Subjective assessments of interference during cognitive tasks in noisy and silent working conditions Helga Sukowski Bundesanstalt für Arbeitsschutz und Arbeitsmedizin (BAuA) 44149 Dortmund Germany

ABSTRACT In a research project on effects of noise on cognitive performance and wellbeing at work, several scales for subjective assessments with respect to the task execution are applied. The aim is to gain insight into how participants experience task processing in different background sound conditions. In the laboratory, participants worked on a proofreading task and a concentration task twice within an interval of around one week, once in a silent condition and once in one of various conditions with a background sound. In both sessions, participants were asked after each task to assess four aspects with respect to task processing: (1) experienced effort, (2) concentration, (3) performance and (4) disturbance by the acoustical condition. The scales range from 0 to 100. So far, one background sound (office noise) was applied. First, data were analysed regarding the differences between the assessments in the condition with and without background sound, separately for the two performance tasks. For both tasks, participants experienced significantly more effort, lower concentration and higher disturbance in the condition with background sound than in the silent condition. Once all sound conditions are completed, an analysis of the comparison between the groups with different background sound conditions will be conducted.

1. INTRODUCTION

In a research project on effects of noise on cognitive performance and wellbeing in employees, in total two experimental studies are carried out. In the first study participants worked on a proofreading task either twice in a silent condition or twice in a condition with speech as background sound. The aim of that study was to investigate, whether there is a training effect from one run to the other and to collect information on the task and the entire test procedure [1]. The aim of the second study is to investigate effects of different background sounds on the proofreading task and a concentration task. In both studies, in addition to the task execution itself, participants were asked about how they expe- rienced the task processing in the different acoustical conditions using four predefined questions. At first glance, it may seem as if the application of these questions is simply an addendum to the tests on cognitive performance, but it is actually a deliberate approach motivated by specific reasons. First, we are generally interested in how participants experience the situations when they work on cognitive tasks in a condition with background sound, in comparison to a situation that is free from any un- wanted sound. The legal demand that adverse effects on employees caused by factors in the work environment have to be minimized is one aspect of “safety and health at work” and therefore espe- cially meant to provide optimal conditions to avoid detrimental effects on health and wellbeing. Since

i, orn inter.noise 21-24 AUGUST SCOTTISH EVENT CAMPUS ? O? ? GLASGOW

unfavorable acoustical conditions are often assessed as a particularly adverse factor in the work en- vironment (e.g. [2]), possible effects of this factor should be considered with more than one method. A further reason is based on the knowledge that task performance and subjective assessments can be different with respect to noise effects. Especially in studies where the cognitive task takes only a short time it might happen that participants can compensate the noise disturbance by higher effort, with the outcome that no effect on the performance is observed, but nevertheless the participants might feel disturbed, interrupted or annoyed by the sound. In these cases it is very important to collect also information on subjective assessments, in order to avoid misleading interpretations of noise effects. Finally, these studies using performance measures and subjective assessments may help to overcome the line between the preferred use of particular methods in field studies and basic research studies, as described in a review on effects of noise on cognitive performance by Schlittmeier & Marsh [3]. Since the research of the BAuA (Federal Institute for Occupational Safety and Health) is focused on occupational settings, further intentions in this project are to carry out the study with employees cov- ering a large age range (intended 30 to 60 years) in contrast to studies with young students, to use a task that is to some extent related to real tasks at a workplace, and as far as possible to use sounds that are realistic for working situations. So far, only one background sound condition could be applied (“sound 1”). Accordingly, the current test set is called “test condition 1”. All information in the following sections refer to the subjective assessments in this test condition. More background sounds are to be applied at a later stage in this ongoing study. 2. METHODS

