The application of mobile functional near-infrared spectroscopy for marketing research – a guideline

Caspar Krampe (Department of Social Science, Marketing and Consumer Behaviour Group, Wageningen University and Research, Wageningen, The Netherlands)

European Journal of Marketing

ISSN: 0309-0566

Article publication date: 16 August 2022

Issue publication date: 19 December 2022

1233

Abstract

Purpose

To advance marketing research and practice, this study aims to examine the application of the innovative, mobile-applicable neuroimaging method – mobile functional near-infrared spectroscopy (mfNIRS) – in the field of marketing research, providing comprehensive guidelines and practical recommendations.

Design/methodology/approach

A general review and investigation of when and how to use mfNIRS in business-to-consumer and business-to-business marketing settings is used to illustrate the utility of mfNIRS.

Findings

The research findings help prospective marketing and consumer neuroscience researchers to structure mfNIRS experiments, perform the analysis and interpret the obtained mfNIRS data.

Research implications

The application of mfNIRS offers opportunities for marketing research that allow the exploration of neural processes and associated behaviour of customers in naturalistic settings.

Practical implications

The application of mfNIRS as a neuroimaging method enables the investigation of unconscious neural processes that control customer behaviour and can act as process variables for companies.

Originality/value

This is one of the first studies to provide comprehensive guidelines and applied practical recommendations concerning when and how to apply mfNIRS in marketing research.

Keywords

Citation

Krampe, C. (2022), "The application of mobile functional near-infrared spectroscopy for marketing research – a guideline", European Journal of Marketing, Vol. 56 No. 13, pp. 236-260. https://doi.org/10.1108/EJM-01-2021-0003

Publisher

:

Emerald Publishing Limited

Copyright © 2022, Caspar Krampe.

License

Published by Emerald Publishing Limited. This article is published under the Creative Commons Attribution (CC BY 4.0) licence. Anyone may reproduce, distribute, translate and create derivative works of this article (for both commercial and non-commercial purposes), subject to full attribution to the original publication and authors. The full terms of this licence may be seen at http://creativecommons.org/licences/by/4.0/legalcode


1. Introduction

To facilitate progress in the use of neuropsychological methods in marketing research (Kannan, 2017; Shaw and Bagozzi, 2018), this article focusses on the mobile-applicable version of functional near-infrared spectroscopy (mfNIRS) and its use in marketing research. mfNIRS, which uses near-infrared (NIR) light to quantify brain activity, is suggested to be capable of extending knowledge beyond the use of existing marketing methods (Gefen, 2014; Krampe et al., 2018a; Krampe et al., 2018b; Gier et al., 2020), whilst providing insights concerning unconscious neural processes.

The method offers advantages over other neuroimaging methods. mfNIRS is, for example, less expensive than the “gold standard” in neuroimaging – functional magnetic resonance imaging (fMRI) – and, most importantly, it is mobile-applicable, which allows the investigation of neural signatures in naturalistic environments. By demonstrating the applicability of mfNIRS, this article highlights new directions for marketing and consumer neuroscience research and demonstrates how marketing researchers from various sub-fields can apply mfNIRS in their work. To this end, the article provides comprehensive guidelines, practical recommendations of when and how to conduct mfNIRS experiments, in addition to explaining how to collect, analyse and interpret mfNIRS data in marketing and consumer neuroscience research.

The article begins with a brief overview of the fields of consumer neuroscience and neuromarketing and addresses the questions “Where are we?” and “Where are we going?”, looking at both the business-to-consumer (B2C) and business-to-business (B2B) marketing environment. Accordingly, the most common neuroimaging methods applied in marketing research and cognate disciplines are described, followed by an explanation of the basic principles, technical parameters and functionalities of mfNIRS. A particular focus is on the methodological questions of when and how to conduct mfNIRS experiments as well as how to analyse data, draw possible statistical conclusions and interpret neural results. Lastly, a status-quo analysis of the current integration of the mentioned guidelines is conducted through a systematic literature review of recent fNIRS/mfNIRS studies in the field of marketing research, leading to practical-normative implications and suggestions for future research.

2. Marketing and neuroscience: where we are and where we are going?

Over the past two decades, consumer neuroscience and neuromarketing research has gained credibility and a degree of acceptance in the broader marketing research community (Harris et al., 2018; Lim, 2018; Lee et al., 2018). This development is primarily driven by the belief that the use of neuroimaging methods will help to gain deeper insights into many types of consumer/customer behaviour and processes in a marketing-relevant context, offering the potential to further advance existing theory and practice (Kenning and Plassmann, 2005; Lim, 2018; Plassmann et al., 2015; Zaltman, 2000). Although neuromarketing and consumer neuroscience are predicted to have a bright future (Mansor and Isa, 2020), the fledgling disciplines face several challenges.

The existing concerns that are associated with the usage of neuroscientific methods in marketing research can be seen as a first challenge (Plassmann et al., 2015). Consumer neuroscience and neuromarketing are often criticised for “only” measuring brain activity and not actual behaviour per se (Plassmann et al., 2015). In response to this concern, it should be evident that the fields of consumer neuroscience and neuromarketing and their associated methods are not intended to replace traditional marketing methods, which measure consumer/customer behaviour. Instead, they should be seen as complementary, with the goal of increasing knowledge about marketing-relevant entities from a neuroscientific perspective. Ideally, consumer neuroscience and neuromarketing research should inspire researchers to use multi-methodological approaches that combine behavioural and neuroscientific methods that capitalise on the strengths of each method (Plassmann et al., 2015).

Another widespread concern associated with the usage of neuroscientific methods relates to the observation that consumer neuroscience and neuromarketing relies primarily on backward inference to identify psychological mechanisms. Although this is problematic for any research that links neuroscience to behaviour, it is expected to be resolved by using a theory-driven approach to designing studies and/or by applying meta-analytical statistical tools to improve the interpretation of research findings.

The third concern identified by Plassmann et al. (2015) refers to the assumption that neuroimaging studies are less reliable and generalisable than traditional marketing studies. This assumption can be refuted based on previous research work demonstrating the possibility of using neuroscientific methods with small sample sizes of 14–29 participants to predict decision-making concerning the choice of products or services with a probability exceeding the level of chance within a population (Berns and Moore, 2012; Boksem and Smidts, 2015; Telpaz et al., 2015). Furthermore, given that the measured neuropsychological processes are based on biological determinants, they can be assumed to occur in similar ways in higher mammals (Seth et al., 2005). There are also statistical and methodological reasons to assume that the relatively small sample size of neuroscientific studies allows the generalisation of findings to the population level (Friston, 1996, 2004, 2012). This is primarily because, according to neuroscientific standards, the stimuli used to gain knowledge are always displayed repeatedly to account for the noise associated with neuroscientific measurements (Plassmann et al., 2015). The actual number of stimuli shown in a neuroscientific study does therefore not differ significantly from that shown in a between-subjects behavioural study, thus providing further confirmation of the statistical power of neuroscientific studies (Plassmann et al., 2015; Hensel et al., 2017).

A second challenge is the fact that previous research has focused almost exclusively on entities that can be attributed to B2C marketing, including research questions on how consumers respond to merchandising elements of in-store communication (Krampe et al., 2017; Cakir et al., 2018) or how neuroscience data can be used to predict future consumer behaviour (Knutson et al., 2007; Berns and Moore, 2012; Gier et al., 2020). Although the research findings provided valuable insights that advanced (consumer) theories and provided practical–normative solutions for companies operating in consumer markets, the exclusive focus on consumer neuroscience neglects the fact that neuroscientific approaches can also provide fruitful insights for marketing research in the broader field of B2B marketing (Lim, 2018). Hence, the advantages of neuroscientific methods should not be disregarded in the field of B2B marketing.

