TUTOR BUSINESS MANAGEMENT A+ WORK, ON TIME, NO PLAGARIZING; ON TIME

DIDGITALASSIGNMENTYoung_consumers_brain_respons.pdf

Home >Business & Finance homework help >Management homework help >TUTOR BUSINESS MANAGEMENT A+ WORK, ON TIME, NO PLAGARIZING; ON TIME

Young consumers’ brain responses to pop music on Youtube

Kyoung Cheon Cha Department of Business Administration,

Dong-A University, Busan, Republic of Korea Minah Suh

Biomedical Engineering, Sungkyunkwan University – Suwon Campus, Suwon, Republic of Korea

Gusang Kwon R&D, Amore-Pacific Research and Development Center, Yongin, Republic of Korea

Seungeun Yang Cheil Communications, Seoul, Republic of Korea, and

Eun Ju Lee Sungkyunkwan University, Seoul, Republic of Korea

Abstract Purpose – The purpose of this paper is to determine the auditory-sensory characteristics of the digital pop music that is particularly successful on the YouTube website by measuring young listeners’ brain responses to highly successful pop music noninvasively. Design/methodology/approach – The authors conducted a functional near-infrared spectroscopy (fNIRS) experiment with 56 young adults (23 females; mean age 24 years) with normal vision and hearing and no record of neurological disease. The authors calculated total blood flow (TBF) and hemodynamic randomness and examined their relationships with online popularity. Findings – The authors found that TBF to the right medial prefrontal cortex increased more when the young adults heard music that presented acoustic stimulation well above previously defined optimal sensory level. The hemodynamic randomness decreased significantly when the participants listened to music that provided near- or above-OSL stimulation. Research limitations/implications – Online popularity, recorded as the number of daily hits, was significantly positively related with the TBF and negatively related with hemodynamic randomness. Practical implications – These findings suggest that a new media marketing strategy may be required that can provide a sufficient level of sensory stimulation to Millennials in order to increase their engagements in various use cases including entertainment, advertising and retail environments. Social implications – Digital technology has so drastically reduced the costs of sharing and disseminating information, including music, that consumers can now easily use digital platforms to access a wide selection of music at minimal cost. The structure of the current music market reflects the decentralized nature of the online distribution network such that artists from all over the world now have equal access to billions of members of the global music audience. Originality/value – This study confirms the importance of understanding target customer’s sensory experiences would grow in determining the success of digital contents and marketing. Keywords Neuromarketing, Digital music, Young consumers, Optimal stimulation Paper type Research paper

Introduction Music is the essence of culture, but how it has been recorded and shared has evolved over time, from vinyl to eight-track to cassette tape to CD, and now to online digital. In the last decade, digital technologies have radically affected the consumption of popular music. No longer is the music industry tied to the whims of a handful of multinational firms that control the global distribution of commercial music (Bennett and Peterson, 2004). Digital technology has so

Asia Pacific Journal of Marketing and Logistics Vol. 32 No. 5, 2020 pp. 1132-1148 © Emerald Publishing Limited 1355-5855 DOI 10.1108/APJML-04-2019-0247

Received 10 April 2019 Revised 8 May 2019 4 June 2019 Accepted 16 June 2019

The current issue and full text archive of this journal is available on Emerald Insight at: www.emeraldinsight.com/1355-5855.htm

This research was supported by the Korean National Research Foundation (2018R1A2B6004658).

1132

APJML 32,5

drastically reduced the costs of sharing and disseminating information, including music, that consumers can now easily use digital platforms to access a wide selection of music at minimal cost (Aguiar and Martens, 2016). Digitization has also spurred music’s move away from firm- led distribution to artist-led online distribution (Bockstedt et al., 2006). The structure of the current music market reflects the decentralized nature of the online distribution network such that artists from all over the world now have equal access to billions of members of the global music audience. The size of music industry grew significantly after streaming service according to a study by Wlömert and Papies (2016).

Nuttall et al. (2011) examined the music consumption of young people and music’s role of music in younger generation consumers’ daily lives by using focus group interview. The generation born in the digital age grew up with digital technologies and listening to music online on web sites via, music streaming service and other digital music channels, which provides access to millions of songs by artists from all over the world. For example, some of the most popular music posted on YouTube was created by lesser-known artists, but social network services (SNS) enabled it to “go viral,” receiving a massive number of hits online as young consumers share their using SNS, unimpeded by national boundaries (with a few exceptions). Digital platform for music listening has no boundary of nationality, device types, or geographic environments, etc. The number of views on digital platform is accrued organically and therefore is a highly meaningful performance metric in digital marketing. The number of views accrues in real-time from the world’s audience’s collective clicks. Hits on music digital platform not only represent a marketing success in global scale, but they also predict customer engagement and advocacy behavior.

What determines the performance of music on digital platforms? It is hard to isolate single success factor of any pop music, because are many variables in music including genre, composer, singer, producer, release timing, as well as external factors that could have increased or decreased the world’s interest in a specific kind of music. Previous research on the effect of music studied the importance of background music on brand attitude and information processing (Park and Young, 1986). Kellaris and Kent (1991) found that tempo can affect arousal and on behavioral intent. Tom (1990) studied the effect of music on the intention to purchase the advertised product. These characteristics holistically can cause the brain’s responses of so-called attention, leading to audience engagement and preference. Although it is difficult to directly explain why a certain pop music succeeds while another fails, we expect to be able to infer individual preference for certain kinds of pop music from the analysis of individual brain responses. Young consumers including the Millennials and the Generation Z, who were born in the digital age, grew up with digital technologies and listen to music by artists from all over the world on web sites like YouTube. Understanding what kind of music your consumers prefer and the brain responses they present when they listen to it is central to developing a successful marketing strategy in entertainment business.

This study defines the young consumers’ optimal sensory load (OSL) for online pop music based on the dimensions of tempo and frequency. The research tests the hypothesis that, in an online listening environment, songs with stimulation levels at or near the OSL capture young consumers’ attention. Functional near-infrared spectroscopy (fNIRS) is a non-invasive imaging method involving the quantification of chromophore concentration resolved from the measurement of near-infrared (NIR) light attenuation or temporal or phasic changes. The use of fNIRS as a functional imaging method relies on the principle of neuro-vascular coupling also known as the haemodynamic response or blood-oxygen-level dependent (BOLD) response. This principle also forms the core of fMRI techniques. Through neuro-vascular coupling, neuronal activity is linked to related changes in localized cerebral blood flow. fNIRS and fMRI are sensitive to similar physiologic changes and are often comparative methods. Studies relating fMRI and fNIRS show highly correlated results in cognitive tasks (Cui et al., 2011), but fNIRS has several advantages in terms of cost and portability over fMRI. Using fNIRS,

1133

Pop music on Youtube

we examine the neural responses of the brain’s attunement to pop music and find that total blood flow (TBF) to the medial prefrontal cortex (mPFC) increases when young consumers listen to pop music online with a level of near OSL. The hemodynamic signals of the attuned brain are also likely to change from a random state to a more rhythmic state, reflecting a listener’s cognitive synchronization with the rhythms of the music (London, 2012). Figure 1 illustrates the conceptual model of the research.