2.1. Participants So far, 23 employees of the Federal Institute for Occupational Safety and Health participated in “test condition 1”. Results from persons who participated only once or only in one of the two tasks are not included in the analysis. For the reading task, 21 participants and for the concentration task, 19 par- ticipants were included. Referring to the number of 21 persons, there were 8 male and 13 female participants. The age of the participants ranged from 29 to 61 (mean = 44 years; median = 41 years), which is very close to the intended range. 2.2. Cognitive tasks In this study, two different cognitive tasks were applied, a proofreading task and a concentration task. Proofreading task: The reading task was developed by the author. The idea and the development of that reading task are explained in detail in previous publications [1; 4]. In short, the main task for the participants is to find mistakes in written sentences. The task exists in two test versions A and B. Each version consists of 52 items with one or two sentences of different length and complexity. The items are either correct or they contain one mistake. The items appear one by one on a computer screen and participants are asked to mark the word on the touch screen that is or that leads to a mistake in the respective item. There is a time limit to process each item. Concentration task: As a concentration task the “Konzentrations-Leistungstest - Revidierte Fassung - (KLT-R)” [5] was applied. This test consists of many small arithmetic problems (additions, subtrac- tions). To solve the problems it is necessary to keep in mind interim results and also the instruction on how to proceed depending on the single results. This demand of the task is further increased by the added time pressure. The test was applied in a paper & pencil format that exists in two test versions (A and B), with 9 columns of 20 arithmetic problems each.

i, orn inter.noise 21-24 AUGUST SCOTTISH EVENT CAMPUS ? O? ? GLASGOW

2.3. Acoustical conditions In this part of the study two acoustical conditions were realized: a silent condition (“silence”) and a condition with office noise as background sound (“sound 1”). The office sound is a recording of a situation in a real multi-space office (27.2 x 13.2 x 3.0 m³) with 30 workplaces. The office includes desks equipped for four employees and also specific rooms (retreat area, single offices, conference room). Room acoustical measures were already in place in that office. The sound recording for this study is provided by the “Institute for Occupational Safety and Health of the German Social Accident Insurance” (IFA). For detailed information on the sound see [6]. The recording is a binaural recording carried out with an artificial head (HEAD acoustics HSU III.2 + HEAD acoustics SQobold). The artificial head was placed at a vacant desk. The occupancy rate during the recording time (123 min) varied between 30 to 56% in that office. The original recording was shortened to the length of 25:52 min, using mainly busy situations in the recording. The energetic mean sound pressure level for both microphones of the artificial head was L Aeq = 46.2 dB(A). To compensate for the transfer function of the reproduction chain, an equalization for realistic listening was carried out. This recording was presented to the participants using closed supra-aural headphones (Sennheiser HD 25). All participants wore the headphones while working on the two tasks and also while answer- ing the questions on the scales. The office noise was only present in the sound condition during the cognitive tasks, but not during the assessment tasks (scales) following the cognitive tasks. The office noise was presented with an Leq of 50.4 dB(A) (left channel) and 47.4 dB(A) (right channel). 2.4. Laboratory and study design The study was carried out in a large, sound insulated laboratory (L: 8.4 m; W: 6.4 m; H: 4.2 m). The laboratory was actually prepared for testing two persons at the same time and therefore equipped with two workplaces separated by a large, non-transparent screen. However, due to the current Covid-19 pandemic situation only one person participated at a time. Each person participated twice. The time between the first test session (T1) and the second test session (T2) varied between one week and 10 days, with only very few exceptions. The participants worked on a different test version (A or B; sequence was balanced) of the developed reading task and the concentration task in each session. Regarding the acoustical condition there were the following sequences: Reading task: T1 silence, T2 sound 1 (10 participants); T1 sound 1, T2 silence (11 participants). Concentration task: T1 silence, T2 sound 1 (9 participants); T1 sound 1, T2 silence (10 participants). 2.5. The four questions and scales for subjective assessments The sections to assess the task execution started immediately after each task. The participants still wore the headphones, but there was no sound presentation in either condition. The use of the scales was explained once during the general instructions and then again using a separate slide on the com- puter screen. After reading the instruction, the participants triggered the start of the questions them- selves. Then the questions appeared one by one, each together with the respective scale, and for all participants and in both tasks in the same order. The scales were designed like small rulers and ranged from 0 to 100. Each scale was labelled at both poles only (for questions and labels, see Table 1). The participants could give their assessments by marking a point on the scale with the touch pen on the touch screen. There was no time limit and participants could change their assessments until they decided to move on to the next question.

i, orn inter.noise 21-24 AUGUST SCOTTISH EVENT CAMPUS ? O? ? GLASGOW

Table 1: Questions and labels used in the assessment section after the reading task (translated from German).

