Investigating the positive effects of music on attentional
control using EEG and an attentional task
Abstract
Many people experience problems efficiently controlling their attention and could benefit greatly from new ways to enhance their attention. The proposed study aims to investigate the effects of music on attentional control as measured in EEG and an attentional task. The proposed study is based on a pilot study which investigated the effects of elevator music on EEG and cognitive performance. No significant effects of elevator music on EEG or performance were found in the pilot study, but there were several limitations in the experimental design which presumably affected the results. The proposed study excludes these limitations and investigates the effects of classical music instead of elevator music on attentional control, since classical music has previously been reported to positively affect cognitive performance. The study combines EEG recordings and an attentional task, being the Attentional Network Task (ANT), in both a music condition and a control condition to investigate the effects of classical music on EEG correlates that represent attentional and on ANT scores that
represent the efficiencies of the alerting, orienting, and executive control networks. Furthermore, questionnaires for self-assessment of trait attentional control and trait anxiety are included to
investigate whether the effects of music on attentional control differ for subjects with different traits.
Summary
The proposed study aims to investigate the effects of classical music on attentional control and how these effects differ for people with different traits including trait attentional control and trait anxiety. Questionnaires are used for self-assessment of traits, and attentional control is assessed using
electroencephalography (EEG) correlates and performance in an attentional task. If music is found to positively affect attentional control, further research can be conducted to develop new attention-enhancing therapies based on music. Such music-based therapies could serve as drug-free methods to enhance attentional control in for example ADHD patients. Also, attentional control in study and work settings could be improved by using background music.
Keywords
Attentional Control – Elevator music – Classical music – Electroencephalography (EEG) – Attentional Network Task (ANT) – Trait attentional control – Trait anxiety
Research topic
From childhood until adult life, people often find themselves in situations where they have to actively control their attention, for instance during studying and working. However, the ability to efficiently control attention is not self-evident for many people including for example patients with attention-deficit hyperactivity disorder (ADHD) who exhibit characteristic symptoms such as inattention and impaired behavioral inhibition. In the past decades, there have been increases in workload and pressure but also in distractions such as mobile phones and television, presumably resulting in a higher demand in people regarding their ability to efficiently control attention. This has led to an increased interest in studying the underlying processes of attentional control and in finding possible ways to enhance attentional control. The cortical system has shown to be of crucial importance to cognitive inhibition processes such as attentional control. Sowell et al. (2003) found reduced white matter density in dorsolateral prefrontal and anterior parietal regions in ADHD patients compared to healthy controls, indicating the relation between reduced cortical efficiency and problems with attention and behavioral inhibition. Thus, it can be suggested that increasing cortical activity in primarily frontal brain regions may result in enhanced attentional control.
The most commonly used method for analyzing brain activity is electroencephalography (EEG), which is a non-invasive method to record electrical brain oscillations measured with electrodes placed on the scalp. Cortical activity is observed as power in the fast wave frequency bands alpha (8-12 Hz), beta (13-30 Hz) and gamma (>30 Hz), while subcortical activity is observed as power in the slow wave frequency bands delta (1-3 Hz) and theta (4-7 Hz). Cortical control over subcortical processes is reflected by negative correlations between slow and fast wave activity, and these negative correlations are associated with behavioral inhibition scores in children and adults (Knyazev & Slobodskaya, 2003). Also, coupling of slow and fast wave activity, reflected by strong correlations between these frequency bands over time, is indicative of cortico-subcortical crosstalk. It has been found that lower resting-state slow wave activity is associated with coupling of slow and fast wave activity, whereas higher resting-state slow wave activity is associated with decoupling (Schutter et al. 2006). These findings indicate that decreased slow wave activity compared to fast wave activity is related to higher cortico-subcortical crosstalk and cortical control. Thus, ratios between slow and fast wave activity, referred to as SW/FW ratios, provide an inversed indication of cortical control and inhibition processes such as behavioral inhibition and attentional control.
One of the most studied SW/FW ratios in relation to attentional control is the theta/beta ratio, which multiple studies have found to be negatively correlated with trait attentional control assessed by the self-report Attentional Control Scale (ACS) (Putman et al., 2010; Putman et al., 2014; Angelidis et
al., 2016). On the other hand, a different study failed to replicate the relation between the theta/beta
ratio and attentional control, which they believed to be due to differences in sample size and age compared to previous studies, but they did find coupling between delta and beta activity in parietal brain regions to be significantly associated with ACS scores (Morillas-Romero et al. 2015). Furthermore, the theta/beta ratio is also reported to be correlated with trait anxiety assessed by the STAI-t (State-Trait Anxiety Inventory), with ACS and STAI-t scores being negatively correlated (Putman et al., 2010; Putman et al., 2014; Angelidis et al., 2016). These findings indicate that a lower theta/beta ratio is associated with better attentional control and lower trait anxiety.
In view of all that has been mentioned so far, it can be proposed that attentional control can be enhanced by lowering SW/FW ratios including the theta/beta ratio, thus by decreasing slow wave activity and increasing fast wave activity, or by increasing coupling between slow and fast wave activity. Most research so far regarding manipulation of the theta/beta ratio has focused on neurofeedback training protocols, but theta/beta training has not yet been found to be effective in reducing the theta/beta ratio and enhancing performance (Doppelmayr & Weber, 2011; Studer et al., 2014). Thus, other ways to decrease the theta/beta ratio and to enhance attentional control should be investigated.