Background and literature review Digital music culture and marketing Previous research on pop music industry has been mainly focused on cyclic diffusion patterns and content features such as innovation and diversity (Peterson and Berger, 1975; Lopes, 1992; Peterson and Anand, 2004). Digitization allows musical content to be packaged audio-visually in the form of a music video. Representing sound as numerical values and recording and reproducing sound using audio signals that are encoded in digital form, digital music is a disruptive technology that has significantly lowered the entry barriers to the music industry by providing direct links between artists and consumers (Larsen, 2009). The music industry’s old business model was based on high up-front investments and high margins, but the new digital music market operates on a fast-diffusion model based on low margins. The success of online person-to-person (P2P) as a music delivery channel has expanded the boundaries of cultural diffusion, as it is consumed and shared on global virtual platforms (Bennett and Peterson, 2004). From the diffusion of virtual platform, there was a study regarding differences in brand-related user-generated content (UGC) between Twitter, Facebook and YouTube (Smith et al., 2012).

The most-viewed music videos on YouTube draw billions of viewers from all around the world. This open music platform is a highly competitive environment, as viewers on the YouTube site have millions of choices at their fingertips. Even so, local music that is produced in subcultural venues can be highly successful on the open music platform, as several artists from developing countries have achieved more than two billion hits, even though their lyrics were written in local languages that are spoken by only a few million nationals. The purpose of the current research is to determine the auditor- sensory characteristics of the digital pop music that is particularly successful on the YouTube website by measuring young listeners’ brain responses to highly successful pop music noninvasively.

Online Music Sound

Near

Under

OSLa

“Tune-In”

“Tune-Out”

Increased Total Blood Flow

• Rhythm (Attuned) Hemodynamics

• Random (Out-of tune) Hemodynamics

R L 2

Fp2 Fp1

NIRS

Decreased Total Blood Flow

Note: aOSL, optimal sensory load

Figure 1. The conceptual model

1134

APJML 32,5

Sensory load Krishna (2011) asked how our senses affect which songs we like and which songs we don’t. The sensory load theory that the neuromarketing literature proposed (Ramsøy, 2015) may provide a clue to the answer. Sensory load can be defined for each of the five human senses, but we focus here on the acoustic saturation point. The OSL for sound can be defined using factors like frequency (pitch), tempo, amplitude and timbre.

The temporal dimension refers to the perception of a song’s speed as fast or slow. The scales for tempo are the number of beats per minute (bpm) or the duration of a single beat, which is called an interonset interval (IOI). While temporal perception is subjective, music theorists have defined the upper and lower bounds of useful tempo for contemporary music. Westergaard (1975) published the range of commonly used tempos in music, noting that normal adults of the time would perceive 120 bpm (or 500 ms IOI) to be a moderately fast speed and 80 bpm (or 700 ms IOI) to be a moderate speed, as it corresponds to a biological standard for gauging musical tempo: heart rate. Normal resting heart rate for adults ranges from 60 to 100 bpm, so tempos that are slower than normal resting heart rate feel slow to most listeners, and a tempo that is twice the normal heart rate feels fast.

The second factor that defines OSL for sound is the hearing range. There is significant variation in individuals’ hearing range, especially in the highest frequency range, as the ability to hear high-frequency tones deteriorates with age. The upper limit of humans’ hearing range goes to 20 kHz for healthy young people (Cutnell and Johnson, 1994) but to only 12–14 kHz for a typical middle-aged adult. In pop songs, the human voice occupies bandwidths of 1–5 kHz, while the frequency of cymbals on a drum set can exceed 15 kHz, the usual limit for analog sounds up used for sound processing. Per the Nyquist–Shannon sampling theorem ( Jerri, 1977), a sample frequency used in digital processing is twice the highest component of the acoustic sound, so if a pop song is processed for online delivery using frequency ranges greater than 30 kHz, sounds at the highest frequency ranges may be perceivable only by young people.

Optimal sensory load for online music listeners Psychology’s perceptual load theory may shed light on the underlying mechanism that causes young people’s minds to wander off when a stimulus fails to provide a perceptual load that is above the mental threshold. The optimal level of stimulation theory suggests that individuals respond more negatively as the level of stimulation gains distance from an optimum point, so there is an inverted-U-shaped relationship between the stimulation level and the response to the stimulation (Steenkamp and Baumgartner, 1992). The perceptual saturation hypothesis suggests that, for the younger generation, the optimal point has moved to the top of the sensory-intensity scale, near the ceiling of the human sensing range.

Maylor and Lavie’s (1998) psychology study noted that young people can handle or control higher levels of sensory load than older people can. Three mechanisms underlie the positive effect of sensory load on young people: human auditory senses peak at adolescence and decrease with age (Olson, 1967); compared to live music, digital music provides limited sensory experiences, as it is devoid of tactile or olfactory sensation, so the loss of certain features that contribute to the feelings of music’s “realness” is unavoidable after sound is compressed to a digital format; and the young consumers may have significantly reduced attention capacity because of their heavy use of computers and smart phones throughout their adolescence, when the brain grows rapidly (Smallwood, 2013). Neuroscientists who study cortical plasticity have suggested that there is a significant and unprecedented difference between how the younger generation’s brains are plastically engaged in digital life and how those of average individuals from earlier generations are so engaged (Bolton, 1894). In 2012, a survey of more than 2,000 US secondary school teachers reported that 87 percent of the teachers who participated in the survey saw digital technologies as creating

1135

Pop music on Youtube

an “easily distracted generation with short attention spans.” Another 2012 survey of four hundred UK teachers reported a significant decline in the attention spans of young students (McAuley et al., 2006). While young consumers may not be able to sustain their attention, they tend to exhibit intermittent bursts of short-duration, high-intensity brain activity. The same directional change in music choices – from those that require a longer attention span to those that do not – suggests that an online song that provides short sparks of stimulation is likely to succeed in capturing their attention.