Finally, the third challenge relates to the limited ecological validity attributed to the mostly stationary neuroimaging methods used in consumer neuroscience and neuromarketing. An aspect that makes it particularly difficult to verify whether research results that have been studied under controlled laboratory conditions or in experiments remain valid in naturalistic contexts or environments.

While the first challenge and related concerns have been –at least partially– resolved in previous research, the second and third challenges remain. There is, consequently, a need for mobile-applicable neuroimaging methods in the field of marketing research that allow to close the ecological validity gap of existing neuroscientific methods and validate data gained from laboratory experiments with data from the real world within the B2C or B2B marketing environment (Boto et al., 2018; Gordon et al., 2018).

3. Neuroscientific methods applied in marketing research

In general, three groups of neuroscientific methods can be identified within the field of consumer neuroscience and neuromarketing (Lim, 2018). The first comprises methods that measure the effects of neural activity patterns (e.g. eye movements, pupil size variation, facial electromyography or skin conductance), and not actual neural activity. A second group consists of neuroimaging methods, which can be defined as methods that allow the production of in vivo images of the human brain (Dimoka et al., 2012). In essence, these methods can be subdivided into two groups: electromagnetic neuroimaging methods, which measure immediate neurophysiological responses (e.g. magnetoencephalography/MEG and electroencephalography/EEG), and metabolic neuroimaging methods, which are based on human metabolic processes (e.g. fMRI).

Other non-neuroimaging methods are also applied. These methods focus on the physiological effects of neural activity patterns, but do not measure neural activity patterns itself. These neuroscientific methods constitute their own methodological group, which is not further considered in this study (for additional information about these methods, Lim, 2018).

In contrast to electromagnetic neuroimaging methods, metabolic neuroimaging methods measure brain activity based on the premise that active brain regions need more energy and, consequently, more oxygenated blood, which is used as a proxy for quantifying neural activity according to physiological characteristics (Kwong et al., 1992). Indirect neuroimaging methods are thus based on a correlative approach. Another distinguishing feature of neuroimaging methods has to do with the prevalence of its usage, which depends primarily on the research focus, technical capabilities and the functionality of the specific method. The final group of neuroscientific methods is related to those that can be used to manipulate neural activity patterns by interrupting the neural system (e.g. transcranial magnetic stimulation and the direct manipulation of neurotransmitter level within the brain) (Hallett, 2000).

The most common electromagnetic and metabolic neuroimaging methods are summarised in Table 1, along with their advantages and disadvantages in the context of marketing research.

As indicated before, given the limited set of mobile-applicable neuroimaging methods available, previous research indicated that mobile neuroimaging methods, such as mfNIRS, could potentially help to close the mentioned ecological validity gap of neuroimaging methods (Boto et al., 2018; Gordon et al., 2018) by answering research questions related to consumer neuroscience and neuromarketing (Kopton and Kenning, 2014; Krampe et al., 2018a, 2018b). mfNIRS might therefore provide a fruitful new avenue for investigating the neural processing and associated behaviour of consumers and customers in real-world B2C or B2B marketing settings.

4. Mobile functional near-infrared spectroscopy

More than 20 years ago, Jobsis (1977) demonstrated the possibility of using NIR light to detect changes in adult cortical oxygenation during hyperventilation (Ferrari and Quaresima, 2012; Kopton and Kenning, 2014). Based on his work and subsequent developments, fNIRS/mfNIRS has recently been applied in several research fields, including neuro-economics (Kopton and Kenning, 2014), shopper neuroscience (Krampe et al., 2018b; Gier et al., 2020), e-commerce (Nissen and Krampe, 2021) and neuro-organisational research (Quaresima and Ferrari, 2016), indicating its potential usage in the field of marketing. Although, recent studies suggest that mfNIRS is a suitable method for research in the fields of consumer neuroscience and shopper neuroscience (Krampe et al., 2017; Gier et al., 2020), its application in the field remains limited. Explanations, next to the identified common challenges associated with the use of neuroscientific methods, include the limited or fragmented literature related to mfNIRS in marketing research, the lack of high-quality, user-orientated methodological primers and the lack of clear guidelines concerning how to apply mfNIRS (Lee et al., 2018).

By addressing the three obstacles that may prevent marketing researchers from using mfNIRS, the current study aims to increase awareness amongst marketing researchers concerning the possibilities of mfNIRS in marketing research and beyond, to establish the fundamental principles of the neuroimaging method and to evaluate whether their research could benefit from the integration of mfNIRS. It, therefore, addresses (4.1) the basic operating principle of mfNIRS, (4.2) the technical parameters and commercially available mfNIRS systems, (4.3) the penetration depth of mfNIRS, and (4.4) potential confounding effects that might interfere with the fNIRS signal.

4.1 Basic operating principle of mobile functional near-infrared spectroscopy

NIR light – or, more precisely, light in a wavelength spectrum of 650–950 nm – has the advantage of being able to pass through biological tissue and non-invasively illuminate cortical tissue in the human brain (Funane et al., 2014; Jobsis, 1977; Kopton and Kenning, 2014; Krampe et al., 2018a, 2018b; McCormick et al., 1992; Scholkmann et al., 2014). The fNIRS method takes advantage of this characteristic, as parts of the emitted and scattered NIR light are absorbed or reflected by oxygenated and de-oxygenated haemoglobin, thereby allowing the indirect quantification of neural activity. More precisely, by using NIR light sources to emit NIR light photons, which always travel through tissue in a “banana shape” (Okada and Delpy, 2003) and by measuring the reflection with NIR-light detectors, this method allows the indirect measurement of brain activity within particular brain regions, defined ex ante (Figure 1).

4.2 Technical parameters and commercially available functional near-infrared spectroscopy/mobile functional near-infrared spectroscopy systems

Many fNIRS systems, which are based on specific techniques, are currently available. Some of the underlying techniques have similar characteristics and are associated with possible advantages and disadvantages. Most of the fNIRS/mfNIRS systems that are currently applied are based on the continuous wave (CW) method (Quaresima and Ferrari, 2016; Scholkmann et al., 2014), which is known to be cost-effective, wireless and portable (Kopton and Kenning, 2014). The CW-fNIRS system allows measurement of oxygenated, de-oxygenated and total haemoglobin concentrations, always in relation to a pre-defined reference value of zero (measured initially as a baseline).

A second system applied is the frequency domain (FD) method, which uses intensity-modulated light to illuminate the head, thus measuring attenuation and delay phases of the transmitted light (Quaresima and Ferrari, 2016; Scholkmann et al., 2014). The FD method could hypothetically be seen as a development of the CW-fNIRS method, as it can measure both light-intensity attenuation and phase shift (Fantini et al., 1999). According to the current research, however, there is no advantageous methodological distinction between the FD- and CW-fNIRS techniques (Davies et al., 2017).

A third system is the time domain (TD) technique, which uses short light pulses to illuminate the head and detect the shape of the pulse after propagation through the tissue (Quaresima and Ferrari, 2016). Compared to CW methods, TD techniques tend to have lower temporal resolution and higher costs, and they are unable to represent weak, functional differences in neural activity, due to the lower intensity of the NIR light (Scholkmann et al., 2014).