Aspect asked for Short name Question Label at 0 Label at 100

Experienced Effort Effort How much effort did you experi- ence while working on the task? no effort very high effort

Experienced Concentration Concentration How well could you concentrate during working on the task?

I could not con- centrate at all.

I could concen- trate very well.

I did not work correctly on any item.

Self-assessed Performance Performance How do you assess your perfor- mance in this task?

I worked correctly on all items.

How disturbing did you experi- ence the acoustical condition while working on the task?

Experienced Disturbance Disturbance

not at all disturbing very disturbing

2.6. Test sessions In Table 2, all elements of the test sessions are listed. Except for (2) and (3), both sessions (T1 and T2) covered the same elements and the sections were conducted in the same order. Table 2: Sequence of different elements in the test sessions. The sections for gathering the subjec- tive assessments are marked in light red.

Sequence of different sections in each test session

1 Welcome

2 General information and informed consent

3 Short questionnaire

4 Questions on mood and tiredness

Reading task

• Verbal and written instruction • Training without time limit • Training with time limit • Reading task itself (10-14 min)

5

6 Subjective assessments regarding the task execution

7 Questions on mood and tiredness

Break (about 10 min)

Concentration task

• Verbal and written instruction • Training with feedback • Concentration task itself (18 min)

8

9 Subjective assessments regarding the task execution

10 Questions on mood and tiredness

11 Time for questions and feedback

12 Farewell

i, orn inter.noise 21-24 AUGUST SCOTTISH EVENT CAMPUS ? O? ? GLASGOW

3. RESULTS

Since at this point only results from one sound condition (test condition 1) are available, the focus in the results section is on the comparison of the differences between the assessments in the silent con- dition and the condition with the office noise as background sound, and also on the comparison be- tween the two different tasks. 3.1. Comparisons of means in silence and noise for each scale and each task In order to investigate whether there are differences between the assessments in the silent condition and the condition with office noise as background sound, t -tests for paired comparisons were calcu- lated separately for each scale and for each cognitive task. The results for the reading task are illus- trated in Figure 1 and those for the concentration task in Figure 2. The respective mean values and the results of the statistical analysis are summarized in Tables 3 and 4. Reading task

i, orn inter.noise 21-24 AUGUST SCOTTISH EVENT CAMPUS ? O? ? GLASGOW

Figure 1: Illustration of mean values for all subjective assessments referring to the reading task, de- picted separately for the acoustical conditions (silence, sound 1). For exact values, see Table 3. Significant differences ( p < 0.05) are marked with an asterisk.

Table 3: Mean values for subjective assessments referring to the reading task in both acoustical con- ditions, and results from the analysis with t -tests for paired comparisons.

Mean values

Aspect asked for

T df p Silence Sound 1

Effort 37.33 50.43 -2.800 20 0.011

Concentration 83.38 55.38 - 5.172 20 <0.01

Performance 69.00 63.14 - 1.350 20 0.192

Disturbance 2.90 56.33 -8.566 20 <0.01

100 - 90 OSilence Sound 1 70 | 60 + * 50 + 30 + 20 + 10 + ¥ Effort Concentration Performance Disturbance

Table 3 and Figure 1 show significant differences between the mean values in the silent condition and the sound condition for the scales “effort”, “concentration” and “disturbance”. The mean values assessing the task execution indicate higher experienced effort, lower experienced concentration and higher experienced disturbance in the sound condition than in the silent condition. For the scale “self- assessed performance”, no significant difference was observed. Concentration task

i, orn inter.noise 21-24 AUGUST SCOTTISH EVENT CAMPUS ? O? ? GLASGOW

Figure 2: Illustration of mean values for all subjective assessments referring to the concentration task, depicted separately for the acoustical conditions (silence, sound 1). For exact values, see Table 4. Significant differences ( p < 0.05) are marked with an asterisk.