Rhythms in brain responses Brain rhythms are a mechanism for integrating neural circuits, which oscillate within a range of frequencies (Buzsáki, 2004). For example, the human brain waves, measured by electroencephalogram (EEG), can be represented by five frequency bands: delta (δ: 1–3 Hz), theta (θ: 4–7 Hz), alpha (α: 8–13 Hz), beta (β: 14–30 Hz) and gamma (g: 31–100 Hz). A simplified view of the relationship between brain oscillations and music is that EEG signals and music oscillate in sync, so brain signals can emulate certain features of the music (McAdams and Bregman, 1979). Single neuron cells fire electrophysiological signals in g rhythms ( Jensen et al., 2007) or θ rhythms (Benchenane et al., 2010), and EEG studies have shown that long-range g oscillations may play a role in music perception (Bhattacharya and Petsche, 2001), as the EEG g oscillations are the brain’s tool for processing music, representing the efficacy of the feature-binding process during perception. As a result, musicians tend to acquire high levels of long-range g synchrony after years of training (Bhattacharya and Petsche, 2001). Fujioka et al.’s (2009) neuroscience study reported that EEG β band (20–30 Hz) and g band (30–50 Hz) oscillations reflect the brain’s responses to rhythmic sounds.

Research hypotheses The human brain generates electrical waves as long as it lives. Since the dynamic nature of brain rhythm is at work in all kinds of human brain function, neuroscientists have used brain rhythm to understand the brain’s functions. Since the work of Gerstein and Mandelbrot (1964), many attempts have been made to use random-walk analyses to account for brain responses like the spiking of neurons, cell migration and motor variability (Abdelnour et al., 2014; Chaisanguanthum et al., 2014; Chiang et al., 2014; Modzelewski et al., 2011; Simen et al., 2011). Like any other biological system, the brain pursues functional efficiency at all levels of operation – in the brain’s case, from the neuronal cell level to the neural network level. Before one can determine the presence of a periodic rhythm vs a random state in brain activation, one must determine whether external stimuli can shape the brain’s modulation pattern. Brain-wave patterns are affected by whether the neural circuit that governs a particular set of brain functions reaches a significant level of activation. The bottom-up processing of external stimuli can be affected by top-down processing; in other words, the execution of higher-order cognitive attention can affect the degree of randomness in the bottom-up processing of external sensory inputs like that of music.

Unlike EEG signals, the rhythms of hemodynamic signals are not commonly calculated, possibly because hemodynamic signals are sluggish. The random-walk test on neural time series has been applied only recently to magnetoencephalography data (Kipiński et al., 2011) and it has rarely been applied to hemodynamic signals measured with magnetic resonance imaging or near-infrared spectroscopy (NIRS). However, since hemodynamic responses are the result of neuro-vascular coupling – a dynamic event among the brain’s neurons, glias and vasculatures – it is possible to calculate the degree of hemodynamic signals’ randomness as surrogates for neuronal activity. While brain activities are inherently random and noisy in their natural state, when the brain rhythm is modified by music that

1136

APJML 32,5

provides appropriate levels of sensory stimulation, the brain’s signals begin to reflect the music’s rhythms. This reflection is called “attunement.”

The effect of sensori-neural stimulation on hemodynamic responses measured by fNIRS has been reported in neuroscience research that found that auditory stimulation and music elevated the concentrations of oxygenated hemoglobin (HbO2) and total hemoglobin (HbT) in blood flow to certain regions of the brain (Hoshi and Tamura, 1993; Kotilahti et al., 2010; Sakatani et al., 1999). However, studies have shown decreases (increases) in children’s (adults’) prefrontal cerebral volume after they play computer games. For example, one study suggested that the level of attention may modulate the directional changes in HbO2 and HbT concentrations (Nagamitsu et al., 2006). According to music theorists, when the brain is entrained, the attention is mobilized with the music (London, 2012). The results of our study provide evidence that, when Young consumers listen to music, their perceptual stimulation level is related to the degree of randomness in their brain responses and to the quality of the sensory experience. Drawing on the optimal level of stimulation theory, we predict that TBF is higher for a stimulus that is above OSL than it is for a stimulus that is below OSL. We also predict that the hemodynamic rhythm of related brain regions to music that is above OSL adopts a regular, predictable pattern. Hence, we propose the following research hypotheses:

H1. Digital music that provides acoustic stimulation near the OSL creates brain responses in the form of higher TBF (a) and lower randomness in HbO2 concentrations (b) than does digital music that provides acoustic stimulation that is below the OSL.

Experiment Participants In total, 56 subjects in their twenties (23 females; mean age 24 years) with normal vision and hearing and no record of neurological disease took part in the study in return for extra course credit and financial incentives. Subjects were recruited to include people who listen to online music frequently and a variety of nationalities. Foreign nationals who participated in the experiment are those who have stayed in Korea for at least six months. Ethical approval was issued by the University Institutional Review Board and informed consent was obtained from all volunteers. Experiments were performed according to the guidelines of the University Institutional Review Board. In general, the number of subjects in neuromarketing studies is very limited because it is expensive and time consuming. For example, Meyerding and Mehlhose (2018) used the data from 31 subjects using fNIRS and Yoon et al. (2006) used 20 subjects only through fMRI. Instead of recruiting a high number of subjects, neuroscience studies tend to use repeated exposures to a controlled set of experimental stimuli. Our subjects spent an average of more than an hour participating in and listening to the musical stimuli (more than 15 times for each song). Because each song lasted 40 s, the multiple repetitions can be tiresome. Moreover, listening to music that is delivered over a headset for an hour can cause fatigue. Therefore, we limited the number of sample songs to reduce subject fatigue.

Stimuli The stimuli used were audio sounds of five pop songs posted on a popular website where digital-generation artists and listeners upload and share music and videos. The five songs were selected because they were the musical genres that are familiar to the Young consumers. Songs were chosen after a pilot study with a small number of subjects of the same age group. While individuals’ musical tastes and preferences can differ widely, we chose several pop genres that many in this age group like. Dunn et al. (2012) investigated that the music preference for various music genres and found that little difference in music

1137

Pop music on Youtube

preference by listeners’ nationality. The number of hits is a variable that reflects how widely distributed a particular song is, but it is difficult to ensure that the participants had never heard these five songs so the study could be generalized. Research on music has used music by Mozart, Bach and Beethoven (Mitterschiffthaler et al., 2007; Verrusio et al., 2015; Wilkins et al., 2014). If one had to avoid any famous music, we would be missing out a lot of useful information, such as how musical masterpieces like those by Mozart can alter our brain activities. We proposed that frequency and tempo influence brain activations and are objective assessments of the musical sound. Familiarity is a subjective and perceptual variable that we do not address, so whether previous exposure to music changes the effects of frequency and tempo should be examined in future research. We note this limitation in the limitation section. The original versions of the pop music selections were about three minutes long, but we selected for the experiment four ten-second segments from each music video that contained the full versions’ artistic and creative elements (e.g. pitches, melodies and rhythms). The genres of the five songs were contemporary rock, hip hop, and dance music, all of which are created and typically enjoyed by the Young consumers. The musical time patterns were all 4/4 time (Table I), but each had a different tempo/frequency mix, as illustrated in Figures 2 and 3. All five songs were released between 2008 and 2013. In the process of song selection, the artists’ vocal factors were less important than the dimensions of tempo and frequency. Average numbers of daily hits were in the order of Songs A–E, as presented in Table I. All five songs were mentioned in the media as “catchy” or otherwise notable or were on the Billboard chart in the year of their release (Table II).