To decide which system to use, the benefits of the individual mfNIRS system compared to its costs as well as the individual research question need to be considered. More generally, however, relative to FD- and TD-fNIRS systems, CW-fNIRS systems offer the advantages of cost-effectiveness and mobile applicability – two important factors relating to the appropriateness of applying mfNIRS in naturalistic marketing-relevant settings, e.g. during a shopping situation or while exploring interactions processes between value chain actors. In turn, FD- and TD-fNIRS techniques offer the possibility of characterising the optical properties of tissue, thereby enabling the determination of changes in absolute oxygenated and de-oxygenated haemoglobin concentration. Nevertheless, these two techniques require more technical expertise to acquire and analyse data than it is the case for CW-fNIRS systems. Findings that might lead to the conclusion that CW-fNIRS systems are most suitable for marketing research and practice.

In addition to the technical parameters, fNIRS/mfNIRS systems vary in their complexity from smaller two-channel portable fNIRS devices (Adorni et al., 2018; Brugnera et al., 2016; Brugnera et al., 2017) to stationary “whole-head” arrays that integrate several dozen channels and cover the whole human cortex. Whereas these advanced “high-density 3-dimensional whole-head fNIRS system[s] (integrating more than 50 fibre optical bundles)” (Liao and Culver, 2014; Quaresima and Ferrari, 2016; Scholkmann et al., 2014) improve data quality by increasing spatial and temporal resolution, they have the disadvantages of requiring stationary usage, being limited in their flexibility and cable length, in addition to being uncomfortable for participants. As indicated before, to solve the ecological validity gap, this study focusses on battery-operated, multichannel, wireless, mfNIRS systems given that they allow measurements in real-world, naturalistic settings – an advantage that seems particularly important in applied marketing research, considering both B2C and B2B marketing research.

4.3 Penetration depth of near-infrared light

The penetration depth of NIR-light is important for marketing researchers to determine which brain regions and consequently which neuropsychological processes can be measured with mfNIRS. The penetration of NIR light depends, thereby, on the light scattering, absorption rate and separation distance between NIR light sources and detectors. As mentioned before, NIR light photons are known to travel in a “banana shape” through tissue, such that fNIRS/mfNIRS measurements are most sensitive to oxygenated/de-oxygenated haemoglobin molecules 1.5–2.5 cm below the skull. For the actual measurement of neural cortical activity, the penetration depth of fNIRS/mfNIRS is 1–2 cm (McCormick et al., 1992; Quaresima and Ferrari, 2016; Torricelli et al., 2014), whilst maintaining an optimal separation distance of approximately 3 cm between NIR light sources and detectors. The selection of the source-detector separation is dependent on the intensity of the NIR light (e.g. strength of sources) and the age and anatomical head properties of the participant (McCormick et al., 1992; Quaresima and Ferrari, 2016; Torricelli et al., 2014).

Given the penetration depth of on average 1–2 cm, mfNIRS allows marketing researchers to measure cortical brain regions that are inter alia associated with decision-making processes (Krampe et al., 2018b), including the dorsolateral prefrontal cortex (dlPFC), the orbitofrontal cortex (OFC) and parts of the cortical ventromedial prefrontal cortex (vmPFC). It does not allow measurement of subcortical brain regions that are inter alia associated with affective information processing (Dalgleish, 2004) such as the anterior cingulate cortex (ACC), insula, nucleus accumbent or amygdala (Ernst et al., 2013; Krampe, Gier et al., 2018; Krampe et al., 2018b). Researchers must therefore clarify in advance whether cortical brain regions and a marketing-relevant entity associated neuropsychological processes can be measured with mfNIRS. To this end, researchers can consolidate the neuroscientific open-source database www.neurosynth.org (Yarkoni et al., 2011) (or other databases: www.open-neuroscience.com/en/), which provides information about the localisation of neuropsychological processes within the brain in a three-dimensional coordinate system arranged according to three axes of the human head, based fMRI data: the x-axis (left to right), the y-axis (posterior to anterior) and the z-axis (bottom to top). Marketing researchers who are particularly interested in cognitive and affective neuropsychological processes (e.g. cognitive load or perception processes associated with shopping-relevant cues or negotiation talks between value chain actors) must therefore be very precise when using MNI coordinates to examine whether a neuropsychological process and the associated brain region are indeed located near the surface of the human skull and can therefore be explored with mfNIRS. In addition, most marketing-relevant neuropsychological processes (e.g. decision-making processes) are known to integrate several brain regions (Broche-Pérez et al., 2016) – a fact that the neurosynth.org database considers, indicating the connectivity of brain regions associated with a particular neuropsychological process or a brain region.

The “event-driven process chain” concerning when to apply mfNIRS can assist marketing researchers to decide whether mfNIRS is a valid and suitable neuroimaging method for investigate a marketing-relevant entity (Figure 2).

4.4 Potential confounding effects and possible solutions

Like most, if not all, scientific methods, mfNIRS systems must cope with several confounding influencing factors, which could have an impact on the measurement procedure and/or data quality. In particular, motion artefacts have been identified as influencing the collection – and therefore the quality – of mfNIRS data. In this regard, abrupt head and facial movements appear to be particularly influential when collecting mobile fNIRS data (Girouard et al., 2010). To prevent head movement artefacts, researchers must ensure that the headgear used has been correctly and stably placed on the participant’s forehead. To do so, several steps are needed.

To correctly place the NIR light sources and detectors that cover the whole cortex or parts of the cortex, the 10–20 system (Jasper, 1958) is typically used in fNIRS studies to ensure that the cap or headband is placed accurately. As the 10–20 system uses data from the entire human cortex to determine the orientation of the NIR-light sources and detectors, this system is less suitable for positioning NIR light sources and detectors as integrated in the headband (Jasper, 1958). There are, however, several alternative solutions to place the headband. The mfNIRS headband can for example be locally standardised on the vertical axis using the craniometric point of the nasion as an orientation point and the middle of the two preauricular points for positioning the headband on the horizontal axis. In addition, the headband could also be placed on the frontal polar lines of the 10–20 system (León-Carrión et al., 2010). Given the limited spatial resolution of mfNIRS, it should be evident that both of these placement procedures are indispensable to guaranteeing at least a minimum level of comparability between participants when analysing group-level data. Moreover, researchers must instruct participants to avoid strong, abrupt head movements. For measuring the cortical brain activity, innovative mfNIRS techniques that also integrate short channels (Brigadoi and Cooper, 2015) allow the simultaneous measurement of confounding signals (e.g. scalp blood circulation), thereby correcting the mfNIRS data as they are collected.

Given that ambient light might also be a cofounding variable, external light sources that might distort the mfNIRS signal during the measurement procedure need to be eliminated. Hence, to reduce the impact of environmental, external light, to assure an appropriate source-detector distance of about 3 cm and to guarantee sufficient light travelling through tissue, a proper skin contact with the NIR light sources and detectors have to be given. The resulting data quality can be checked by using an instrument-control software. Although data quality can be further evaluated and quantified later during offline data analysis, in many cases, it can be improved only within the limitations imposed by the quality of the raw data. Optimising the measurement setup is therefore of paramount importance. A range of software packages are available for data acquisition and analysis (e.g. Homer 3; NIRS Toolbox; AtlasViewer). Many are delivered along with fNIRS/mfNIRS devices or can be obtained online (e.g. https://nirx.net/software; www.nitrc.org/projects/homer2; https://openfnirs.org/software/homer/).