Table 4: Mean values for subjective assessments referring to the concentration task in both acoustical conditions, and results from the analysis with t -tests for paired comparisons.

Mean values

Aspect asked for

T df p Silence Sound 1

Effort 57.21 72.74 -2.554 18 0.020

Concentration 75.11 45.32 - 4.742 18 <0.01

Performance 54.42 46.05 - 1.810 18 0.087

Disturbance 5.47 68.16 -10.981 18 <0.01

Regarding the differences between both acoustical conditions Table 4 and Figure 2 show the same results for the concentration task as for the reading task. Again, there were significant differences between the mean values on the questions referring to “effort”, “concentration” and “disturbance”, with higher experienced effort, lower experienced concentration and higher experienced disturbance in the sound condition than in the silent condition. For the scale “self-assessed performance” there was also no significant difference observed.

100 - 90 80 70 | 60; _* 50 | * @Silence Sound 1 30 + 20 10 + * Effort Concentration Performance Disturbance

3.2. Comparisons of means in both tasks for each scale and each acoustical condition In order to investigate whether one of the tasks leads to significantly higher or lower amounts in the assessments, again t -tests for paired comparisons were calculated for each question on subjective assessments. Now, the comparison was between the assessments for processing the reading task and the assessments for processing the concentration task. The four analyses carried out for the silent condition revealed significant differences between the assessments observed for both tasks regarding the questions on “effort” [ t (18) = -4.411, p < 0.05] and “performance” [ t (18) = 3.335, p < 0.05]. The participants experienced significantly higher effort (m eans: M Reading = 37; M Concentration = 57.2) and rated there performance lower ( M Reading = 70,95; M Concentration = 54.4) in the concentration task than in the reading task (n = 19). A similar result was found for the condition with background sound. Again, the analyses showed significant results for the questions on “effort” [ t (18) = -4.595, p < 0.05] and “performance” [ t (18) = 2.408, p < 0.05], with higher effort ( M Reading = 48.37; M Concentration = 72.74) and lower self-assessed performance ( M Reading = 64.53; M Concentration = 46.05) in the concentration task than in the reading task (n = 19). 4. DISCUSSION AND CONCLUSIONS

At this point of the study, subjective assessments referring to situations where cognitive tasks had to be carried out in silence or in noise were analysed. The results showed clear findings for the test condition 1, reported here. In both tasks the participants experienced significantly more effort, less concentration and more disturbance when they worked on the tasks in a condition with office noise as background sound in comparison to a silent condition. At first glance, this finding can be considered as not surprising or even expected. The results show detrimental effects on the assessments when the participants worked in noisy conditions. To feel af- fected by noise during a cognitive task is an experience we all know more or less from everyday life. However, looking closer at the test conditions, it is not that obvious to observe these clear effects as it seems. The sound condition and the test condition were moderate in comparison to a real working day. The time working on the tasks was short (10 to 18 min) in comparison to an entire working day, and the sound with an Leq around 50 dB(A) was comparatively soft. For example, in Germany the Technical Rules for Workplaces “Noise” (ASR A3.7) [7] state that during activities at workplaces with a high demand on concentration or speech intelligibility, a rating level of 55 dB(A) must not be exceeded (rating level: LpAeq during a particular activity plus supplements (max. 6 dB) in case of impulsiveness and/or tonality and/or informational contents). The current findings demonstrate: Even in situations with comparatively short periods of high concentration and a moderate sound level, per- sons experience detrimental effects in comparison to situations without any background sound. The analyses carried out for the comparisons between the two tasks showed similar results in both acoustical conditions. Although the level of experienced effort was higher for both tasks when carried out in a condition with background sound, the difference between the tasks was similar in both acous- tical conditions. From this finding, it can be concluded that the concentration test required in general more effort than the reading task. The higher effort experienced in the concentration task seems to be linked to the task itself rather than to the sound condition. Since it is known that tasks with higher demands can lead to a higher amount of experienced effort or annoyance when carried out in noisy conditions than less demanding tasks, this observation possibly will not hold for further sound con- ditions to be applied in this study. Regarding the subjective assessments, the other “side” of this observation is, that the reading task - although potentially less demanding than the concentration task - is equally affected by the sound condition in comparison to the silent condition.