There is a growing interest in the effects of music on attentional control, since music has shown to be capable of influencing a variety of cognitive processes and stress measures. For example, Alexander (2018) reported that listening to music which varied in loudness in accordance with subjects’ alpha EEG oscillator amplitude resulted in a reduction of stress-sensation and positive shifts in mental and emotional status. Many studies investigating the effects of music on attention have focused on ‘The
Mozart Effect’, which was first proposed by Rauscher et al. (1993) and refers to an increase in performance on a standard IQ spatial reasoning task after listening to Mozart’s music. Verrusio et al. (2015) investigated changes in EEG related this possible Mozart Effect and found alpha power to significantly increase while listening to Mozart’s music. Since the Mozart Effect was proposed, most studies conducted to investigate the effects of music on attentional processes have used Mozart’s music. However, Mozart’s music was recently reported to elevate mood in both ADHD patients and healthy controls (Zimmerman et al., 2019). Thus, the proposed Mozart Effect may be an indirect effect of induced mood which in turn affects performance rather than a direct of music on attentional control, and so mood should be controlled for while investigating the effects of music on attentional control. Despite the exclusion of Mozart’s music in specific, classical music is still one of the most promising genres for enhancing attentional control compared to genres such as techno (Chen et al., 2008), HipHop (Chou et al., 2010), and hard rock (Geethanjali et al., 2012). This can be explained by other important characteristics of music, which are valence and tempo. Baldwin & Lewis (2017) reported that only music with a slow tempo and positive valence was able to positively affect performance in a Go-NoGo task reflecting behavioral inhibition. Above all, the degree to which participants find the music likeable seems to be of great influence, as shown by a deterioration of performance in the Go-NoGo task for participants who disliked the music regardless of the music’s valence and tempo (Baldwin & Lewis, 2017). Thus, music with a slow tempo and positive valence seems to be the best candidate to enhance attentional control, but controls need to be included to assess possible changes in mood. Music genres with slow tempo and positive valence are for example classical music or mellow background music such as elevator music.
Considering that a lot is yet to be discovered regarding the effects of music on attentional control, the proposed study aims to further investigate these effects. If listening to music can positively affect attentional control, students and workers could benefit from this when they have troubles controlling their attention. Also, if music is found to positively affect attentional control, this provides
possibilities for new attention-enhancing therapies based on music listening, thus providing a cheap and healthy alternative for manipulating attentional control.
Proposed study
Attentional control can be assessed in multiple ways including questionnaires, attentional tasks, and EEG correlates such as the theta/beta ratio. Trait attentional control is often assessed with the self-report Attentional Control Scale (ACS), which provides a total score reflecting the general ability to control attentional and scores for subscales reflecting the abilities to focus and shift attention (Derryberry & Reed, 2001). However, ACS scores were not found to be related to state attentional control measures (Williams et al., 2017), thus additional measures such as attentional tasks are needed to assess changes in state attentional control. The Attentional Network Task (ANT) developed by Fan
et al. (2002) is often used to analyze the efficiencies of the three main attentional networks proposed
by Posner & Peterson (1990) being the alerting, orienting, and executive control networks. As mentioned before, attentional control can also be analyzed by using EEG correlates that represent attentional control including the theta/beta ratio (Putman et al., 2010; Putman et al., 2014; Angelidis
et al. 2016) and coupling between slow and fast wave activity (Morrilas-Romero et al, 2015).
Many studies have investigated attentional control using one of these separate methods (ACS, ANT, EEG), but not many have combined these methods in order to investigate the effects of music on these different parameters for attentional control. The proposed study aims to investigate whether listening to music can positively affect attentional control as measured is EEG and ANT scores. An additional aim of the proposed study is to investigate whether the effects of music differ for people with different traits, including trait attentional control reflected by ACS scores and resting-state theta/beta ratio. These differences are interesting to investigate, considering that a previous study investigating EEG differences in good and bad responder to methylphenidate reported higher absolute theta power, a higher theta/beta ratio, and lower relative beta power in good responders compared to poor
responders (Clarke et al., 2002). It may be possible that the effects of music on attentional control are also more beneficial for people with a higher theta/beta ratio and lower trait attentional control. A pilot study has been carried out with the aim of investigating the effects of different sounds, being elevator music, nature sounds, and environmental sounds on EEG. The pilot study also contained a
cognitive task following the listening period, which enabled collection of additional data regarding the effects of music on performance measures including task duration and accuracy. The pilot study was conducted in 2019 at the NS station-lab located in Amsterdam, The Netherlands.
Methods pilot study
Experimental design
Before the start of the experiment, subjects were asked to complete a questionnaire which contained questions regarding their personal information such as age and gender, but also how stressed they were feeling at that moment and whether they could be classified as either more introverted or extraverted. The experimental design consisted of EEG measurements during a 90-seconds listening period and a subsequent cognitive task (see figure 1). The listening period contained three conditions over which subjects were randomly divided; (1) an audio fragment of elevator music, (2) an audio fragment of nature sounds, and (3) environmental sounds of the station without an audio fragment. Subjects were instructed to close their eyes and remain calmly seated during the listening period. After the listening period, subjects were instructed to complete the upcoming cognitive task and to press ‘b’ to continue to the task. The complete task consisted of 20 questions divided over 3
categories; (1) performing small calculations (10x), (2) having to come up with 3 rhyming words (5x), and (3) supplementing series of numbers (5x), which were presented in the same order for each subject (see Appendix I). Subjects were not specifically instructed to answer as quickly as possible but were expected to not take longer than necessary.
During the whole session, subjects were seated in front of a computer and EEG measurements were made using a single frontal midline electrode (Fz) placed directly on the forehead. The reference and ground electrodes were placed on the earlobe and forehead, respectively.
Complete data sets for each subject were ought to contain questionnaire results, EEG recordings, and files containing the task results. The task-result files contained all time-tracked actions performed while running the task in Stimulus-Presenter, including elapsed time during instructions and listening periods as well as keys pressed on the keyboard and typed in answers. From these files, performance-measures including total reaction time (RT) for all questions combined, RT for each task-category separately, and percentage of questions answered correctly (score 0-100).
Differences in EEG among groups were explored to investigate the effects of the different sound conditions on EEG. Performance measures including task durations and scores were also compared between groups to investigate the effects of different sound conditions on these performance parameters.
Figure 1. Experimental design NS pilot study. The listening period consisted of 90 seconds and the
cognitive task consisted of 20 questions.
Data analysis
The goals of data-analysis were (1) to explore group-differences in EEG frequency bands and SW/FW ratios during the listening period, with a specific interest in the theta/beta ratio, and (2) to explore group-differences in performance on the cognitive task, using RT and score as performance-measures.