5. Applying mobile functional near-infrared spectroscopy in marketing research

5.1 Experimental procedure

To measure the neural processes and associated behaviour of customers in real-world situations – like at the point-of-sale or during negotiation talks – it is imperative to note that, to date, the environmental setting must be prepared to account and control for potential confounding environmental factors. Given the relative difficulty of identifying all occurrences of stimuli (e.g. unforeseen human interaction, background noise, functional pitfalls caused by movement-related artefacts) in naturalistic, mobile experiments, it is useful to combine mobile fNIRS with other neurophysiological methods, such as eye-tracking, to control for external, environmental cues whilst synchronising gaze directions and the neural mfNIRS data to be collected. Behavioural measurements should also be considered, as they allow for the interpretation of neural activity patterns measured through mfNIRS, in addition to hypotheses that are formulated according to theory. Given the necessity of such additional effort, and in the absence of innovative algorithms, the analysis of data from mfNIRS experiments currently requires significantly more time and effort, as the stimuli onsets, id est the time that a stimulus occurred in the environment, must be defined ex post, using additional data sources. Moreover, because researchers cannot define the exact times at which stimuli occurred, the time series defined ex post might not be as accurate as the time series defined within a stationary experiment, which makes it possible to send a trigger signal at the exact time a stimulus occurs. Recent developments show that marketing researcher and practitioners can be supported in data analysis through the use of innovative algorithms and data analysis toolkits. Most of the offered packages include comprehensive manuals explaining how to use the toolkits and, consequently, to analyse fNIRS data. Given the variation in the application complexity of software packages, marketing researchers must draw on their experiences with neuroimaging methods to determine which software package to use. For example, toolkits that rely solely on MATLAB analysis software interfaces (ww2.mathworks.cn/en/products/matlab.html) might not be as user-friendly for novices as MATLAB-based toolkits that have already been integrated within the fNIRS/mfNIRS system (Figure 3).

5.2 Data pre-processing

Although the data analysis toolkits may differ in their user-friendliness, some general steps can be identified that are (more or less) valid for all procedures to analyse mfNIRS data. As a first step, the raw data (NIR light absorption rates) maybe pre-processed. The pre-processing of fNIRS data can be subdivided into three steps. In the first step, the signal quality of each mfNIRS channel (i.e. source–detector combination, as previously defined) is checked. In case of poor signal quality, e.g. due to light oversaturation or strong movement-related artefacts, the associated channel needs to be excluded from further analysis. It should be noted that this procedure inevitably results in the loss of data. Researchers should therefore be very careful about excluding channels and the data associated with them, and they should attempt to apply the aforementioned preventive measurements to increase the mfNIRS data quality. In a second step, discontinuities and movement (spike) artefacts from the data time series need to be eliminated. In addition, abnormalities with two or more adjacent channels that have t-values greater than three standard deviations from the group average (Fishburn et al., 2014) or that indicate significantly more spike artefacts need to be excluded from further analysis. In this case as well, researchers should be very careful about the exclusion of data and, if applicable, such procedures should be reported.

In the final and third step of the pre-processing of mfNIRS data, a frequency filter can be used to smooth fNIRS data. In general, there are three different filter types that improve data quality based on individual data characteristics: low-, high- and band-pass filters. The low-pass filter deletes all fluctuations evolving over more rapid time scales from the data time series. This filter is often applied to de-noise data (removing aspects from the data that are not related to the measured effect). The high-pass filter deletes all fluctuations evolving over less rapid time scales. It is thus often used for de-trending (removing aspects of the data that is causing distortion). Lastly, the band-pass filter combines the low- and high-pass filters. It is particularly relevant when experimental events occur at regular time intervals, as is often the case in stationary mfNIRS experimental settings using a mixed-event/block design (Petersen and Dubis, 2012).

5.3 Data analysis

The analysis of oxygenated/de-oxygenated haemoglobin signals requires converting the pre-processed or raw NIR light absorption and attenuation data into concentrations of oxygenated and/or deoxygenated haemoglobin. Although previous research indicated significant correlations between fNIRS data signals and the sophisticated fMRI blood-oxygenation level-dependent (BOLD) signal (Noah et al., 2015; Wijeakumar et al., 2017), there is currently no consensus regarding whether the correlation is stronger with oxygenated haemoglobin (Hoshi et al., 2001; Strangman et al., 2002) or de-oxygenated haemoglobin signals (Huppert et al., 2006; Toronov et al., 2003), both of the signals that can be measured with mfNIRS. Recent literature concerning the validity and reliability of fNIRS/mfNIRS therefore suggests reporting both signals to avoid false positives (Hocke et al., 2018; Tachtsidis and Scholkmann, 2016). The most commonly used algorithm to convert raw mfNIRS data into (de-)oxygenated haemoglobin signals is the modified Beer–Lambert law (Kocsis et al., 2006; Kopton and Kenning, 2014; Scholkmann et al., 2014). The Beer–Lampert law is defined by the following equation (Lloyd-Fox et al., 2010):

(1) ΔA=Δc  *  L  *  DFC

It considers a constant optical scattering of the light and relates the change to the chromophore concentration (the smallest unit of light: photons) to changes in light attenuation (A), absorption coefficient (α), the concentration of the specific chromophore (c), source–detector separation (L) and the differential path length factor (DFC), all of which vary according to specific wavelengths, gender, age and differences in the tissue types (Kopton and Kenning, 2014; Lloyd-Fox et al., 2010).

After the raw light absorption and attenuation data have been transformed into oxygenated/de-oxygenated haemoglobin concentrations, the following step consists of statistically analysing the haemodynamic-state time series at the within-session and/or within-subject level, or across multiple sessions or subjects. The standard approach involves fitting a general linear model (GLM) to each haemodynamic time series that has been defined ex ante (Kenning et al., 2007). This model explains the response variable Yc (e.g. the haemodynamic response function at the smallest unit of the cortex surface, mostly pre-defined in a topographical layout, referred to as channel c = 1, … C) in terms of a linear combination of the explanatory variables plus an error term:

(2) Yc=xc1β1++xclβl+xcLβL+εc

Here, βl is the unknown parameter, corresponding to each of the L explanatory variables xcl (where l = 1, … L). The error terms εc are independent.

Writing this equation in full for each channel c yields the following set of simultaneous equations:

(3a) Y1=x11β1++x1lβl+x1LβL+ε1
(3b) Yc=xc1β1++xc1βl+xcLβL+εc
(3c) YC=xC1β1++xC1βl+xCLβL+εC

It has the equivalent matrix form:

(4) (Y1YcYC)=(x11x1lx1Lxc1xclxcLxC1xClxCL)(β1βcβC)+(ε1εcεC)
and it can be written in matrix notation as follows:
(5) Y=Xβ+ε
where Y is the column vector of observation, ε is the column vector of error terms and β is the column vector of parameters. The C × L matrix X with the clth element xcl is the design matrix. For example, this matrix can be used to specify the covariates corresponding to the different conditions in the experiment. If the participant repeatedly alternated between ten measurements of the experimental condition and ten of the control conditions, the model might include a covariate valued at 1 for each measurement corresponding to the control condition. After the design matrix has been defined, the different parameters are estimated based on the GLM (Kenning et al., 2007). It should be noted that the design matrix is always dependent on the research questions. The regression coefficients associated with each covariate in the best fit are referred to as β values. They are used to compute statistical values (e.g. t or F) associated with each channel for a given contrast of covariates, as previously defined in a topographical map. The most common method to do so is the contrast analysis, in which various contrasts can be analysed in terms of the pre-defined experimental design and research questions (e.g. comparing the neural activity patterns of two different experimental settings, such as experimental versus control group). Once the contrasts have been defined and the statistics have been computed, neural activity differences can be displayed for every channel in a statistical parametric map: a brain image in which the value of each channel is its corresponding statistic. The significance p-value threshold should be defined according to the expected effect size (Cohens’ d) (Cohen, 1988), and the results can be displayed in an anatomical representation of the brain. Most commonly, this consists of an atlas of the “standard brain”: a computer-generated 3D model derived from the brain anatomy of an individual or a group of subjects. In addition to contrast analyses, more advanced brain connectivity analyses that have recently been developed might be interesting in some situations, always depending on the research question (Hubert et al., 2012; Zhang et al., 2010; Lu et al., 2010).