Figure 2 presents the sonic characteristics of the five songs. Song A, which is dance music, had the highest average number of daily hits (more than 1.6m) among the five stimuli. The acoustic sonogram of Song A shows that it almost always uses the highest frequency ranges – up to ~35 kHz – during the four ten-second segments used in the experiment. Song A used two tempos: a base tempo of 150 bpm (or 400 ms IOI) and a tempo as high as 480 bpm (125 ms IOI) in one segment. Based on this tempo/frequency mix, song A’s level of perceptual stimulation is above the perceptual threshold. Song B has the second-highest number of hits (an average of 1.5m hits per day). At 140 bpm (430ms IOI), its tempo is the second-fastest of the five selections, and its frequency range is more than 30 kHz. The perceptual load of Song B is also high, so Songs A and B meet the OSL for both tempo and frequency.

Song Average daily hits Rhythm

A 1,662,280 4/4 B 1,531,718 4/4 C 787,512 4/4 D 3,151 4/4 E 82 4/4

Table I. Five song stimuli

Spectral View

OSL: 30 (kHz)

~ Near Optimal Frequency Level?

Almost Always

Often Frequently Never Never

A B C D E

0 (sec) 40 0 (sec) 40 0 (sec) 40 0 (sec) 40 0 (sec) 40 Figure 2. Sonic frequencies

1138

APJML 32,5

Song C uses frequency ranges as high as 30 kHz during most of four ten-second segments used in the experiment, but the tempo, which is at the threshold level 120 bpm (500 ms IOI), is slower than the tempos of songs A and B. Song D is hip hop music with a slow tempo (700 ms IOI or 86 bpm), which is moderate speed based on the older tempo perception scale (Figure 3) but may be slow for the Young consumers. The frequency range used for the four ten-second segments used in the experiment is the narrowest among the five stimuli (~ 15 kHz), and the upper-frequency limit is also the smallest among the five because hip hop primarily features artists’ voices without the back-up of a full band. Song E is a contemporary rock song with the slowest tempo among the five selections (840 ms IOI or 71 bpm), and its frequency range is the second-lowest (~18 kHz).

In sum, two songs (A and B) provide hearing stimulation that meets the OSL in terms of both tempo and frequency and two songs (D and E) provide levels of stimulation that are under the OSL. Song C is at the OSL for tempo, and it meets the OSL at its highest frequency.

Functional near-infrared spectroscopy The transparency of biological tissue to light in the near-infrared wavelengths makes NIRS possible. NIRS is non-invasive and portable, and it has a cost advantage. The incident near-infrared light from a transmitting optode (source) is scattered through the tissues, and the reflected light is detected by a receiving optode (detector). The amount of the source light that a tissue absorbs depends on the light’s wavelength, and the oxygenation status determines the brain’s absorption of the light. The loss of the intensity that is due to the absorption of the photons can be measured in units of optical density (Zaramella et al., 2001). The changes in (oxy-Hb) and (deoxy-Hb) can be calculated according to themodified Beer–Lambert law (Kocsis et al., 2006) using two wavelengths of near-infrared light – in our case, 780 and 850 nm.

IOI (ms) 2,000 1,414 1,000 700 500 350 250

bpm 30 42 60 80 120 168 240 (Heart Rate) (Double HR)

OSL

Average IOI

840 700 500 430 400 275 125 (ms)

E D C B A

Too Slow (800+ ms)

Very Slow (700 ms)

Moderate (400 ms)

Fast (~300 ms)

Tempo Perception in the Digital Age

Figure 3. Temporal perceptions

fNIRS Song Average daily hits Mean TBF after 40 s Mean BORP

A 1,662,280 0.194 (0.556) 0.303 (0.040) B 1,531,718 −1.366 (0.393) 0.311 (0.041) C 787,512 −1.527 (0.561) 0.381 (0.046) D 3,151 −2.324 (0.398) 0.499 (0.043) E 82 −2.544 (0.518) 0.482 (0.047) Notes: fNIRS data were recorded at a channel over the right mPFC (chapter 5). Standard errors are listed in parentheses

Table II. Total blood flow (TBF) and brain

oxygenation random probability (BORP)

1139

Pop music on Youtube

We used a 12-channel wireless fNIRS system (Biomedical Optics Lab, Korea University) with sampling rates of 8 ~10 Hz to measure the participants’ hemodynamic response while they watched the videos. The system consists of three light sources and five detectors (a 3 × 7 grid). Channel locations are shown in Figure 1. The fNIRS probe was attached to each subject’s forehead, and the detectors of the lowest line were set along the Fp1 and Fp2 electrode line according to the international 10/20 system. Measurements from channels 1, 2, 11 and 12, which contained noise from movements of the subjects’ heads, hair and sweat, were excluded from further analysis.

Neuroscience research has recorded acoustic stimulation in various regions of the brain, including the temporal brains of newborn babies (Hoshi and Tamura, 1993) and the frontal brains of adults (Sakatani et al., 1999). Since infrared light cannot penetrate hair, brain regions that are not covered by hair, such as the prefrontal cortex, are well-suited to an fNIRS study. When a Millennial listens to music online, the motivation is usually enjoyment, so changes in brain activity that are due to popular music should occur in brain areas that are associated with reward-related processing – that is, the medial pre-frontal cortex (mPFC) (Haber and Knutson, 2010). As the mental function of pleasurable experience that is modulated in the medial frontal cortex increases, TBF to this region increases. In particular, the processing of sounds is dominantly modulated by the brain’s right hemisphere (Kaiser et al., 2000). We analyzed TBF to the right brain area of the mPFC at channel five using fNIRS.

Procedure After arriving at the lab, participants received instructions on the tasks they would be asked to perform. They were then seated, and the experimenter applied the fNIRS equipment to their forehead areas and gave them headphones with which to listen to the songs. Five songs were presented to the subjects in a random order while the subjects’ brains were scanned by fNIRS.

All participants heard all five songs which were repeated four times. Such use of within- subject design is common in neuroscience research (Herzmann et al., 2012) due to high cost and inter- and intra-individual variances. The volume levels of the five songs were kept constant. We focused on the acoustic OSL and conducted additional analysis on the time of exposure to sound. All subjects listened to the sounds of five songs as a within-subject design. In total, 22 resting periods were provided between each of the songs’ four ten-second segments. After the participants listened to the five songs, the task ended.

Analysis of fNIRS data We recorded the changes in the 56 subjects’ HbT, as measured by fNIRS, while they listened to each song. The blood oxygenation random probabilities (BORP) data over the four ten-second periods were calculated using the augmented Dickey–Fuller (ADF), as explained next.