EEG analysis
EEG data were analyzed separately for each subject using EEG Recorder Software running in a MATLAB environment. The first 70 seconds of the listening period were selected and portioned into 14 trials of 5 seconds each. To remove the 50 Hz artifact caused by electrical networks, a 0.5-48 Hz bandpass filter was applied. Each 5-second trial was visually inspected for artefacts, determined as having an amplitude exceeding +/- 100 µV, and the trials containing artefacts were discarded. In total, no more than three 5-second trials were allowed to be discarded due to artefacts. Average powers for delta, theta, alpha, and beta frequencies were calculated for each trial and inversed SW/FW ratios were calculated. Average power values were processed into more intuitive values by dividing -100 by the average power. Outlier analyses were performed for each frequency band and SW/FW ratio in each group and outlier values were excluded for further analysis regarding that specific frequency band or ratio.
Task performance analysis
The first question of the task was excluded for analysis, since no practice round was included and so subjects were still getting acquainted with executing the task. Task durations were calculated as the total summed up number of seconds used to complete all the tasks, according to the time indications in the results file. Outlier analysis was performed to identify subjects with an abnormally high or low task duration. Answers were only graded ‘correct’ if the answer was fully complete and accurate. Scores were calculated as the percentage of questions answered correctly, where subjects’ score should be at least 50.
These analyses were also performed for the three separate task categories in each group, being calculations, rhyming tasks, and number series tasks. The same inclusion criteria were applied for each category as described above for the total task measures.
One-way ANOVAs were performed in R studio to test for group-differences in average powers, SW/FW ratios, and performance measures and correlation analyses were performed to test for correlations between performance measures and powers or ratios. Shapiro-Wilk tests for normality and F-tests for equal variation were performed for each frequency band, ratio, or performance measure. A significance level of 0.05 was used for all statistical tests.
Results pilot study
Subjects & groups
All inclusion criteria were met by a total of 21 subjects with a mean age of 33 years (µ[SD] =
32.95[16.38]), including 14 males with a mean age of 33 years (µ[SD] = 32.71[17.48]) and 7 females with a mean age of 33 years (µ[SD] = 33.43[15.23]).
The music-, nature- and environment-groups contained 6, 9, and 6 subjects, respectively, with a male : female ratio of 2 : 1 in each group. One-way ANOVA showed that mean age between the music-group (µ[SD] = 29[13.8]), the nature-music-group (µ[SD] = 34.78[18.54]), and the environment-music-group (µ[SD] = 34.17[17.51) did not significantly differ (F2/18 = 0.228, P = 0.798).
EEG frequency bands & SW/FW ratios
Outlier analyses were performed for each frequency band and each SW/FW ratio within groups and outlier values were excluded from statistical analyses regarding that variable. Shapiro-Wilk normality test were also performed for each variable within groups and all power and ratios were found to be normally distributed within groups (P > 0.05). One-way ANOVAs were used to analyze group-differences for each frequency band and ratio separately (see table 1 for group statistics). No significant group-differences were observed for average power in any of the frequency bands including delta (F2/18 = 1.255, P = 0.309), theta (F2/18 = 1.638, P = 0.222), alpha (F2/18 = 0.546, P = 0.589), and beta (F2/18 = 0.205, P = 0.816) nor for any of the SW/FW ratios including theta/beta (F2/17 = 0.489, P = 0.621), delta/beta (F2/18 = 0.896, P = 0.426), and theta/alpha (F2/18 = 0.812, P = 0.459). See figure 2 for average powers and SW/FW ratios in each group.
Group Frequency band / ratio
Music µ(SD) Nature µ(SD) Environment µ(SD) Delta (1-3 Hz) 9.3337(0.0964) 9.1423(0.291) 9.2597(0.233) Theta (4-7 Hz) 8.6874(0.106) 8.5555(0.148) 8.6648(0.195) Alpha (8-12 Hz) 8.4724(0.123) 8.4431(0.179) 8.5339(0.179) Beta (13-30 Hz) 8.0408(0.077) 8.0284(0.0772) 8.0577(0.108) Theta/Beta 1.0739(0.0198) 1.0657(0.0187) 1.0755(0.0274) Delta/Beta 1.1609(0.017) 1.1388(0.0356) 1.1494(0.0354) Theta/Alpha 1.0254(0.0087) 1.0135(0.0163) 1.0156(0.0266) Delta/Alpha 1.1017(0.00531) 1.0832(0.0391) 1.0852(0.0279)
Table 1. Group statistics (mean and standard deviation) for average powers and SW/FW ratios.
Figure 2. Box plots representing (A) average powers of the individual frequency bands and (B)
SW/FW ratios for all three groups. The music-group is shown in blue, the nature-group in orange, and the environment-group in gray. Median, upper and lower quartiles, and outlier whiskers are shown. No significant group-differences were found for any of the frequency bands or SW/FW ratios.
Performance measures (RT and score)
Shapiro-Wilk normality tests were performed for task durations and scores within each group to analyze departures from normality. Task durations were normally distributed within each group, but scores were not normally distributed in the music-group and so non-parametric Kruskal-Wallis tests were used for comparing scores between groups while one-way ANOVA was used to compare task durations between groups. No significant group-effects were found for either task durations (F2/18 = 1.803, P = 0.193) or for scores (X2 = 2.587, df = 2, P = 0.274). See table 2 for group statistics and see figure 3 for mean task durations and scores.
Also, no significant group-effects were observed for durations and scores of the separate task categories (P > 0.05), except for a higher duration for the calculation category in the music-group compared to the nature- and environment-group (F2/18 = 3.898, P < 0.05). However, this difference was not significant after post-hoc Tukey HSD’s test. See figure 4 for the mean durations and scores on the separate task categories of each condition.
Performance measure Group Task duration (s) Calculations duration (s) Rhyming duration (s) Number series duration (s) Total score Music µ(SD) 229.89(62.24) 87.72(25.8) 99.29(42.99) 42.88(14.43) 90.35(14.67) Nature µ(SD) 164.91(41.21) 59.25(16.98) 68.58(20.29) 30.47(6.15) 87.72(6.96) Environment µ(SD) 225.95(117.65) 56.95(24.31) 120.5(66.10) 48.50(29.75) 93.86(8.43)
Table 2. Means and standard deviations for performance measures in each group. No significant
group-effect on performance was observed. Delta (1-3 Hz) Theta (4-7 Hz) Alpha (8-12 Hz) Beta (13-30 Hz)
-1 00 / A ve ra ge p owe r ( µV ) 7.8 8 8.2 8.4 8.6 8.8 9 9.2 9.4 9.6 9.8
Averaged powers frequency bands
Music Nature Environment
A
Theta/Beta Delta/Beta Theta/Alpha Delta/Alpha
Ra tio S W /F W 0.95 1 1.05 1.1 1.15 1.2 1.25 SW/FW ratios Music Nature Environment
B
Figure 3. Performance measures (A) task duration and (B) scores for all three groups (music, nature,
environment).