5.4 Data interpretation

To interpret the mfNIRS findings, marketing researchers and practitioners – who are assumed to have only limited knowledge of the anatomic structure of the human brain and the associated neuropsychological processes – can consult the open data approach of neurosynth.org (Yarkoni et al., 2011) to acquire advanced knowledge on the specific research focus, the neuropsychological processes and its localisation and connectivity within the brain. One special feature of neurosynth.org is that it allows to identify all the research work that has been related to particular brain regions, research topics and/or cognitive constructs. This feature is of value as it automatically preselects the relevant literature, which might advance the consumer neuroscience and/or neuromarketing relevant knowledge of researchers and practitioners, helping them to interpret the mfNIRS related data and research findings.

6. Functional near-infrared spectroscopy/mobile functional near-infrared spectroscopy in marketing research: a status-quo analysis

To examine the current use of fNIRS/mfNIRS in marketing research, to assess whether and how the indicated guidelines are applied and to formulate theoretical and practical normative implications, a systematic review of recently published journal articles using mfNIRS/fNRIS experiments was conducted.

6.1 Applied methodology

The review was conducted according to a keyword search of “fNIRS”, “marketing”, “consumer neuroscience” and “neuromarketing” in the Google Scholar and Web of Science search engines (Martín-Martín et al., 2021). The review includes only scientific journal articles concerning fNIRS/mfNIRS studies conducted from 2009 to 2021. It excludes, editorials, review articles, extended abstracts and conference submissions. Moreover, it focuses only journal articles published in the broader field of marketing research. This search procedure identified 13 articles using fNIRS/mfNIRS to answer marketing-relevant research questions.

The majority of the identified articles have focussed on validating fNIRS/mfNIRS in the field of consumer neuroscience (Kim et al., 2016; Krampe et al., 2018b; Liu et al., 2018; Cakir et al., 2018; Burns et al., 2018), followed by studies using fNIRS/mfNIRS to predict the future consumer behaviour (Shimokawa et al., 2009; Krampe et al., 2018a; Cha et al., 2019; Gier et al., 2020; He et al., 2021) and the exploration of cultural (Burns et al., 2018) and/or gender-specific differences (Nissen and Krampe, 2021; Duan et al., 2021). The scientific articles identified can, thus, be categorised according to four overarching themes in the use of fNIRS/mfNIRS: “validation” in marketing research, “behaviour prediction” based on neural cortical activity, exploration of “cultural differences” and evaluation of “gender differences”. The results of the systematic literature review also reveal similarities and differences between the methodological approaches that have been used to conduct mfNIRS experiments (Figure 4).

6.2 Application-oriented commonalities

The results of the systematic literature review indicate that the mfNIRS headband is the most commonly used device for measuring brain activity patterns of the prefrontal cortex within the field of marketing research. The focus on brain regions ascribed to the prefrontal cortex is supported by recent research (Carlén, 2017), indicating three central brain regions that are known to play an essential role in explaining marketing-relevant entities. Firstly, the dlPFC has been investigated in many studies (Harris et al., 2018; Lim, 2018; Gier et al., 2020). The results obtained from these investigations include the finding that complex cognitive perceptions and decision-making processes can be associated with the dlPFC (Deppe et al., 2005; Knutson et al., 2007; Krampe et al., 2018b). Secondly, the OFC has been identified as an essential brain region. It has been related to decision-making processes and marketing-relevant entities, including the willingness to pay or the “liking” of objects, products, or advertising activities (Kühn et al., 2016; Plassmann et al., 2008; Wallis, 2007; Krampe et al., 2018a). Third, the medial/ventro-medial cortex continues to be of crucial importance in consumer neuroscience and neuromarketing research and has been linked to neuropsychological processes that underly individual value refraction and self-control in decision-making (Deppe et al., 2005; Hare et al., 2009).

The overarching role of the prefrontal cortex in consumer neuroscience and neuromarketing research should therefore be evident, explaining the research focus of recent mfNIRS studies. It is nevertheless important to note that other sub-cortical brain regions – such as the nucleus accumbens, which is generally associated with reward aspects (Knutson et al., 2007); the amygdala, which is associated with emotional/affective processing (Sergerie et al., 2008); and the ACC, which integrates cognitive processes of conflict recognition and conflict resolution (Kerns et al., 2004) – are also essential to the study of marketing-relevant entities. To date, however, technological limitations have prevented the exploration of these brain regions with mfNIRS, stressing again the importance to consider in advance whether mfNIRS can be used to answer specific research questions, given the correlation of the construct of interest and the associated brain regions (Figure 2).

Another interesting finding of the systematic literature review is that most fNIRS/mfNIRS studies concentrated primarily on oxygenated haemoglobin levels. As indicated before, there is an ongoing discussion concerning whether the oxygenated, de-oxygenated or total haemoglobin level is the best predictor of cortical neural activity (Dravida et al., 2017). Following this discussion and the associated mixed research findings, both haemoglobin levels might be reported in future studies, or if chosen for one, it should be explained – in terms of the research questions and the fNIRS/mfNIRS system to be used – which signals are analysed and why. In addition, haemodynamic signals might be improved with anti-correlation methods or by integrating short channels to the mfNIRS headband or cap (Zhou et al., 2020).

In the majority of the fNIRS/mfNIRS studies reported, data collection is harmonised by using one particular data collection procedure. Some studies use the 10–20 system (Jesper, 1958), while others use a more applied methodological approach (focussing on the nasion and/or the two preauricular points) (Figure 4). The 10–20 system appears particularly suitable for full-head measurement procedures using a whole-head cap. This observation is only conditionally valid for the studies reviewed, as they are based on data collected with a headband, and not with a cap covering the whole cortex. To date, there does not appear to be any single, accepted method for harmonising mfNIRS data collection. Such an approach is, however, needed as it could help to set standards, thereby allowing the comparison of marketing-relevant mfNIRS research findings across neuromarketing or consumer neuroscience studies. To ensure a minimum of comparability, when using the headband to measure brain regions of the prefrontal cortex, the craniometric point of the nasion can be used as a landmark on the vertical axis, and the middle of the two preauricular points can be used to position the headband on the horizontal axis to ensure that the same standardisation is used in all mfNIRS experiments in the future.

Most of the fNIRS/mfNIRS studies reviewed apply preventive and/or post-experiment data correction methods to cope for confounding effects, such as external light sources or abrupt head movements. These preventive applications are essential to increasing the quality of mfNIRS data, and they need to be considered in future research by applying preventive and post-data correction methods.