Calculating the random walk. As a time series, a random-walk process is an autoregressive model in which the dependent variable yt depends linearly on its own previous value, yt–1, after controlling for a linear time trend and a constant term:

yt ¼ pþbtþfyt�1þet ; (1)

where yt is a variable of interest, et is the error term at time t, α is a constant (intercept), β is the coefficient on a deterministic time trend (t ¼ 1, 2,…, n), and f is a coefficient for autoregressive dependence or order 1 (or p), with yt−1 (or yt−p). The test of the coefficient f’s statistical significance as autoregressive dependence requires the calculation of probabilities for random walks such that, when there is sufficient support that f ¼ 1, the process can be described as random walk.

1140

APJML 32,5

The variance could also be unstable in random-walk processes. In the following equation, B is a back shift operator, so:

yt ¼ pþbtþyt�1þet (2)

- 1�Bð Þyt ¼ pþbtþet

-yt ¼ pþbt 1�Bð Þ þ

et 1�Bð Þ

-yt ¼ pþbt 1�Bð Þ þ 1þBþB2þB3þ . . .

h i et :

Therefore, a random-walk process is nonstationary, and its variance increases with time t. “Shocks” gradually die away in a stationary time series, but they do not in a random

walk. Stock price is an example of a random-walk process. An ADF test, an augmented version of the Dickey–Fuller test for serial correlation, is a

test for a random walk in a time series sample. There are three options in an ADF test: with a constant, with a constant and a time trend, and with no constant or time trend. We used the model with a constant and a time trend because most blood oxygen-generation data has a time trend. The following equation is the ADF test equation:

Dyt ¼ pþbtþryt�1þ XP

p¼1

jpDyt�pþet ; (3)

where α is a constant, β is the coefficient on a deterministic time trend, ρ is the coefficient on yt−1, and p is the lag order of the autoregressive process, which could be selected using a Schwarz Information Criterion. The dependent variable of Equation (3) changes from that of Equation (1) by differencing (Δyt ¼ yt−yt−1). If ρ ¼ 0, the test equation becomes yt ¼ p + bt + yt−1 +…+ et, similar to Equation (2). The more negative the ADF statistic of ρ, the stronger the rejection of the hypothesis that the time series is a random walk at some level of confidence. The ADF test is then carried out under the null hypothesis ρ ¼ 0 against the alternative hypothesis of ρ o 0. Since the test is done over the residual term rather than using raw data, it is not possible to use standard t-distribution to provide critical values. Therefore, the statistic t has a specific distribution, the Dickey–Fuller table. Once a value for the test statistic:

DFt ¼ r̂

Standard Error of r̂ ;

is computed, it can be compared to the relevant critical value that Dickey and Fuller (1979) tabulated. If the test statistic is less than the (larger negative) critical value, then the null hypothesis of ρ ¼ 0 is rejected and no random walk process is present. (This test is non-symmetrical, so we do not consider an absolute value for the test statistic.) Therefore, the p-value is the probability (greater than |DFτ|) of obtaining an ADF test statistic result that is at least as extreme as the one that was observed, assuming that the null hypothesis (ρ ¼ 0, that is, random walk) is true. The p-values associated with detecting random walks in ADF tests are used as the degree of randomness, referred to hereafter as the BORP.

1141

Pop music on Youtube

We used custom-written MATLAB codes to perform ADF tests on the fNIRS data with a specified model of lag consisting of the data points of a two-second cycle. When time series data are constant and/or have no variance, the ADF test cannot be carried out because of the multicollinearity of the independent variables.

Results The research hypothesis predicted that songs that provide strong sensory stimulation above the OSL increase the TBFs of those in the Young consumers more than do songs that present a sensory stimulation level that is under the perceptual threshold. Figure 4 shows the grand average time-course fNIRS plot of the five-song stimuli. TBF can be directly obtained as a product of HbT (Wyatt et al., 1990). After the music began in the experiment, the subjects’ concentrations of HbT increased until HbT reached its peak at around five to eight seconds, after which it decreased for the next 30 s (Figure 4). This result is generally consistent with an industry study that reported that consumers’ attention span lasts no longer than eight seconds. There was a divide in the hemodynamic responses between the two songs that had more than a million hits per day (A and B) and the remaining three (C–E).

Before the ANOVA, we conducted the Shapiro–Wilk test for the five songs. Normality assumption (pW0.05) was met for Songs A–E, so we make ad hoc group comparisons among these songs. For our correlation analysis, we measured all of the variables on a continuous scale, and each participant listened to all five songs. We also checked for and removed outliers that were greater than +/− SD 3 since outliers can skew the results of the correlation. We conducted repeated measures of ANOVA on TBF, measured at 40 s, for the five songs since TBF at the end of each set of song segments can represent the level of sustained attention. The multivariate test for the model was significant, and the songs’ main effect on TBF was significant (Wilk’s λ 0.75, F(4, 51)¼ 4.249, p¼ 0.005, η2 ¼ 0.25). After the four ten-second segments of each song had been played, the TBF levels song A were at the highest level, followed in order by Songs B–E (Figure 4). Pairwise comparisons after the Bonferroni correction showed that there was a significant divide in the length of time that the TBF levels remained at the highest level between songs A/B and songs D/E (po0.05). These results support H1a.

Figure 5 presents the relationship between TBF and daily hits and that between BORP and daily hits. The Pearson correlation coefficient between TBF and daily hits was 0.88 (po0.05), and the correlation coefficient between BORP and daily hits was −0.96 (po0.05). Pairwise comparisons after the Bonferroni correction showed that there was a significant

–20

0 10 20 30 40

TIME (Sec)

A B C D E

To ta

l H b

C on

ce nt

ra tio

n (�

M )

Note: Concentrations of HbT measured at Channel 5 over the right mPFC area

Figure 4. Average Time-Course at the Right Medial Prefrontal Cortex (rMPFC)

1142

APJML 32,5

divide in the length of time that the TBF levels remained at the highest level between songs A/B and songs D/E (po0.05). Other differences were not significant, possibly because neural data contains a large portion of individual variance (between-subject F-test: F(1, 54) ¼ 372.675, po0.001, η2 ¼ 0.873). These results are consistent with ourH1b, that pop music that presents stimulation above OSL can reduce the randomness in hemodynamic signals.

The changes in the participants’ hemoglobin concentrations while they listened to popular songs show a mean reverting tendency with low BORP numbers, a “rhythm” such that a system recovers order and balance in due time. The brain’s response to less popular songs were random-walk processes, which represents a neural drain, a process in which the brain fails to recover from oxygen depletion because of boredom.