Figure 4. Mean (A) duration and (B) scores for the separate task categories and sound conditions.
Significant group-differences are indicated with ‘*’.
Correlations between EEG correlates and performance parameters
No significant correlations were found between listening period EEG correlates and total task duration (see table 3), except for an unexpected positive correlation between beta power and total duration of the task. The same was observed for correlations between EEG correlates and task scores (see table 4); no significant correlations were found, except for an almost significant positive
correlation between beta power and scores.
Predictor variable ~ outcome variable Spearman S Correlation rho P-value
Theta/Beta Ratio ~ Task duration 1748 -0.314 0.18
Delta/Beta Ratio ~ Task duration 1694 -0.1 0.67
Theta/Alpha Ratio ~ Task duration 1646 -0.0688 0.77
Delta/Alpha Ratio ~ Task duration 1466 0.0481 0.84
Delta power ~ Task duration 1494 0.0299 0.9
Theta power ~ Task duration 1670 -0.0844 0.72
Alpha power ~ Task duration 1440 0.0649 0.78
Beta power ~ Task duration 868 0.436 0.049*
Table 3. Results of correlation analyses EEG correlates and task duration. Variables used for
comparison and Spearman’s test statistics are presented. Significant p-values are marked with ‘*’. 0 20 40 60 80 100 120 140
Music Nature Environment
RT
(s)
RT task categories
RT calculations RT rhyming RT number series
Du ra tio n (s )
Calculations Rhyming Number series Duration task categories
A
* 0 50 100 150 200 250Music Nature Environment
RT
(s)
RT Total
A
Total task durationDu ra tio n (s ) 50 55 60 65 70 75 80 85 90 95 100
Music Nature Environment
Sco
re
(50
-100)
Score task categories
Score calculations Score rhyming Score number series
B
Sco
re
Calculations Rhyming Number series Scores task categories
50 55 60 65 70 75 80 85 90 95 100
Music Nature Environment
Sco
re
(50
-100)
Score
B
Total task scoresSc
or
Predictor variable ~ outcome variable Spearman S Correlation rho P-value
Theta/Beta Ratio ~ Score 1442.2 -0.0844 0.72
Delta/Beta Ratio ~ Score 1754.8 -0.139 0.55
Theta/Alpha Ratio ~ Score 988.5 0.358 0.11
Delta/Alpha Ratio ~ Score 1314.9 0.146 0.53
Delta power ~ Score 1370.6 0.11 0.64
Theta power ~ Score 1048.4 0.319 0.16
Alpha power ~ Score 1673.2 -0.0865 0.71
Beta power ~ Score 882.2 0.427 0.053
Table 4. Results of correlation analyses between SW/FW ratios, frequency band powers, and task
scores. Variables used for comparison and Spearman’s test statistics are presented.
Discussion of pilot results & implications for proposed study
In summary, the results of the pilot study did not show any significant effects of the different sound conditions including elevator music, nature sounds, and environmental sounds on EEG correlates related to attentional control or on performance measures. It was proposed that if music enhances attentional control, slow wave activity and SW/FW ratios, with a specific interest in the theta/beta ratio, would be decreased in the music-group. However, no significant group differences were
observed for any of the frequency band powers including delta, theta, alpha and beta, or for any of the SW/FW ratios including theta/beta, delta/beta, theta/alpha and delta/alpha. These findings indicate no significant effects of elevator music on the theta/beta ratio or any of the other EEG correlates related to attentional control. Furthermore, no significant effects of sound condition on performance during the cognitive task were found as measured by task duration and scores, except for a longer duration of the calculation tasks in the music-group compared to the other groups. Correlations between listening period EEG correlates and performance during the cognitive task were also analyzed but no
significant correlations were found, except for one unexpected result being a positive correlation between beta power and reaction time on the calculation tasks.
Although listening to an elevator music fragment did not positively affect EEG correlates that represent attentional control or performance during a cognitive task, these results should be interpreted with caution since there were several limitations in the experimental design of the pilot study. One major drawback is the lack of baseline EEG measurements, resulting in an unpaired design in which measurements in the three sound conditions are derived from different subjects. In such an unpaired design, sample sizes should be sufficiently large to correct for possible effects of sampling bias, but the sample sizes in the pilot study were small and so sampling bias may have affected the results. Also, if SW/FW ratios or frequency band powers initially differed between groups and actually were affected by the listening to music, this will have gone unnoticed due to the lack of baseline measurements. Additionally, the inclusion of baseline measurements and thus initial values for the EEG correlates related to attentional control enables analysis of differential effects of music on attentional control for subjects with for example high compared to low initial theta/beta ratio values. Another limitation of the pilot study arises from the continuous presence of artefacts in the EEG recordings during the cognitive task, which are thought to be due to movement of the hands during typing of the answers, movement of the eyes across the screen and blinking, and possibly other excessive movements during the task. This resulted in exclusion of the EEG data during the task for further analysis. However, analysis of EEG during the task is necessary to analyze attention-related changes in brain activity and how these are affected by the different sound conditions. Since subjects were not instructed to actively control attention during the listening period, the EEG data during the listening period are less informative regarding attentional control. More controlled experimental conditions should reduce the presence of artefacts and inclusion of electro-oculography (EOG) recordings enables correction of EEG data for ocular artefacts. Also, the gamma frequency band was not included in analysis due to application of a bandpass filter which filtered out all frequencies above 48 Hz. By using a band-stop filter instead of a bandpass filter, possible effects of music the gamma frequency band can also be analyzed.