Although the combination of mfNIRS with other neuroscientific methods (defined as non-neuroimaging methods, like eye-tracking, pupillometry, electrocardiogram or heart rate variability) appears to be useful (Venkatraman et al., 2015), none of the studies reviewed attempts to combine mfNIRS with other neuroscientific methods and tools. The combination of mfNIRS with existing (neuroimaging) methods, acting as control elements to manage potential confounding effects that occur in a naturalistic research environment, might therefore increase the explained variance when investigating marketing-relevant entities. Future research in consumer neuroscience and neuromarketing might therefore explore the possibility of combining mfNIRS with other neuroscientific methods taking software and hardware issues into account, further validating mfNIRS in marketing research.

6.3 Research-oriented commonalities

As with any emerging methodology, it remains imperative to validate mfNIRS in the field of marketing research. To this end, the vast majority of the studies reviewed use multiple approaches to validate mfNIRS in marketing research. In general, four validation approaches may be used to validate a neuroimaging method in marketing research (Krampe et al., 2018a):

  1. Predictive validity, which is given when a measurement accurately predicts behaviour according to theory;

  2. Construct validity, which requires measurements obtained from a measurement procedure to behave exactly the same as the actual variable;

  3. Divergent validity, which is demonstrated when two different methods measure two different constructs without revealing any relationship; and

  4. Convergent validity, which is given when a strong relationship is found between the scores obtained from two different methods measuring the same construct (Gravetter and Fozano, 2018).

The results of the systematic review show that two of the four validity approaches have been used to validating mfNIRS in marketing research.

Following the predictive validity approach, Kim et al. (2018) report on a fNIRS study in which participants were requested to evaluate food images whilst wearing a mfNIRS headband. The results show that fNIRS is capable of predicting whether participants will favour or reject food images, based on increased neural activity in the right prefrontal cortex. Studies by Liu et al. (2018) and Cakir et al. (2018) add to these findings by showing that using fNIRS/mfNIRS to measure neural brain activity is helpful in merchandising activity evaluation and examining the related decision-making processes. Just recently, He et al. (2021) indicated further that mfNIRS is capable of exploring the neural correlates, which can predict attitudes towards advertisements, measuring brain regions of the right inferior frontal gyrus, which is ascribed to empathy and emotional contagion.

In addition, studies by Burns et al. (2018) and Krampe et al. (2017) adopt a convergent validity approach to validate the use of mfNIRS in marketing research. Both studies replicate a frequently cited neural effect that has previously been explored with fMRI to determine whether mfNIRS is capable of replicating fMRI results, given its technical capabilities. The results are consistent with previous fMRI findings, indicating significant neural activity in brain regions ascribed to specific neural and marketing-relevant effects.

It should, however, be mentioned that the indicated studies only targeted research questions that can be attributed to B2C marketing settings – more specifically to the field of consumer neuroscience. Though, B2C and B2B companies are different, both are confronted with dynamic changing market environments to which they need to respond (Lim et al., 2019). Future research might therefore approach the validation of mfNIRS also in a B2B marketing setting (Lim, 2018), focussing on research questions that examine inter alia interactions between actors of the value chain. Aspects that would contribute to the further validation of mfNIRS in the field of marketing research, as a whole.

mfNIRS has also been applied to predict consumer behaviour. More precisely, the neural activity elicited by marketing-relevant stimuli (e.g. auditory-sensory characteristics of a video sequence or point of sale merchandising elements) have been linked to behavioural/sales data to predict actual purchasing or consumption behaviour (Krampe et al., 2017; Cha et al., 2019; Gier et al., 2020). Moreover, mfNIRS has been applied to predict decision-making under risk and to explore information to predict investment behaviour (Shimokawa et al., 2009). The advantage of using mfNRS in real-world environments to predict customer behaviour adds value to marketing research, as it opens up the possibility of validating research findings that have been studied in an artificial laboratory environment. The application of mfNIRS could therefore strengthen future research approaches by allowing a marketing-relevant entity to be explored through different real-world perspectives focusing on neuropsychological, behavioural and traditional marketing methods insights in a real-world setting. Also, mfNIRS has recently been used to explore cultural and/or gender-specific differences in perceptions of persuasive messages (Burns et al., 2019), e-commerce websites (Nissen and Krampe, 2021) or in transnational brand purchase decisions associated with original or mixed culture advertisements (Duan et al., 2021). Aspects that provide additional evidence of the ability of mfNIRS to investigate individual differences between customers throughout the entire online and offline value chain.

7. Conclusion

The presented guidelines for the use of mfNIRS show that the methodological issues and performance of mfNIRS are relatively high for newcomers to the field of consumer neuroscience and neuromarketing. However, compared to other neuroimaging methods, such as fMRI, mfNIRS is less costly and can be applied in naturalistic contexts – two arguably fundamental factors in marketing research. As outlined in the review of recently published fNIRS/mfNIRS studies in marketing research, the application of mfNIRS in marketing has the potential to fundamentally improve the understanding of marketing-relevant entities by increasing the ecological validity in B2C and B2B marketing settings. However, the latter requires future research efforts with the aim of validating mfNIRS as a neuroimaging method also in the B2B context. Researchers and marketers are therefore advised to broaden the scope of mfNIRS research to include the study of (neural) exchange processes between actors in the value chain. This offers opportunities to explore marketing-relevant entities in their entirety, which advances marketing research as a whole and which goes beyond the previous investigation of impact-oriented research approaches. The presented guidelines support this development by proposing harmonised standards and procedures for the application of mfNIRS in marketing research that enable comparability between research findings.

Figures

mfNIRS functionality

Figure 1.

mfNIRS functionality

Event-driven process chains for using mfNIRS

Figure 2.

Event-driven process chains for using mfNIRS

Processual display of conducting mfNIRS experiments

Figure 3.

Processual display of conducting mfNIRS experiments

Recent fNIRS/mfNIRS experiments in the fields of consumer neuroscience and neuromarketing

Figure 4.

Recent fNIRS/mfNIRS experiments in the fields of consumer neuroscience and neuromarketing

The most common direct and indirect neuroimaging methods applied in consumer neuroscience and cognate research

Neuroimaging method Advantages Disadvantages Related studies Mobile applicability
Electromagnetic neuroimaging methods
Magnetencephalography MEG
Registration of changes in magnetic streams elicited by the electrophysical signal of neurons
+ Good temporal resolution
(neural activity changes can be measured in s)
+ Non-invasive
– High measurement costs
– Relatively complicated data analysis
– Poor spatial resolution (limited potential to localise brain activity on the cortex)
Ioannides et al. (2000), Zaltman and Kosslyn (2000), Boto et al. (2018) Partial
Electroencephalography EEG
Measurement of voltage fluctuations at the surface of the cortex
+ Good temporal resolution (neural-activity changes can be measured in ms)
+ Moderate equipment costs
+ Relatively uncomplicated data analysis
+ Non-invasive
– Poor spatial resolution (limited potential to localise brain activity on the cortex)
– Relatively complicated data collection
Boksem and Smidts (2015), Harris et al. (2018); Müller-Putz et al. (2015) Partial
Metabolic neuroimaging methods
Positron-Emission-Tomography PET
Nuclear medical technique for analysing indirect, metabolic processes in neurons
+ Good spatial resolution
(neural activity of the whole brain can be measured and localised)
– Very poor temporal resolution (compared to EEG)
– Invasive
Harris et al. (2018) No
Functional magnetic resonance imaging fMRI
Measurement of indirect, metabolic activity using the magnetic properties of blood (the BOLD signal)
+ Very good spatial resolution
(neural activity of the whole brain can be measured and localised)
+ Non-invasive
– High measurement costs
– Relatively complicated data analysis
– Poor temporal resolution (compared to EEG)
Dimoka and Davies (2008), Dimoka (2010), Falk et al. (2016), Hubert et al. (2018), Kenning et al. (2007), Ogawa et al. (1990) No

References

Adorni, R., Brugnera, A., Gatti, A., Tasca, G.A., Sakatani, K. and Compare, A. (2018), “Psychophysiological responses to stress related to anxiety in healthy aging”, Journal of Psychophysiology, Vol. 33 No. 3.