i, orn inter.noise 21-24 AUGUST SCOTTISH EVENT CAMPUS ? O? ? GLASGOW

The results presented in this contribution are in line with findings from a previous study on the reading task [8], where differences between two different groups were found. In that study the participants from a group working on the reading task twice in a speech sound condition experienced significantly more effort and more disturbance than the participants working twice in silence, and they assessed their concentration and performance lower than the participants in a group working on the reading task twice in silence. Irrespective of the results achieved in the cognitive tasks itself (performance), the results from the previous and the current study underline the necessity to protect employees from unwanted sounds at the workplace, in order to minimize disturbance or effort and to allow concentrated working. An eye should also be kept on effective room acoustic measures and noise reduction measures even when the background sound does not exceed the permitted limits. According to the principles of occupational safety and health, it can be summarized: Reduce noise at the workplace by as much as possible! 5. ACKNOWLEDGEMENTS

Many thanks to my BAuA colleagues for participating in this study. Helpful support was also given by colleagues in the laboratory unit, mainly by Ulrich Hold, who did the programming of the com- puter-based version of the reading task and the implementation of the scales. Colleagues of the BAuA information center provided support during the scheduling of the participants. Nina Ahrweiler did a great job with testing participants and working on data entry and proofreading of the manuscript. Many thanks to Jan Selzer and Florian Schelle from the IFA for providing the realistic sound material, for installing the sound in our laboratory and the entire helpful cooperation. 6. REFERENCES

1. Sukowski, H. Effects of the acoustical work environment on reading performance in employees:

A laboratory study on the evaluation of a reading task. Proceedings of the 26th International Congress on Sound and Vibration ; pp. 689-705. Montreal, Canada, Jul 7-11 2019. 2. Kim, J. & de Dear, R. Workspace satisfaction: The privacy-communication trade-off in open-

plan offices. J. Environ Psychol. 36 , 18-26 (2013). 3. Schlittmeier, S. & Marsh, J.E. Review of research on the effects of noise on cognitive perfor-

mance 2017-2021. Proceedings of 13 th ICBEN Congress on Noise as a Public Health Problem (virtual conference), paper 28062 [48 pages]. Stockholm, 14-17 Jun 2021. 4. Sukowski, H. & Romanus, E. Effects of background speech on reading performance in adults.

Proc Mtgs Acoust . 28(1) , article 050002 [8 pages] (2017). 5. Düker, H. & Lienert, G.A., revised edition: Lukesch, H. & Mayrhofer, S. Konzentrations-Leis-

tungs-Test - Revidierte Fassung , 1st edition, Hogrefe, 2001. 6. Selzer, J. & Schelle, F. Untersuchung der Raumakustik und auditiver Belastung im Mehrperso-

nenbüro. Fortschritte der Akustik - DAGA 2021 , pp. 601-604. Berlin: DEGA e.V. (2021). 7. Technische Regeln für Arbeitsstätten. Lärm (ASR A3.7). Gemeinsames Ministerialblatt, 2021

März; 24:543-557. 8. Sukowski, H. Subjective assessments of interference by noise during a proofreading task. Ab-

stracts of 13 th ICBEN Congress on Noise as a Public Health Problem (virtual conference), ab- stract 28639, p. 79. Stockholm, 14-17 Jun 2021.

i, orn inter.noise 21-24 AUGUST SCOTTISH EVENT CAMPUS ? O? ? GLASGOW