Furthermore, changes should be made to the setup of the listening period. The listening period consisted of only 90 seconds, which is possibly too short to induce significant effects and does not allow for analysis of EEG changes over time. Also, since the listening period was separated from the task, direct effects of music on performance could not be analyzed. These problems can easily be solved by extending the listening period and by continuing the music during the task period. Moreover, most of the previous studies have focused on the effects of classical music, while there is no earlier evidence about the use of elevator music in specific. Thus, the absence of music-induced effects can also be due to the wrong type of music being used.
Lastly, the pilot study did not include direct parameters for attentional control. Analyzing solely task durations and scores on a cognitive task provides an incomplete picture, since these variables are mainly related to cognitive performance rather than attentional performance and they may depend on numerous other factors such as subjects’ mathematic and linguistic skills, their familiarity with typing on a keyboard, and effort put in completing the task without any reward being offered.
Considering all that has been mentioned so far, it is clear that several changes to the experimental design need to be made in order to collect sufficient data for a reliable conclusion regarding the effects of music on attentional control. A summary of the problems observed in the pilot and how these will be solved in the proposed study is presented in table 5.
Design Problem in pilot Implication proposal
Starting values attentional control
- No baseline theta/beta ratio - No trait scores
- Include baseline EEG measurements - Questionnaires for trait attentional
control & trait anxiety
EEG during task - Continuous artefacts Due to: - Typing of answers - Ocular artefacts - Excessive movements - Reduce artefacts How: - One-press answers - EOG recordings - Controlled environment
Music & listening period - Elevator music
- Listening period too short - Music separated from task
- Classical music
- Extend resting-state listening period - Continuously play music
Direct parameters attentional control
- Cognitive task
- No EEG analysis during task
- Attentional task
- EEG analysis during task => Task-related EEG changes
Table 5. Problems observed in the experimental design of the pilot study and how these are solved in
the proposed study.
Objectives & expectations
The proposed study aims to further investigate the effects of music on attentional control as measured in EEG and performance in an attentional task.
Regarding the effects of music on EEG, the proposed study aims to investigate music-induced changes in EEG correlates that represent attentional control during both resting-state and engagement in an attentional task. As mentioned before, these EEG correlates include SW/FW ratios with a specific interest in the theta/beta ratio, individual slow wave (delta and theta) and fast wave (alpha and beta) frequency band powers and coupling of slow and fast wave activity. If music positively affects attentional control, we expect to observe a decrease of the theta/beta ratio and possibly other SW/FW ratios in the music condition compared to the control condition.
To further investigate the effects of music on performance measures for attentional control, the Attentional Network Task (ANT) (Fan et al., 2002) is included which provides scores for the efficiency of the three important attentional networks being the alerting, orienting, and executive control networks (Peterson & Posner, 1990). It was reported in a previous study investigating interactions between these networks that the efficiency of the executive control network is enhanced by the orienting network but inhibited by the alerting network (Callejas et al. 2005). Thus, we expect to observe higher efficiencies of the orienting and executive control networks reflected by ANT-scores if the hypothesis is met that music positively affects attentional control.
Furthermore, the proposed study aims to investigate whether the effects of music on attentional control differ for people with different traits. Considering that the theta/beta ratio is a useful marker for trait attentional control and trait anxiety (Putman et al., 2010; Putman et al., 2014; Angelidis et al., 2016; Angelidis et al., 2019), baseline recordings of the theta/beta ratio will be used to separate subjects with high and low theta/beta ratios and to compare the effects of music on attentional control between these subjects. Additionally, questionnaires are included for self-assessment of trait
attentional control, trait anxiety, and problems with executive control. Based on previous research, it is expected that a higher baseline theta/beta ratio is associated with lower trait attentional control and higher trait anxiety (Putman et al., 2010; Putman et al., 2014; Angelidis et al., 2016; Angelidis et al., 2019). Also, it is expected that the effects of music on EEG and attentional performance are greater for subjects with a higher theta/beta ratio and lower trait attentional control, based on the previous finding that ADHD patients with a higher theta/beta ratio and more pronounced inattentive symptoms respond better to treatment (Clarke et al., 2002).
Approach
Subjects & location
The study will be conducted at the Science Park location of the University of Amsterdam. Subjects will include 100 students from the University of Amsterdam. Benefits accompanying the inclusion of only University students is that age and academic abilities will not confound the results since these are relatively equal among students. Exclusion criteria are an average GPA below 5.5 (Dutch grading system 0-10), age above 25 years, and diagnosed mental illness. ADHD patients are not excluded but are included as a separate group to enable comparison between the effects of music on subjects with and without ADHD. Subjects are requested to abstain from alcohol, caffeine, drugs, and medication affecting brain function, such as methylphenidate, in the 24 hours preceding the experiment as well as during the experiment.
Figure 5. Experimental design of the proposed study. Time indications of each block are presented in
minutes. Questionnaires include the ACS, SAS, webexec, and BMIS.
Overall approach and rationale
The experimental design consists of a paired design in which subjects’ attentional control as measured using EEG and the Attentional Network Task (ANT) is compared between a music condition and a control condition, which are divided over two days (see figure 5). Subjects will be randomly divided over two groups for which the conditions are reserved, thus one group starts with the music condition while the other starts with the control condition. Both conditions consist of a resting-state period of 8 minutes, in which subjects are calmly seated with their eyes closed during the first 4 minutes and open for the second 4 minutes, followed by a task period consisting of the ANT (Fan et al., 2002). In the music condition, classical music is played throughout the whole session, while no music is played in the control condition. EEG recordings are made continuously throughout the resting state and task periods, with markers indicating the starting and ending time of each period. Simultaneously, EOG recordings are made to enable correction of EEG data for eye movements and blinking.
Questionnaires for self-assessment of trait attentional control, trait anxiety, and mood are included at the start and end of each day.
