The Prospects of Using EEG in Tourism and Hospitality Research

EEG themes/theories/concepts

There are probably five aspects that the neuroscientific approach, including EEG, can be applied to in order to improve the understanding of marketing theories and consumer behavior (Plassmann et al., 2015). The first is to identify the underlying mechanism of the cognition and emotional processes, which can help better understand associated consumer behaviors. For example, Lee et al. (2014) found that frontal theta EEG activation is a neural indicator of green consumers’ cognitive engagement with environmentally friendly product messages. Second, neuroscientific techniques can be applied to measure implicit processes, which it is hard capture by using traditional approaches. When social context, such as the presence of others, was presented, passive viewing of luxury branded products induced greater late positive potential amplitude measured by EEG than viewing basic branded products (Pozharliev et al., 2015). The third is to dissociate between psychological processes. Pozharliev et al. (2015) identified that being alone or with another person when viewing luxury versus basic branded products involved different neural processes for the P2 and P3 components over visual cortex sites. Fourth, individual differences in consumer behavior and the source of heterogeneity can be evaluated by using neuroscientific methods. Balthazard et al. (2012) found significant correlations between transformational leadership and EEG variables, including amplitude asymmetry, coherence, and phase lock duration, based on which transformational and nontransformational leaders can be classified. Fifth, neuroscience measures can improve predictions regarding consumer behavior. Neural similarity calculated by alpha oscillations is a predictor of recall of movie trailers and population-level sales (Barnett & Cerf, 2017). It should be noted that one study may involve more than one aspect.

The behavioral inhibition system/behavioral activation system (BIS/BAS) theory has been widely applied in EEG studies. Introduced by Gray (1982), the BIS/BAS theory is also known as reinforcement sensitivity theory in psychology. The BIS refers to avoidance or withdrawal behavior when responding to threats, which may lead to punishment or novelty (van Zeeland & Henseler, 2018). The BAS is associated with motivation in responding to rewarding incentives. Davidson et al. (1979) proposed a theoretical framework that links emotions with EEG measures: The left frontal area of the brain measures positive or approach-related emotions, and the right frontal area measures negative or withdrawal-related emotions. According to Davidson’s (2003) model, EEG can measure hemispheric activity in the prefrontal cortex. If the activity of the left hemisphere is greater than that of the right hemisphere, it leads to higher BAS scores. Conversely, if the activity of the right hemisphere is greater than that of the left hemisphere, it leads to higher BIS scores (Sutton & Davidson, 1997). Emotion regulation refers to conscious and unconscious processes when negative or positive emotions are increased or decreased (Gross, 1998). Research on emotion regulation has shown that emotion-focused coping can reduce the experience of negative emotions by avoiding negative information (Mogg et al., 2004), which has also been observed in EEG studies (e.g., Denson et al., 2012).

A number of studies have used EEG to study emotion. Emotion can be defined as “any brief conscious experience that intense mental activity and an elevated level of pleasure or displeasure characterises” (Gordon et al., 2018, p. 96). Emotion is considered to be one of the main factors that violates the “rational man” rule of traditional economics, as it is likely to affect consumers’ decision-making processes (LaBarbera & Tucciarone, 1995). Thus, it is important to capture the emotional responses in marketing research. It is widely accepted that brain activity patterns are closely linked to consumers’ cognition and behavior (Matukin et al., 2016). Therefore, using neurophysiological measurement methods, such as EEG, can capture more objectively the psychophysiological responses of consumers that they cannot control or manipulate (Burns et al., 2010). The concept of affect is associated with emotion as affect can be seen as the outward expression of an emotion, which consists of two dimensions, valence, and arousal (Venkatraman et al., 2015). Venkatraman et al. (2015) found that higher values of frontal asymmetry are related to affect in response to more effective ads.