Berns, G.S. and Moore, S.E. (2012), “A neural predictor of cultural popularity”, Journal of Consumer Psychology, Vol. 22 No. 1, pp. 154-160.

Boksem, M.A. and Smidts, A. (2015), “Brain responses to movie trailers predict individual preferences for movies and their population-wide commercial success”, Journal of Marketing Research, Vol. 52 No. 4, pp. 482-492.

Boto, E., Holmes, N., Leggett, J., Roberts, G., Shah, V., Meyer, S.S., Muñoz, L.D., Mullinger, K.J., Tierney, T.M., Bestmann, S. and Barnes, G.R. (2018), “Moving magnetoencephalography towards real-world applications with a wearable system”, Nature, Vol. 555 No. 7698, pp. 657-661.

Brigadoi, S. and Cooper, R.J. (2015), “How short is short? Optimum source-detector distance for short-separation channels in functional near-infrared spectroscopy”, Neurophotonics, Vol. 2 No. 2, p. 25005.

Brugnera, A., Adorni, R., Compare, A., Zarbo, C. and Sakatani, K. (2016), “Cortical and autonomic patterns of emotion experiencing during a recall task”, Journal of Psychophysiology, Vol. 32 No. 2, pp. 53-63.

Brugnera, A., Zarbo, C., Adorni, R., Tasca, G.A., Rabboni, M., Bondi, E., Compare, A. and Sakatani, K. (2017), “Cortical and cardiovascular responses to acute stressors and their relations with psychological distress”, International Journal of Psychophysiology, Vol. 114, pp. 38-46.

Burns, S.M., Barnes, L.N., Katzman, P.L., Ames, D.L., Falk, E.B. and Lieberman, M.D. (2018), “A functional near infrared spectroscopy (fNIRS) replication of the sunscreen persuasion paradigm”, Social Cognitive and Affective Neuroscience, Vol. 13 No. 6, pp. 628-636.

Burns, S.M., Barnes, L.N., McCulloh, I.A., Dagher, M.M., Falk, E.B., Storey, J.D. and Lieberman, M.D. (2019), “Making social neuroscience less WEIRD: using fNIRS to measure neural signatures of persuasive influence in a Middle east participant sample”, Journal of Personality and Social Psychology, Vol. 116 No. 3, p. e1.

Cakir, M.P., Çakar, T., Girisken, Y. and Yurdakul, D. (2018), “An investigation of the neural correlates of purchase behavior through fNIRS”, European Journal of Marketing, Vol. 52 Nos 1/2, pp. 224-243.

Carlén, M. (2017), “What constitutes the prefrontal cortex?”, Science, Vol. 358 No. 6362, pp. 478-482.

Cha, K.C., Suh, M., Kwon, G., Yang, S. and Lee, E.J. (2019), “Young consumers’ brain responses to pop music on YouTube”, Asia Pacific Journal of Marketing and Logistics, Vol. 32 No. 5, pp. 1132-1148.

Cohen, J. (1988), Statistical Power Analysis for the Behavioral Sciences, 2nd ed., Erlbaum Associates, Hillsdale.

Dalgleish, T. (2004), “The emotional brain”, Nature Reviews Neuroscience, Vol. 5 No. 7, pp. 583-589.

Daugherty, T. and Thomas, A.R. (2018), “Guest editorial”, European Journal of Marketing, Vol. 52 Nos 1/2, pp. 2-3.

Deppe, M., Schwindt, W., Kugel, H., Plassmann, H. and Kenning, P. (2005), “Nonlinear responses within the medial prefrontal cortex reveal when specific implicit information influences economic decision making”, Journal of Neuroimaging, Vol. 15 No. 2, pp. 171-182.

Dimoka, A. (2010), “What does the brain tell us about trust and distrust? Evidence from a functional neuroimaging study”, MIS Quarterly, Vol. 34 No. 2, pp. 373-396.

Dimoka, A., Davis, F.D., Gupta, A., Pavlou, P.A., Banker, R.D., Dennis, A.R., Ischebeck, A., Müller-Putz, G., Benbasat, I., Gefen, D. and Kenning, P. (2012), “On the use of neurophysiological tools in IS research: developing a research agenda for NeuroIS”, MIS Quarterly, Vol. 36 No. 3, pp. 679-702.

Dravida, S., Noah, J.A., Zhang, X. and Hirsch, J. (2017), “Comparison of oxyhemoglobin and deoxyhemoglobin signal reliability with and without global mean removal for digit manipulation motor tasks”, Neurophotonics, Vol. 5 No. 1, p. 11006.

Duan, L., Ai, H., Yang, L., Xu, L. and Xu, P. (2021), “Gender differences in transnational brand purchase decision toward mixed culture and original culture advertisements: an fNIRS study”, Frontiers in Psychology, Vol. 12, p. 654360.

Ernst, L.H., Plichta, M.M., Lutz, E., Zesewitz, A.K., Tupak, S.V., Dresler, T., Ehlis, A.C. and Fallgatter, A.J. (2013), “Prefrontal activation patterns of automatic and regulated approach – avoidance reactions – a functional near-infrared spectroscopy (fNIRS) study”, Cortex, Vol. 49 No. 1, pp. 131-142.

Falk, E.B., O’Donnell, M.B., Tompson, S., Gonzalez, R., Dal Cin, S., Strecher, V., … and An, L. (2016), “Functional brain imaging predicts public health campaign success”, Social Cognitive and Affective Neuroscience, Vol. 11 No. 2, pp. 204-214.

Ferrari, M. and Quaresima, V. (2012), “A brief review on the history of human functional near-infrared spectroscopy (fNIRS) development and fields of application”, Neuroimage, Vol. 63 No. 2, pp. 921-935.

Fishburn, F.A., Norr, M.E., Medvedev, A.V. and Vaidya, C.J. (2014), “Sensitivity of fNIRS to cognitive state and load”, Frontiers in Human Neuroscience, Vol. 8, p. 76.

Funane, T., Atsumori, H., Katura, T., Obata, A.N., Sato, H., Tanikawa, Y., Okada, E. and Kiguchi, M. (2014), “Quantitative evaluation of deep and shallow tissue layers' contribution to fNIRS signal using multi-distance optodes and independent component analysis”, NeuroImage, Vol. 85, pp. 150-165.

Gefen, D., Ayaz, H. and Onaral, B. (2014), “Applying functional near infrared (fNIR) spectroscopy to enhance MIS research”, AIS Transactions on Human-Computer Interaction, Vol. 6 No. 3, pp. 55-73.

Gier, N.R., Strelow, E. and Krampe, C. (2020), “Measuring dlPFC signals to predict the success of merchandising elements at the point-of-sale – a fNIRS approach”, Frontiers in Neuroscience, Vol. 14, p. 575494.

Girouard, A., Solovey, E.T., Hirshfield, L.M., Peck, E.M., Chauncey, K., Sassaroli, A., Fantini, S. and Jacob, R.J. (2010), “From brain signals to adaptive interfaces: using fNIRS in HCI”, Brain-Computer Interfaces, Springer, London, pp. 221-237.