Questionnaires
Round 1 Adaptation period
Questionnaires
Round 3 Adaptation period
Resting-state EEG (condition 1) Resting-state EEG (condition 2) ANT + EEG (Condition 1) ANT + EEG (Condition 2) Questionnaires Round 2 Questionnaires Round 4
+/- 30 minutes 5 minutes 8 minutes 30 minutes +/- 30 minutes
+/- 30 minutes 5 minutes 8 minutes 30 minutes +/- 30 minutes
EEG & EOG recordings
Practice block ANT Practice block ANT 5 minutes 5 minutes
Controls
Controls need to be included to correct for the possible influence of confounding factors such as mood and fatigue. To control for effects of fatigue or better acquaintance with the task during the second condition, subject will be randomly divided over two groups for which the music and control conditions are reversed. To rule out the possibility that performance during the first condition is affected by unfamiliarity with the task, a practice block of 5 minutes is included before the start of the ANT. Furthermore, the BMIS will be included with the other questionnaires to control for changes in mood.
Questionnaires
Questionnaires are included at the start and end of each day for self-assessment of traits and to assess possible changes in mood. These questionnaires are the ACS (Attentional Control Scale) for
attentional control (Derryberry & Reed, 2001), the SAS (Zung Self-Rating Anxiety Scale) for trait anxiety (Zung, 1971), the webexec for problems with executive control (Buchanan et al., 2010), and the BMIS (Brief Mood Introspection Scale) for current mood (Mayer & Gaschke, 1988), see
appendices II-V.
ACS, SAS, and webexec scores will afterwards be used to separate subjects with high and low
attentional control and trait anxiety. The effects of music on EEG and performance in the ANT will be compared between the groups with high and low trait scores. Paired t-test will be used to analyze changes in BMIS scores before and after each condition, reflecting changes in mood during the experiment.
EEG & EOG recordings
EEG recordings are continuously made during the resting-state period and the task period, starting after completion of the first round of questionnaires. Recordings are made using 9 scalp-electrodes placed according to the international 10-20 system at positions F3, Fz, F4, C3, Cz, C4, P3, Pz, and P4. Reference and ground electrodes are placed on the left earlobe and forehead, respectively.
Simultaneously, EOG recordings will be made using two electrodes placed on the left and right temples.
EEG analysis will be performed according to the methods described in the pilot experiment with a few modifications. In short, a band-stop filter of 48-52 Hz will be applied to remove the 50 Hz artifact caused by electrical networks while still including the gamma frequency band in the data. Resting-state and task conditions for all EEG channels will be divided into 1-minute trials for visual inspection of artefacts. Also, ocular artefacts will be removed using the EOG recordings. For further analysis, subjects’ EEG data are only included when at least 80% of the 1-minute trials is artefact-free. Average power densities and SW/FW ratios are calculated separately for the resting-state and task conditions with and without music in the same way as in the pilot study, but also including the gamma frequency band. Additionally, correlation analyses will be performed between slow and fast wave power densities to produce correlation measures reflecting coupling of slow and fast wave activity. Paired t-tests will be used to compare average power densities, SW/FW ratios, and correlation measures between the music and control conditions to analyze the effects of music on EEG, but also between the resting-state and task conditions to analyze task-related EEG changes.
Attentional Network Task - ANT
The ANT described by Fan et al. (2002) consists of a 30-minutes task to analyze the efficiency of the alerting, orienting, and executive control networks. This task has been used successfully in studies investigating attentional performance in ADHD patients (Lundervond et al., 2011), effects of methylphenidate on attention (Kratz et al., 2012), effects of music-induced mood on attention (Jiang
et al. 2011), and many more. Subjects are instructed to focus on a central fixation point and to quickly
and accurately report which way a central arrow (target) is pointing by pressing a specific key for left or right. The target appears above or below the fixation point and is accompanied by neutral,
congruent, or incongruent flankers (see figure 6). The target may or may not be preceded by a cue, with different cue conditions being no cue, double cue, center cue, and spatial cue (see figure 6). The spatial cue is the only cue condition that provides spatial information about where the target will appear. The same time windows are applied as used by Fan et al. (2002), which are shown in figure 7.
Individual markers for each cue and flanker condition will be implemented in the task in Stimulus Presenter. Timings of task conditions and reaction times of subjects will be logged, and ANT scores will be calculated using the algorithm of Fan et al. (2002). Regarding the efficiencies of the alerting and orienting networks, higher scores reflect higher efficiency, while for the efficiency of the executive network a lower score reflects higher efficiency. These scores are calculated for each subject in both the music and control conditions. Paired t-test will be performed to compare the efficiency scores between the music and control conditions. Additionally, unpaired t-test will be performed to compare ANT scores between groups with different traits.
Figure 6. Cue conditions and flanker conditions. The fixation point is visualized by a cross, the cue by
a star, and the target by a central arrow. Cue conditions include (1) no cue, (2) center cue, (3) double cue, and (4) spatial cue. Flanker conditions include (a) neutral, (b) congruent, and (c) incongruent.
Figure 7. Time windows in milliseconds (ms) for fixation point, cue, and target presentation in a
single trial. This example contains a spatial cue. T1 represents the time of fixation point presentation before the cue is shown. T1 is varied to control for anticipation and so attention has to be actively sustained. One total trial consists of 3500 ms.
Timeline
The study is expected to be completed in approximately 3 months. Subjects will be recruited in two weeks, but the start of data collection can already start after 1 week. Collection of all data, including questionnaire scores, EEG and EOG recordings, and ANT results will take approximately one month. Statistical analyses and the processing of results into an article are also expected to last one month each. See figure 8 for the complete timeline.
Figure 8. Expected timeline of the proposed study. The complete study will take approximately 3
months. 0 2 4 6 8 10 12 Recruitment of subjects Conducting experiments Questionnaire/EEG/ANT analyses Statistical analyses Processing of results into article
Time (weeks) Timeline study
T1 = 400 – 1600 ms Cue 100 ms 400 ms RT < 1700 ms 3500 ms - RT - T1
Ethical considerations
The study will be conducted according to the ethical guidelines of the University of Amsterdam.
Scientific & societal relevance
This study provides further insights in the effects of music on attentional control. Previous studies have mainly focused on the effects of music on cognitive performance, but not many have specifically focused on attentional control.
If music is found to positively affect attentional control, this can be applied in different societal aspects. First of all, this enables possibilities for future research into new attention-enhancing therapies based on listening to music. Many people, including for example ADHD patients, could benefit greatly from new no-drug methods to enhance their attention. A big advantage of new music-based therapies over current therapies such as methylphenidate is the expected absence of side effects which are commonly experienced in drug-based therapies. Furthermore, if music positively affects attentional control this can be applied to improve study and work settings by playing background music to enhance attentional control.