Attention, which refers to people’s ability to focus on certain parts of the environment or objects, has also been widely studied using EEG (Venkatraman et al., 2015). In marketing research, attention has been used to study attitudinal and behavioral outcomes in consumers (Pieters & Warlop, 1999). By recording the EEG data associated with attention to visual stimuli, Barnett and Cerf (2017) examined whether cross-brain correlation can predict movie trailer recall and future ticket sales. They found that neural similarity can be a predictor of free recall. Attention is also seen as a type of emotional state by Rosenbaum et al. (2019). They examined the effects of the presence of greenery in consumption settings on consumers’ brain activation data associated with six emotional states such as attention, excitement, and interest. Specifically, attention restoration theory was applied to explore the effects of the incorporation of biophilic design elements in consumption settings (i.e., shopping malls) on reducing stress.

Another concept that links to emotion is empathy, which reflects people’s social nature and includes cognitive and emotional components. It can be defined as the ability for people to feel others’ emotions, which enables them to predict others’ behavior and to motivate altruistic behavior (Smith, 2006). It is believed that empathy relates to the neural mechanism that is involved in altruistic prosocial behavior (Lee, 2016). Using EEG, it has been shown that the neural mechanism of customer empathy plays a key role in responding to corporate social responsibility messages, in terms of reducing price sensitivity to prosocial products and increasing consumers’ willingness to pay for these products (Lee, 2016).

EEG has also been widely applied to memory theories. There are two types of memory processes: long-term memory and working memory. Long-term memory “stores an individual’s fundamental knowledge organised into central concepts that have features and linkages to related concepts” (Minas et al., 2018, p. 791). Working memory as short-term memory retrieves knowledge from long-term memory and applies it to a new domain through deliberate conscious cognition (Minas et al., 2018), by processing the results of idea generation guiding behavior in the very near future (Lee et al., 2014). Many of the studies reviewed linked working memory with attention, suggesting that consumer behavior can be influenced when they pay continuous attention to marketing stimuli and transfer their working memory to long-term memory (Gordon et al., 2018). More recent research has set out to explore various theories/concepts in one study. For example, Laaksonen et al. (2019) evaluated the effects of media brand knowledge on not only consumers’ emotional and attentional responses to but also their memory of news messages.

Other less used concepts considered to be related to neural activity include preferences and self-control. The frontal asymmetry theory has also been used to explain the relationship between hemispheric asymmetry in EEG signals and preferences/choice behavior (Sutton & Davidson, 2000). EEG activity has been found to be related to a trait of preference (Telpaz et al., 2015). Self-control has been defined as “resistance to temptation” in the neuroscience literature (Lopez et al., 2014). It is believed that self-control is connected with the prefrontal cortex, in particular its right region (Hu et al., 2015).

EEG methods and indices

Most of the reviewed articles used a mixed-methods approach, which can overcome the weakness of each method and thus generate more reliable outcomes. Also, different methods are able to process various information, which can generate rich and comprehensive research results compared with single method research (Gountas et al., 2019). Three types of mixed approaches are identified.

First, EEG data can be integrated with traditional self-reported approach including quantitative and/or qualitative data to explore people’s cognitive and emotional responses when expressing attitudes toward marketing or management issues (Clark et al., 2018). For example, the results of EEG and a postexperiment survey were applied, both showing that attractive apparel products triggered positive emotions, while unattractive apparel products triggered negative emotions (Touchette & Lee, 2017). Traditional methods that capture subjective attitudes and opinions are useful complements for psychophysiological methods that explore arousal, affective valence, cognitive effort, and motivation associated with brain activation (Clark et al., 2018). For example, Gountas et al. (2019) evaluated the effects of an anti-binge-drinking campaign and how young people react to binge-drinking. They discovered that focus group discussions were able to capture consumers’ deeper opinions regarding the messages while EEG was capable of identifying cognitive and affective engagement.