Discussion and conclusion In this study, we redefined the sensory load of sound and music for the Young consumers using fNIRS for brain investigation. A popular perfusion neuroimaging technique, fMRI is based on hemodynamics, but its high cost is a barrier in real-world market research. fNIRS

0.5

–0.5

–1.5

–2

–2.5

–3.5

–3

–1

TBF (40 sec) Daily Hits

10,000,000

1,000,000

100,000

10,000

1,000

100

10,000,000

1,000,000

100,000

10,000

1,000

100

0.55

0.50

0.45

0.40

0.35

0.30

0.25

BORP (40 sec)

Daily Hits

(� M

) (�

M )

A B C D E

(a)

(b)

Notes: Daily hits are presented in logarithmic scale. Pearson correlation coefficients (r) are (a) between TBF and average daily hits (r= 0.88, p<0.05) and (b) between BORP for sounds and average daily hits (r= –0.99, p<0.05)

Figure 5. TBF, BORP and

YouTube average daily hits

1143

Pop music on Youtube

enables researchers to observe the frontal brain’s hemodynamic responses in an affordable and user-friendly manner. As NIRS uses infrared light penetration into the frontal brain up to only 2.2 inches of depth, its major disadvantage as a neuromarketing device is its short penetration, which usually prevents it from accessing the mesolimbic brain areas, such as the nucleus accumbens, that modulate reward-processing.

We obtained fNIRS measurements from 56 subjects while they listened to popular online music. We predicted that popular music, as an acoustic stimulus, has to present intense forms of sensory stimulation at a level above the OSL to achieve worldwide commercial success (as measured by, e.g. YouTube daily hits). The results show that Young consumers may require a level of stimulation that is above OSL to sustain their attention. With our fNIRS experiment, we identified the neural indicators of attunement to online music in hemodynamic signals as TBF to the right mPFC, which was at a higher level when subjects listened to songs that provided stimulation above the perceptual threshold than when they listened to other songs. The Young consumers’s brain responses to popular online music create a musical reality that unfolds in real time, showing signs of synchronizing perception of the musical reality. Taken together, our results suggest that, for Young consumers, music has to be faster and use higher frequencies than what was stimulating for past generations.

Songs that have attracted more than a million hits per day on YouTube appeared to create a distinctive rhythm in the brain as a consequence of a song-to-brain match. Randomness in brain responses was reduced by riding on the rhythm of a hit song that presented a high level of acoustic stimulation. The more popular music, songs A and B, elicited more predictable hemodynamic responses in the fNIRS channel near the right mPFC, which is related to reward processing of sound stimuli, than did the less popular songs, songs D and E.

Songs D and E were well-known, as they were on the 100 Billboard chart or well-covered by the press, and musical genres with slow tempos and low-frequency ranges are far from second-rate. Therefore, our results were likely due to the online factor, as when people listen to the raw sounds coming from a live performance, they may sense frequency ranges over 15 kHz through tactile (skin or bone) conduction. With these vibrations ulterior to hearing, the live audience may well enjoy hip-hop or slow-rock performances. However, when the music is processed, extracted and compressed for quick delivery in an online channel, such tactile conduction becomes improbable. This technologically limiting factor may be the cause of systematic disadvantages for certain musical genres when they are played online.

Taken together, the unified perception of sensorymodalities plays a central role inmodulating rhythmic brain activity. Changes in TBF and the degree of randomness of the right mPFC’s hemodynamic responses may be related to Young consumers’ preference for online music.

Implications Music is the essence of culture. With digital technology, young consumers can enjoy music from almost anywhere in the world, no longer constrained by geography or cultural rules in determining their musical preference. Numerous characteristics holistically contribute to the success of a pop music. Although we try to avoid taking an oversimplified reductionist approach of decomposing music into its elements, we speculate that certain sound features of pop music can stimulate the brain of young consumers. In the young consumers’ brains which are still maturing, stimulation can be a gate way opener for audience attention, interest and preference. Although it is difficult to directly explain the success of pop music, but we expected to be able to infer the preference of pop music through the analysis of brain responses. Understanding what kind of music the young consumers prefers and knowing their brain responses that are linked to preference is central to developing a pop music marketing strategy in entertainment business. Most firms have underestimated the sensory aspect of marketing, but that is changing as firms begin to recognize its power (Krishna, 2011). Sensory marketing engages consumers’ senses in an attempt to change their purchasing

1144

APJML 32,5

behavior in a positive direction (Krishna, 2011). Many consumers love products that engage their senses. The current study shows that the attunement of consumers’ brains to popular music can be indexed by the decreased randomness (and increased regularity) of the brain’s signals. An important lesson for marketing managers is the importance of determining whether their marketing uses OSL, as way to measure what will capture customers’ attention in a cluttered media environment.

Music’ effect is well-studied in traditional media context, such as radio, TV and physical retail settings. Previous literature on media advertisements has identified the role of background music in affecting content recall, increasing purchase intention and changing attitudes toward the brand and advertisement itself. The literature also suggested that there must be “fit” between the ad content and song content, or fit between the product and ad contents. Previously, music was treated as an executional ad element being embedded in traditional media delivery. What the literature did not tell us is what is the new role of music in the online network environment. For the young generation, music now is not a mere background factor, but it is the focal experiential product. The main contribution of the current study is increased attention to the fit between musical features as they related to the young listener. Theoretical implication of this research is our new findings on the importance of OSL of marketing stimuli in eliciting meaningful brain activations of young generation consumers. This study shows that certain factors when used properly may stimulate the young brain and elicit positive responses from those who are immersed in the digital life.

The ability to entrain the sensori-motor attention to external voices and actions is a cognitive capacity that has had significant implications for the success of humans as a social species (Merker, 2000). The reasons that individuals can withstand more intense forms of sensory stimulation in cyberspace than they can in physical space are not clear, although the phenomenon could be related to cyberspace’s being a mediated environment in which the laws of physics, biology and physiology are replaced with scale-free subjective, perceptual and virtual reality (Lee and Schumann, 2009). For digital marketing, managers of online digital media must learn how to enrich customers’ sensory experiences in an environment that is inherently devoid of stimuli to all but two of the human senses.

The importance of understanding customers’ sensory experiences will grow in the age of digital marketing, aided by artificial intelligence (AI). AI marketing uses customer data to anticipate a customer’s next move and prepare for the customer’s journey. By measuring customers’ brain responses with a neuro device like NIRS, AI can learn a customer’s musical taste and preferences without interrupting his or her perceptual experience. With the aid of neuromarketing data, AI can provide more intelligent search results and more refined content delivery that can heighten customer satisfaction using product-related experiences.

