 | .60 | 1.54 | 1.13 | .91 | 1.36 | |
| Previous year | 91 | 1 | 1.00 | .92 | 1.08 | .94 | .35 | 1.52 | .84 | .34 | 1.34 | .94 | .70 | 1.19 | .26 | |
| 6 | Observation year | 31 | 1 | .94 | .65 | 1.24 | .89 | .48 | 1.29 | 1.01 | .61 | 1.41 | .96 | .76 | 1.16 | |
| Previous year | 31 | 1 | .94 | .64 | 1.24 | 1.28 | .87 | 1.69 | 1.12 | .72 | 1.53 | 1.09 | .88 | 1.29 | .38 | |
| 7 | Observation year | 188 | 1 | 1.16 | 1.05 | 1.26 | 1.33 | 1.22 | 1.44 | 1.22 | 1.08 | 1.36 | 1.18 | 1.11 | 1.25 | |
| Previous year | 188 | 1 | 1.14 | 1.04 | 1.25 | 1.25 | 1.14 | 1.35 | 1.28 | 1.14 | 1.43 | 1.17 | 1.10 | 1.24 | .87 | |
| 8 | Observation year | 21 | 1 | .93 | .76 | 1.10 | .98 | .80 | 1.16 | .79 | .62 | .96 | .92 | .82 | 1.03 | |
| Previous year | 21 | 1 | 1.09 | .92 | 1.26 | .95 | .77 | 1.13 | .90 | .73 | 1.07 | .98 | .88 | 1.09 | .44 | |
| 9 | Observation year | 34 | 1 | 1.36 | 1.10 | 1.62 | .82 | .70 | .95 | .85 | .69 | 1.01 | 1.01 | .89 | 1.12 | |
| Previous year | 34 | 1 | 1.44 | 1.18 | 1.70 | .75 | .62 | .88 | .89 | .73 | 1.05 | 1.02 | .90 | 1.14 | .88 | |
| 10 | Observation year | 122 | 1 | 1.31 | 1.22 | 1.40 | 1.06 | .98 | 1.14 | 1.17 | 1.07 | 1.26 | 1.13 | 1.08 | 1.19 | |
| Previous year | 122 | 1 | 1.07 | .98 | 1.16 | 1.02 | .94 | 1.09 | 1.15 | 1.06 | 1.25 | 1.06 | 1.00 | 1.12 | .07 | |
| 11 | Observation year | 42 | 1 | .77 | .67 | .87 | .95 | .74 | 1.15 | 1.03 | .82 | 1.23 | .94 | .83 | 1.05 | |
| Previous year | 42 | 1 | .81 | .71 | .91 | .92 | .71 | 1.12 | .93 | .72 | 1.14 | .91 | .80 | 1.02 | .78 | |
| 12 | Observation year | 297 | 1 | .91 | .76 | 1.06 | .98 | .80 | 1.16 | 1.11 | .94 | 1.28 | 1.00 | .89 | 1.11 | |
| Previous year | 297 | 1 | .94 | .79 | 1.09 | .98 | .80 | 1.17 | 1.19 | 1.02 | 1.36 | 1.03 | .92 | 1.14 | .72 | |
| 13 | Observation year | 8 | 1 | .66 | .23 | 1.09 | .77 | .42 | 1.13 | .78 | .40 | 1.15 | .80 | .55 | 1.05 | |
| Previous year | 8 | 1 | 1.28 | .85 | 1.71 | 1.31 | .96 | 1.66 | 1.31 | .94 | 1.68 | 1.23 | .98 | 1.47 | .02 | |
| 14 | Observation year | 297 | 1 | 1.13 | .99 | 1.27 | 1.10 | .75 | 1.44 | 1.04 | .71 | 1.36 | 1.07 | .87 | 1.26 | |
| Previous year | 306 | 1 | 1.25 | 1.11 | 1.38 | 1.58 | 1.24 | 1.91 | 1.54 | 1.22 | 1.86 | 1.34 | 1.15 | 1.53 | .05 | |
| 15 | Observation year | 63 | 1 | .97 | .86 | 1.07 | .76 | .63 | .89 | .98 | .82 | 1.13 | .93 | .84 | 1.01 | |
| Previous year | 63 | 1 | .97 | .86 | 1.08 | 1.12 | 1.00 | 1.25 | 1.12 | .96 | 1.27 | 1.05 | .97 | 1.13 | .04 | |
CI, confidence interval.
Results of subjective survey by office between the presence and absence of background music. Of 10 points, 7 points or more was estimated as a high score, and the ratio of 7 points or more in the period with or without background music was compared using the chi-square test. Asterisk (*) means the significant difference (p<0.05).