References
1. Alexander, F. (2018). Stress Coping Via Musical Neurofeedback. Advances in mind-body medicine, 32(2), 17-20.
2. Angelidis, A., van der Does, W., Schakel, L., & Putman, P. (2016). Frontal EEG theta/beta ratio as an electrophysiological marker for attentional control and its test-retest reliability. Biological psychology, 121, 49-52.
3. Angelidis, A., Solis, E., Lautenbach, F., van der Does, W., & Putman, P. (2019). I’m going to fail! Acute cognitive performance anxiety increases threat-interference and impairs WM performance. PloS one, 14(2).
4. Baldwin, C. L., & Lewis, B. A. (2017). Positive valence music restores executive control over sustained attention. PloS one, 12(11).
5. Buchanan, T., Heffernan, T. M., Parrott, A. C., Ling, J., Rodgers, J., & Scholey, A. B. (2010). A short self-report measure of problems with executive function suitable for administration via the Internet. Behavior Research Methods, 42, 709-714.
6. Callejas, A., Lupianez, J., Funes, M. J., & Tudela, P. (2005). Modulations among the alerting, orienting and executive control networks. Experimental brain research, 167(1), 27-37.
7. Clarke, A. R., Barry, R. J., McCarthy, R., Selikowitz, M., & Croft, R. J. (2002). EEG
differences between good and poor responders to methylphenidate in boys with the inattentive type of attention-deficit/hyperactivity disorder. Clinical Neurophysiology, 113(8), 1191-1198. 8. Chen, C. Y., Wong, W. K., Kuo, C. D., Liao, Y. T., & Ke, M. D. (2008). Wavelet real time
monitoring system: a case study of the musical influence on electroencephalography. WSEAS Transactions on Systems, 7(2), 56-62.
9. Chou, P. T. M. (2010). Attention Drainage Effect: How Background Music Effects Concentration in Taiwanese College Students. Journal of the Scholarship of Teaching and Learning, 10(1), 36-46.
10. Derryberry, D., & Reed, M. A. (2001). Attentional control, trait anxiety, and the regulation of irrelevant response information. Manuscript in review.
11. Doppelmayr, M., & Weber, E. (2011). Effects of SMR and theta/beta neurofeedback on reaction times, spatial abilities, and creativity. Journal of Neurotherapy, 15(2), 115-129. 12. Fan, J., McCandiss, B. D., Sommer, T., Raz, A., $ Posner, M. I. (2002). Testing the efficiency
and independence of attentional networks. Journal of cognitive neuroscience, 14(3), 3430-347.
13. Geethanjali, B., Adalarasu, K., & Rajsekaran, R. (2012). Impact of music on brain function during mental task using electroencephalography. World Academy of Science, Engineering and Technology, 66, 883-887.
14. Jiang, J., Scolaro, A. J., Bailey, K., & Chen, A. (2011). The effect of music-induced mood on attentional networks. International Journal of Psychology, 46(3), 214-222.
15. Knyazev, G. G., & Slobodskaya, H. R. (2003). Personality trait of behavioral inhibition is associated with oscillatory systems reciprocal relationships. International journal of psychophysiology, 48(3), 247-261.
16. Kratz, O., Studer, P., Baack, J., Malcherek, S., Erbe, K., Moll, G. H., & Heinrich, H. (2012). Differential effects of methylphenidate and atomoxetine on attentional processes in children with ADHD: an event-related potential study using the Attention Network Test. Progress in Neuro-Psychopharmacology and Biological Psychiatry, 37(1), 81-89.
17. Lundervold, A. J., Adolfsdottir, S., Halleland, H., Halmøy, A., Plessen, K., & Haavik, J. (2011). Attention Network Test in adults with ADHD-the impact of affective fluctuations. Behavioral and Brain Functions, 7(1), 27.
18. Mayer, J. D., & Gaschke, Y. N. (1988). The experience and meta-experience of mood. Journal of Personality and Social Psychology, 55, 102-111.
19. Morillas-Romero, A., Tortella-Feliu, M., Bornas, X., & Putman, P. (2015). Spontaneous EEG theta/beta ratio and delta–beta coupling in relation to attentional network functioning and self-reported attentional control. Cognitive, Affective, & Behavioral Neuroscience, 15(3), 598-606.
20. Posner, M. I., & Petersen, S. E. (1990). The attention system of the human brain. Annual review of neuroscience, 13(1), 25-42.
21. Putman, P., van Peer, J., Maimari, I., & van der Werff, S. (2010). EEG theta/beta ratio in relation to fear-modulated response-inhibition, attentional control, and affective traits. Biological psychology, 83(2), 73-78.
22. Putman, P., Verkuil, B., Arias-Garcia, E., Pantazi, I., & van Schie, C. (2014). EEG theta/beta ratio as a potential biomarker for attentional control and resilience against deleterious effects of stress on attention. Cognitive, Affective, & Behavioral Neuroscience, 14(2), 782-791. 23. Rauscher, F. H., Shaw, G. L., & Ky, C. N. (1993). Music and spatial task performance.
Nature, 365(6447), 611-611.
24. Schutter, D. J., Leitner, C., Kenemans, J. L., & van Honk, J. (2006). Electrophysiological correlates of cortico-subcortical interaction: A cross-frequency spectral EEG analysis. Clinical Neurophysiology, 117(2), 381-387.
25. Sowell, E. R., Thompson, P. M., Welcome, S. E., Henkenius, A. L., Toga, A. W., & Peterson, B. S. (2003). Cortical abnormalities in children and adolescents with attention-deficit
hyperactivity disorder. The Lancet, 362(9397), 1699-1707.
26. Studer, P., Kratz, O., Gevensleben, H., Rothenberger, A., Moll, G. H., Hautzinger, M., & Heinrich, H. (2014). Slow cortical potential and theta/beta neurofeedback training in adults: effects on attentional processes and motor system excitability. Frontiers in human
neuroscience, 8, 555.