Limitations and suggestions for future research. The human brain is a highly efficient and precise processor of sensory information and it is possible to explore the nature of market offerings as sensory stimuli that are quantifiable in accurate physical scales. More research is needed to establish the science of neuromarketing as a cause-and-effect paradigm, beyond the current practice of identifying neural responses. Because experimental results obtained from artificial stimuli cannot provide meaningful implications for marketing practice, the current research categorizes songs on the dimensions of tempo and frequency using real-world marketing stimuli for laboratory experiments instead of stripped-down versions. In live music, tempo and frequency are highly related dimensions that combine with other stimuli to contribute to the experience of perceptual saturation. Therefore, one limitation of our research concerns our focus on tempo- and frequency-related factors and the lack of analysis of other parameters that may be equally important to auditory attention. Research is also needed that will identify the interactive brain processes for other elements of sound. Future research also needs to investigate how the rapid development of digital technologies, such as virtual and

1145

Pop music on Youtube

augmented realities, will require a new media marketing strategy that can provide a sufficient level of sensory stimulation by optimally recruiting both audio and visual neural pathways of young consumers and create a high impact on the market.

References

Abdelnour, F., Voss, H.U. and Raj, A. (2014), “Network diffusion accurately models the relationship between structural and functional brain connectivity networks”, Neuroimage, Vol. 90, pp. 335-347.

Aguiar, L. and Martens, B. (2016), “Digital music consumption on the internet: evidence from clickstream data”, Information Economics and Policy, Vol. 34, pp. 27-43.

Benchenane, K., Peyrache, A., Khamassi, M., Tierney, P.L., Gioanni, Y., Battaglia, F.P. and Wiener, S.I. (2010), “Coherent theta oscillations and reorganization of spike timing in the hippocampal- prefrontal network upon Learning”, Neuron, Vol. 66 No. 6, pp. 921-936.

Bennett, A. and Peterson, R.A. (2004), Music Scenes: Local, Translocal and Virtual, Vanderbilt University Press.

Bhattacharya, J. and Petsche, H. (2001), “Musicians and the gamma band: a secret affair?”, NeuroReport, Vol. 12 No. 2, pp. 371-374.

Bockstedt, J.C., Kauffman, R.J. and Riggins, F.J. (2006), “The move to artist-led pnline music distribution: a theory-based assessment and prospects for structural changes in the digital music market”, International Journal of Electronic Commerce, Vol. 10 No. 3, pp. 7-38.

Bolton, T.L. (1894), “Rhythm”, The American Journal of Psychology, Vol. 6 No. 2, pp. 145-238. Buzsáki, G. (2004), “Large-scale recording of neuronal ensembles”, Nature Neuroscience, Vol. 7 No. 5,

pp. 446-451. Chaisanguanthum, K.S., Shen, H.H. and Sabes, P.N. (2014), “Motor variability arises from a slow

random walk in neural state”, The Journal of Neuroscience, Vol. 34 No. 36, pp. 12071-12080. Chiang, C.-W., Wang, Y., Sun, P., Lin, T.-H., Trinkaus, K., Cross, A.H. and Song, S.-K. (2014),

“Quantifying white matter tract diffusion parameters in the presence of increased extra-fiber cellularity and vasogenic edema”, Neuroimage, Vol. 101, pp. 310-319.

Cui, X., Bray, S., Bryant, D.M., Glover, G.H. and Reiss, A.L. (2011), “A quantitative comparison of NIRS and fMRI across multiple cognitive tasks”, NeuroImage, Vol. 54 No. 4, pp. 2808-2821.

Cutnell, J.D. and Johnson, K.W. (1994), Physics, Wiley, New York, NY. Dickey, D.A. and Fuller, W.A. (1979), “Distribution of the estimators for autoregressive time series with

a unit root”, Journal of the American Statistical Association, Vol. 74 No. 366a, pp. 427-431. Dunn, P.G., de Ruyter, B. and Bouwhuis, D.G. (2012), “Toward a better understanding of the relation

between music preference, listening behavior, and personality”, Psychology of Music, Vol. 40 No. 4, pp. 411-428, available at: https://doi.org/10.1177/0305735610388897

Fujioka, T., Trainor, L.J., Large, E.W. and Ross, B. (2009), “Beta and gamma rhythms in human auditory cortex during musical beat processing”, Annals of the New York Academy of Sciences, Vol. 1169 No. 1, pp. 89-92.

Gerstein, G.L. and Mandelbrot, B. (1964), “Random walk models for the spike activity of a single neuron”, Biophysical Journal, Vol. 4 No. 1, pp. 41-68.

Haber, S.N. and Knutson, B. (2010), “The reward circuit: linking primate anatomy and human imaging”, Neuropsychopharmacology, Vol. 35 No. 1, pp. 4-26.

Herzmann, G., Jin, M., Cordes, D. and Curran, T. (2012), “A within-subject ERP and fMRI investigation of orientation-specific recognition memory for pictures”, Cognitive Neuroscience, Vol. 3 Nos 3-4, pp. 174-192.

Hoshi, Y. and Tamura, M. (1993), “Dynamic multichannel near-infrared optical imaging of human brain activity”, Journal of Applied Physiology, Vol. 75 No. 4, pp. 1842-1846.

Jensen, O., Kaiser, J. and Lachaux, J.-P. (2007), “Human gamma-frequency oscillations associated with attention and memory”, Trends in Neurosciences, Vol. 30 No. 7, pp. 317-324.

Jerri, A.J. (1977), “The Shannon sampling theorem – its various extensions and applications: a tutorial review”, Proceedings of the IEEE, Vol. 65 No. 11, pp. 1565-1596.

1146

APJML 32,5

https://doi.org/10.1177/0305735610388897

Kaiser, J., Lutzenberger, W., Preissl, H., Ackermann, H. and Birbaumer, N. (2000), “Right-hemisphere dominance for the processing of sound-source lateralization”, The Journal of Neuroscience, Vol. 20 No. 17, pp. 6631-6639.

Kellaris, J.J. and Kent, R.J. (1991), “Exploring tempo and modality effects, on consumer responses to music”, in Holman, R.H. and Solomon, M.R. (Eds), Advances in Consumer Research, Vol. 18, Association for Consumer Research, Provo, UT, pp. 243-248.

Kipiński, L., König, R., Sielużycki, C. and Kordecki, W. (2011), “Application of modern tests for stationarity to single-trial MEG data”, Biological Cybernetics, Vol. 105 Nos 3-4, pp. 183-195.

Kocsis, L., Herman, P. and Eke, A. (2006), “The modified Beer–Lambert law revisited”, Physics in Medicine and Biology, Vol. 51 No. 5, p. N91.

Kotilahti, K., Nissilä, I., Näsi, T., Lipiäinen, L., Noponen, T., Meriläinen, P., Huotilainen, M. and Fellman, V. (2010), “Hemodynamic responses to speech and music in newborn infants”, Human Brain Mapping, Vol. 31 No. 4, pp. 595-603.

Krishna, A. (2011), Sensory Marketing: Research on the Sensuality of Products, Routledge.