The second type of mixed methods involve using EEG with other psychophysiological sensors. As a consumer neuroscience measurement tool, EEG can be used with physiological metrics such as heart rate and skin conductance to evaluate levels of emotion and arousal (Stikic et al., 2014). The results generated by each method can be compared with those measured with other methods, thus increasing the validity of the study. Moreover, EEG research has generally been supplemented by eye tracking and facial expression measurements (Clark et al., 2018). Eye tracking, which measures participants’ fixation and period of gazing, can be used to identify objects that evoke participants’ emotional responses (Kamienkowski et al., 2012; Matukin et al., 2016). Real-time facial expressions can help determine emotional valence related to motivation (Ohme et al., 2009).

The third type of mixed approach refers to the use of EEG, other psychophysiological sensors and traditional self-report methods (e.g., Clark et al., 2018; Laaksonen et al., 2019). Boshoff’s (2017), for example, used both neurophysiological measures (EEG and Galvanic Skin Response) to measure subconscious responses and a traditional pencil-and-paper experiment to measure conscious responses to the attractiveness of frontline employees in a service failure and recovery situation. The results showed that while the answers were neutral at the conscious level, the responses varied depending on the attractiveness of the service provider at the subconscious level. This implies that using EEG to measure neural data can avoid effort and conscious reporting biases. Figure 1 summarized the number of articles retrieved using various EEG approaches, and showed that EEG with quantitative methods is the most commonly used combination.

OPEN IN VIEWER

Figure 1

The Number of Articles Using Various Electroencephalography (EEG) Approaches

EEG is ‘a neurophysiological measurement of postsynaptic electrical potentials on the surface of the scalp’ (Minas et al., 2018, p. 796). Electrodes collect the summation of the synchronized activity, which is compared either with a reference electrode or a common average reference (Harmon-Jones & Peterson, 2009). Oscillations of brain activity collected by electrodes are “complex waveforms that can be decomposed into simple waveforms of different periodicity at varying amplitudes” (Minas et al., 2018, p. 796). There are two basic analytical approaches to analyze EEG signals: time domain analysis and frequency domain analysis. Time domain analysis is mainly used to analyze the geometric characteristics of EEG waveforms, such as amplitude, mean, variance, skewness, and kurtosis, while frequency domain analysis can be used to estimate all frequency-related features when a signal is represented by its frequency component (Harpale & Bairagi, 2016). The two analytical approaches will be further discussed in the following paragraphs.

First, the event-related potential (ERP) method, a useful derivative of EEG as a time domain analysis method, can be applied to evaluate brain activity that is time-locked to a stimulus presentation (Deitz et al., 2016). ERP measures “the changes in the voltage level in response to a stimulus presented as a function of time” (Telpaz et al., 2015, p. 513). The two most frequently used ERP components are the P300 wave with P, meaning a positive deflection in the scalp, and 300, meaning the deflection potential appearing 300 ms after the stimulus; and similarly the N200 wave, indicates a negative deflection starting after 200 ms since stimulus has been presented (Telpaz et al., 2015). The advantage of ERP is that it can capture rapid processing brain activity and investigate relevant cognitive and affective processes (Luck, 2005). A limitation of ERPs is that the signal-to-noise ratio is low, which implies that the results can be largely affected by “noises,” that is, task-unrelated brain activity (Lin et al., 2018). Therefore, it is crucial to include a sufficient number of trials through presenting iterations of stimuli in order to improve the results.

Regarding the second analytical approach—frequency domain analysis, EEG studies have usually focused on five frequency bands: delta (<4 Hz), theta (4-8 Hz), alpha (8-13 Hz), beta (13-20 Hz) and gamma (>20Hz; Harmon-Jones & Peterson, 2009). Among these, the alpha band has received a lot of attention because of its cortical “idling” process (Minas et al., 2018). It is believed that the alpha band is related to relaxation, contemplation, internal focus, and mental performance (Hannah et al., 2013). Synchronization occurs when a brain region is at rest, while desynchronization occurs when a brain region is activated, reducing the alpha frequency (Potter & Bolls, 2012). Studies on attention and working memory have shown that the alpha band is negatively associated with intensive attention and working memory resources inputting into completing a task (Guo et al., 2018). Gamma power has also been found to be positively related to memory encoding (Guo et al., 2018). The theta band has been linked with the inhibition of elicited responses and the beta band with alertness (Telpaz et al., 2015). However, it should be noted that because each band is related to different cognitive processes, caution should be exercised when judging a mental process by observing changes in a frequency band (Poldrack, 2006).