27. Verrusio, W., Ettorre, E., Vicenzini, E., Vanacore, N., Cacciafesta, M., & Mecarelli, O. (2015). The Mozart effect: a quantitative EEG study. Consciousness and cognition, 35, 150-155.
28. Williams, P. G., Rau, H. K., Suchy, Y., Thorgusen, S. R., & Smith, T. W. (2017). On the validity of self-report assessment of cognitive abilities: Attentional control scale associations with cognitive performance, emotional adjustment, and personality. Psychological
Assessment, 29(5), 519.
29. William WK Zung. A Rating Instrument for Anxiety Disorders. 12(6): Psychosomatics 371-379. 1971.
30. Zimmermann, M. B., Diers, K., Strunz, L., Scherbaum, N., & Mette, C. (2019). Listening to mozart improves current mood in adult ADHD-a randomized controlled pilot study. Frontiers in psychology, 10, 1104.
Appendix I
The table below contains the complete cognitive task. Columns represent the number of the question, the question itself, and which category the question belonged to. For the rhyming tasks, the English translation for the given Dutch word is also included.
Question number
Category Question English translation
1 Calculation 800 + 366 = …
2 Rhyming Name 3 words that rhyme with ‘wachten’ ‘waiting’
3 Calculation 33 / 11 = …
4 Number series Finish the series: 6 12 18 24 …
5 Calculation 68 – 24 = …
6 Rhyming Name 3 words that rhyme with ‘spoor’ ‘railway’
7 Calculation 12 x 5 = …
8 Number series Finish the series: 83 77 71 65 …
9 Calculation 125 + 33 = …
10 Rhyming Name 3 words that rhyme with ‘brein’ ‘brain’
11 Calculation 36 / 9 = …
12 Number series Finish the series: 240 120 60 30 …
13 Calculation 48 + 95 = …
14 Rhyming Name 3 words that rhyme with ‘geel’ ‘yellow’
15 Calculation 578 – 386 = …
16 Number series Finish the series: 2 4 8 16 …
17 Calculation 9 x 8 = …
18 Rhyming Name 3 words that rhyme with ‘reis’ ‘trip’
19 Calculation 88 / 4 = …
Appendix II
The table below contains all questions of the Attentional Control Scale (ACS) as developed by Derryberry & Reed (2002). Items are scored on a 4-point scale (1 = almost never, 2 = sometimes, 3 = often, 4 = always). R = reverse-scored item.
Number Item
1 (R) It’s very hard for me to concentrate on a difficult task when there are noises around.
2 (R) When I need to concentrate and solve a problem, I have trouble focusing my attention.
3 (R) When I am working hard on something, I still get distracted by events around me.
4 My concentration is good even if there is music in the room around me.
5 When concentrating, I can focus my attention so that I become unaware of what’s
going on in the room around me.
6 (R) When I am reading or studying, I am easily distracted if there are people talking in the
same room.
7 (R) When trying to focus my attention on something, I have difficulty blocking out
distracting thoughts.
8 (R) I have a hard time concentrating when I’m excited about something.
9 When concentrating I ignore feelings of hunger or thirst.
10 I can quickly switch from one task to another.
11 (R) It takes me a while to get really involved in a new task.
12 (R) It is difficult for me to coordinate my attention between the listening and writing
required when taking notes during lectures.
13 I can become interested in a new topic very quickly when I need to.
14 It is easy for me to read or write while I’m also talking on the phone.
15 (R) I have trouble carrying on two conversations at once.
16 (R) I have a hard time coming up with new ideas quickly.
17 After being interrupted or distracted, I can easily shift my attention back to what I was
doing before.
18 When a distracting thought comes to mind, it is easy for me to shift my attention away
from it.
19 It is easy for me to alternate between two different tasks.
20 (R) It is hard for me to break from one way of thinking about something and look at it
Appendix III
The table below contains all questions of the SAS (Zung Self-Rating Anxiety Scale) as developed by Zung (1971). Items are scored on a 4-point scale (1 = a little of the time, 2 = some of the time, 3 = good part of the time, 4 = most of the time).
Number Item
1 I feel more nervous and anxious than usual.
2 I feel afraid for no reason at all.
3 I get upset easily or feel panicky.
4 I feel like I’m falling apart and going to pieces.
5 I feel that everything is all right and nothing bad will happen.
6 My arms and legs shake and tremble.
7 I am bothered by headaches neck and back pain.
8 I feel weak and get tired easily.
9 I feel calm and can sit still easily.
10 I can feel my heart beating fast.
11 I am bothered by dizzy spells.
12 I have fainting spells or feel like it.
13 I can breathe in and out easily.
14 I get numbness and tingling in my fingers and toes.
15 I am bothered by stomach aches or indigestion.
16 I have to empty my bladder often.
17 My hands are usually dry and warm.
18 My face gets hot and blushes.
19 I fall asleep easily and get a good night’s rest.
20 I have nightmares.
Appendix IV
The table below contains all questions of the webecex (an alternative for the DEX) as developed by Buchanan et al. (2010). Items are scored on a 4-point scale (1 = no problems experienced, 2 = a few problems experienced, 3 = more than a few problems experienced, 4 = a great many problems experienced). The total score is calculated by summation of the scores for all items.
Number Item
1 Do you find it difficult to keep your attention on a particular task?
2 Do you find yourself having problems concentrating on a task?
3 Do you have difficulty carrying out more than one task at a time?
4 Do you tend to ‘lose’ you train of thoughts?
5 Do you have difficulty seeing through something that you have started?
Appendix V
The table below contains the BMIS (Brief Mood Introspection Scale) as developed by Mayer & Gascke (1988). Subjects are instructed to circle the response on the scale that indicates how well each adjective or phrase describes their present mood (XX = definitely do not feel, X = do not feel, V = slightly feel, VV = definitely feel).
Lively XX X V VV Drowsy XX X V VV Happy XX X V VV Grouchy XX X V VV Sad XX X V VV Peppy XX X V VV Tired XX X V VV Nervous XX X V VV Caring XX X V VV Calm XX X V VV Content XX X V VV Loving XX X V VV Gloomy XX X V VV Fed up XX X V VV Jittery XX X V VV Active XX X V VV
Overall, my mood is:
Very unpleasant Very pleasant