Larsen, J.P. (2009), The Music Industry and Digital Music: Disruptive Technology and the Value Network Effects on Industry Incumbents, Copenhagen Business School, Copenhagen.

Lee, E.-J. and Schumann, D.W. (2009), “Proposing and testing the contextual gender influence theory: an examination of gender influence types on trust of computer agents”, Journal of Consumer Psychology, Vol. 19 No. 3, pp. 440-450.

London, J. (2012),Hearing in Time: Psychological Aspects of Musical Meter, Oxford University Press, Oxford.

Lopes, P.D. (1992), “Innovation and Diversity in the popular music industry, 1969 to 1999”, American Sociological Review, Vol. 57 No. 1, pp. 56-71.

McAdams, S. and Bregman, A. (1979), “Hearing musical streams”, Computer Music Journal, pp. 26-60.

McAuley, J.D., Jones, M.R., Holub, S., Johnston, H.M. and Miller, N.S. (2006), “The time of our lives: life span development of timing and event tracking”, Journal of Experimental Psychology: General, Vol. 135 No. 3, p. 348.

Maylor, E.A. and Lavie, N. (1998), “The influence of perceptual load on age differences in selective attention”, Psychology and Aging, Vol. 13 No. 4, pp. 563-573.

Merker, B. (2000), “Synchronous chorusing and numan origins”, The Origins of Music, MIT Press, Cambridge, MA.

Meyerding, S.G.H. and Mehlhose, C.M. (2018), “Can neuromarketing add value to the traditional marketing research? An exemplary experiment with functional near-infrared spectroscopy (fNIRS)”.

Mitterschiffthaler, M.T., Fu, Cynthia H.Y., Dalton, J.A., Andrew, C.M. and Williams, Steven C.R. (2007), “A functional MRI study of happy and sad affective states induced by classical music”, Human Brain Mapping, Vol. 28 No. 11, pp. 1150-1162.

Modzelewski, R., Lepretre, S., Martinaud, O., Hannequin, D., Hitzel, A., Habert, M.-O., Tilly, H. and Vera, P. (2011), “Brain perfusion in adult patients with acute myeloblastic leukemia before and after cytosine arabinoside”, Molecular Imaging and Biology, Vol. 13 No. 4, pp. 747-753.

Nagamitsu, S., Nagano, M., Yamashita, Y., Takashima, S. and Matsuishi, T. (2006), “Prefrontal cerebral blood volume patterns while playing video games – a near-infrared spectroscopy study”, Brain and Development, Vol. 28 No. 5, pp. 315-321.

Nuttall, P., Sally, A., Carless, L., Crockford, L., Finnamore, K., Frazier, R. and Hill, A. (2011), “Understanding music consumption through a tribal lens”, Journal of Retailing and Consumer Services, Vol. 18 No. 2, pp. 152-159.

Olson, H.F. (1967), Music, Physics and Engineering, Courier Corporation.

Park, C.W. and Young, S.M. (1986), “Consumer response to television commercials: the impact of involvement and background music on brand attitude”, Journal of Marketing Research, Vol. 23 No. 1, pp. 11-24, available at: https://doi.org/10.1177/002224378602300102

Peterson, R.A. and Anand, N. (2004), “The production of culture perspective”, Annual Review of Sociology, Vol. 30, pp. 311-334.

1147

Pop music on Youtube

https://doi.org/10.1177/002224378602300102

Peterson, R.A. and Berger, D.G. (1975), “Cycles in symbol production: the case of popular Music”, American Sociological Review, Vol. 40 No. 2, pp. 158-173.

Ramsøy, T.Z. (2015), Introduction to Neuromarketing & Consumer Neuroscience, Neurons Inc.

Sakatani, K., Chen, S., Lichty, W., Zuo, H. and Wang, Y.-P. (1999), “Cerebral blood oxygenation changes induced by auditory stimulation in newborn infants measured by near infrared spectroscopy”, Early Human Development, Vol. 55 No. 3, pp. 229-236.

Simen, P., Balci, F., Cohen, J.D. and Holmes, P. (2011), “A model of interval timing by neural integration”, The Journal of Neuroscience, Vol. 31 No. 25, pp. 9238-9253.

Smallwood, J. (2013), “Distinguishing how from why the mind wanders: a process–occurrence framework for self-generated mental activity”, Psychological Bulletin, Vol. 139 No. 3, p. 519.

Smith, A.N., Fischer, E. and Yongjian, C. (2012), “How does brand-related user-generated content differ across YouTube, Facebook, and Twitter?”, Journal of Interactive Marketing, Vol. 26 No. 2, pp. 102-113.

Steenkamp, J.-B.E. and Baumgartner, H. (1992), “The role of optimum stimulation level in exploratory consumer behavior”, Journal of Consumer Research, Vol. 19 No. 3, pp. 434-448.

Tom, G. (1990), “Marketing with music”, Journal of Consumer Marketing, Vol. 7 No. 2, pp. 49-53, available at: https://doi.org/10.1108/EUM0000000002576

Verrusio, W., Ettorre, E., Vicenzini, E., Vanacore, N., Cacciafesta, M. and Mecarelli, O. (2015), “The mozart effect: a quantitative EEG study”, Consciousness and Cognition, Vol. 35, pp. 150-155.

Westergaard, P. (1975), An Introduction to Tonal Theory, Norton, New York, NY.

Wilkins, A.S., Wrangham, R.W. and Fitch, W.T. (2014), “The ‘Domestication Syndrome’ in mammals: a unified explanation based on neural crest cell behavior and genetics”, GENETICS, Vol. 197 No. 3, pp. 795-808.

Wlömert, N. and Papies, D. (2016), “On-demand streaming services and music industry revenues— insights from spotify’s market entry”, International Journal of Research in Marketing, Vol. 33 No. 2, pp. 314-327, available at: https://doi.org/10.1016/j.ijresmar.2015.11.002

Wyatt, J., Cope, M., Delpy, D., Richardson, C., Edwards, A., Wray, S. and Reynolds, E. (1990), “Quantitation of cerebral blood volume in human infants by near-infrared spectroscopy”, Journal of Applied Physiology, Vol. 68 No. 3, pp. 1086-1091.

Yoon, C., Angela, H., Gutchess, F.F. and Polk, T.A. (2006), “A functional magnetic resonance imaging study of neural dissociations between brand and person judgments”, Journal of Consumer Research, Vol. 33 No. 1, pp. 31-40.

Zaramella, P., Freato, F., Amigoni, A., Salvadori, S., Marangoni, P., Suppjei, A., Schiavo, B. and Chiandetti, L. (2001), “Brain auditory activation measured by near-infrared spectroscopy (NIRS) in neonates”, Pediatric Research, Vol. 49 No. 2, pp. 213-219.