The cerebral cortex of the brain, which contains four lobes, the frontal, the parietal, the occipital and the temporal lobes, “controls higher order functions, such as planning, organizing and decision-making” (Touchette & Lee, 2017, p. 6). The frontal lobe has the largest number of neurons that “control the pleasure chemical dopamine, which has a crucial effect on emotion” (Touchette & Lee, 2017, p. 6), and thus attracts the most attention from neuromarketing. The two parts of the frontal lobe, the right and left hemispheres, show a lateralization relationship, which has been described by Richard Davidson’s theory of frontal asymmetry (Morin, 2011). Davidson’s (1995) theory suggests that the left frontal area is associated with positive emotions, related to approach motivation, and the right frontal area is connected to negative emotions, related to withdrawal motivation. Among the different frequency bands in frontal asymmetry research, compared with the other bands, the alpha band was used by the majority of the reviewed articles, as it can show the respondents’ attentiveness (Touchette & Lee, 2017). For example, Minas et al. (2018) showed that an achievement prime led to a greater alpha wave desynchronization compared with a neutral prime, suggesting an increase in cognitive activity in the right frontal areas.

Lee (2016) evaluated the neural correlates of empathy, focusing on the anterior cingulate cortex (ACC). The ACC, a part of the brain’s frontal and limbic connectivity, is connected to various affective responses believed to be associated with empathy (Lloyd et al., 2004). By analyzing the theta band (4-8 Hz), Lee (2016) found an increase in brain activation in the ACC among consumers choosing a prosocial product in response to a corporate social responsibility–related message. Boksem and Smidts (2015) showed that medial frontal beta power was positively related to individual preferences: the higher the beta power when watching the movie trailer, the higher the participants ranked that trailer. In addition to analyzing the power of different bands, such as the alpha and beta bands, the articles reviewed applied ERP analysis. For example, it was shown that different N200 components were presented when comparing highly and less preferred goods, while the P300 component was not significant (Telpaz et al., 2015). Pozharliev et al. (2015) also found an emotional effect in luxury brand product pictures based on the parietal distribution of late positive potentials, demonstrating strong positivity.

In EEG-based neuromarketing experiments, stimuli are developed to reflect changes in brain patterns and indices relevant to the research aims. The EEG stimuli used in the retrieved articles can be divided into three categories: visual materials, scenario-based materials and at-rest assessments. Eighteen articles used visual materials, including videos (e.g., Guo et al., 2018; Boshoff, 2017), pictures (e.g., Minas et al., 2018; Touchette & Lee, 2017) and websites (e.g., Clark et al., 2018; Gregor et al., 2014). Twelve articles used scenario-based materials as stimuli (e.g., Hariharan et al., 2016; Lee, 2016; M. Li et al., 2014). Finally, three articles conducted at-rest assessments using an eyes-closed resting (but alert) procedure (e.g., Hannah et al., 2013; Waldman et al., 2017).

Literature applying EEG in tourism and hospitality research

Although EEG has been used in the business and management field over the past decades, its application in tourism and hospitality research is scarce with only one article identified. Bastiaansen et al. (2018)’s EEG experiments found that watching movies related to the destination images induced a positive emotional response to these destination pictures. While shed some light on the future research, Bastiaansen et al. (2018) used the EEG method alone, so it was not possible to further assess the correlation between tourists’ emotional responses to the destination pictures and their intention to visit or purchase behavior. Second, it used a between-subjects design, which may lead to error variance in the between-subjects effects due to individual differences, criticized in most EEG-based experiments.