Gordon, R., Ciorciari, J. and van Laer, T. (2018), “Using EEG to examine the role of attention, working memory, emotion, and imagination in narrative transportation”, European Journal of Marketing, Vol. 52 Nos 1/2, pp. 92-117.

Gravetter, F.J. and Forzano, L.A.B. (2018), Research Methods for the Behavioral Sciences, Cengage Learning.

Hallett, M. (2000), “Transcranial magnetic stimulation and the human brain”, Nature, Vol. 406 No. 6792, pp. 147-150.

Hare, T.A., Camerer, C.F. and Rangel, A. (2009), “Self-control in decision-making involves modulation of the vmPFC valuation system”, Science, Vol. 324 No. 5927, pp. 646-648.

Harris, J.M., Ciorciari, J. and Gountas, J. (2018), “Consumer neuroscience for marketing researchers”, Journal of Consumer Behaviour, Vol. 17 No. 3, pp. 239-252.

He, L., Pelowski, M., Yu, W. and Liu, T. (2021), “Neural resonance in consumers' right inferior frontal gyrus predicts attitudes toward advertising”, Psychology and Marketing, Vol. 38 No. 9, pp. 1538-1549.

Hocke, L.M., Oni, I.K., Duszynski, C.C., Corrigan, A.V., Frederick, B.D. and Dunn, J.F. (2018), “Automated processing of fNIRS data-a visual guide to the pitfalls and consequences”, Algorithms, Vol. 11 No. 5, p. 67.

Hoshi, Y., Kobayashi, N. and Tamura, M. (2001), “Interpretation of near-infrared spectroscopy signals: a study with a newly developed perfused rat brain model”, Journal of Applied Physiology, Vol. 90 No. 5, pp. 1657-1662.

Hubert, M., Linzmajer, M., Riedl, R., Kenning, P. and Hubert, M. (2012), “Introducing connectivity analysis to NeuroIS research”, Proceedings of the International Conference for Information Systems.

Hubert, M., Hubert, M., Linzmajer, M., Riedl, R. and Kenning, P. (2018), “Trust me if you can–neurophysiological insights on the influence of consumer impulsiveness on trustworthiness evaluations in online settings”, European Journal of Marketing.

Huppert, T.J., Hoge, R.D., Diamond, S.G., Franceschini, M.A. and Boas, D.A. (2006), “A temporal comparison of BOLD, ASL, and NIRS hemodynamic responses to motor stimuli in adult humans”, NeuroImage, Vol. 29 No. 2, pp. 368-382.

Jasper, H. (1958), “The ten-twenty electrode system of the international federation”, Electroencephalography in Clinical Neurophysiology, Vol. 10, pp. 367-380.

Jobsis, F.F. (1977), “Noninvasive, infrared monitoring of cerebral and myocardial oxygen sufficiency and circulatory parameters”, Science, Vol. 198 No. 4323, pp. 1264-1267.

Kannan, P.K. (2017), “Digital marketing: a framework, review and research agenda”, International Journal of Research in Marketing, Vol. 34 No. 1, pp. 22-45.

Kenning, P. and Plassmann, H. (2005), “NeuroEconomics: an overview from an economic perspective”, Brain research bulletin, Vol. 67 No. 5, pp. 343-354.

Kenning, P., Plassmann, H. and Ahlert, D. (2007), “Applications of functional magnetic resonance imaging for market research”, Qualitative Market Research: An International Journal, Vol. 10 No. 2, pp. 135-152.

Kerns, J.G., Cohen, J.D., MacDonald, A.W., Cho, R.Y., Stenger, V.A. and Carter, C.S. (2004), “Anterior cingulate conflict monitoring and adjustments in control”, Science, Vol. 303 No. 5660, pp. 1023-1026.

Kim, J.Y., Kim, K.I., Han, C.H., Lim, J.H. and Im, C.H. (2016), “Estimating consumers' subjective preference using functional near infrared spectroscopy: a feasibility study”, Journal of Near Infrared Spectroscopy, Vol. 24 No. 5, pp. 433-441.

Knutson, B., Rick, S., Wimmer, G.E., Prelec, D. and Loewenstein, G. (2007), “Neural predictors of purchases”, Neuron, Vol. 53 No. 1, pp. 147-156.

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.

Kopton, I.M. and Kenning, P. (2014), “Near-infrared spectroscopy (NIRS) as a new tool for neuroeconomic research”, Frontiers in Human Neuroscience, Vol. 8, p. 549.

Krampe, C., Gier, N.R. and Kenning, P. (2018b), “The application of mobile fNIRS in marketing research – detecting the ‘First-Choice-Brand’ effect”, Frontiers in Human Neuroscience, Vol. 12 No. 433.

Krampe, C., Gier, N. and Kenning, P. (2017), “Beyond traditional neuroimaging: can mobile fNIRS add to NeuroIS?”, Information Systems and Neuroscience, Springer, pp. 151-157.

Krampe, C., Strelow, E., Haas, A. and Kenning, P. (2018a), “The application of mobile fNIRS to ‘shopper neuroscience’ – first insights from a merchandising communication study”, European Journal of Marketing, Vol. 52 No. 1/2, pp. 244-259.

Kühn, S., Strelow, E. and Gallinat, J. (2016), “Multiple ‘buy buttons’ in the brain: forecasting chocolate sales at point-of-sale based on functional brain activation using fMRI”, Neuroimage, Vol. 136, pp. 122-128.

Kwong, K.K., Belliveau, J.W., Chesler, D.A., Goldberg, I.E., Weisskoff, R.M., Poncelet, B.P., Kennedy, D.N., Hoppel, B.E., Cohen, M.S. and Turner, R. (1992), “Dynamic magnetic resonance imaging of human brain activity during primary sensory stimulation”, Proceedings of the National Academy of Sciences, Vol. 89 No. 12, pp. 5675-5679.

Lee, N., Chamberlain, L. and Brandes, L. (2018), “Welcome to the jungle! The neuromarketing literature through the eyes of a newcomer”, European Journal of Marketing, Vol. 52 Nos 1/2, pp. 4-38.

León-Carrión, J. and León-Domínguez, U. (2012), “Functional near-infrared spectroscopy (fNIRS): principles and neuroscientific applications”, Neuroimaging Methods, pp. 48-74.

Liao, S. and Culver, J. (2014), “Near infrared optical technologies to illuminate the status of the neonatal brain”, Current Pediatric Reviews, Vol. 10 No. 1, pp. 73-86.

Lim, W.M. (2018), “Demystifying neuromarketing”, Journal of Business Research, Vol. 91, pp. 205-220.

Lim, W.M., Ahmed, P.K. and Ali, M.Y. (2019), “Data and resource maximization in business-to-business marketing experiments: methodological insights from data partitioning”, Industrial Marketing Management, Vol. 76, pp. 136-143.

Liu, X., Kim, C.S. and Hong, K.S. (2018), “An fNIRS-based investigation of visual merchandising displays for fashion stores”, Plos One, Vol. 13 No. 12, p. e0208843.

Lloyd-Fox, S., Blasi, A. and Elwell, C. (2010), “Illuminating the developing brain: the past, present and future of functional near infrared spectroscopy”, Neuroscience and Biobehavioral Reviews, Vol. 34 No. 3, pp. 269-284.

Lu, C.M., Zhang, Y.J., Biswal, B.B., Zang, Y.F., Peng, D.L. and Zhu, C.Z. (2010), “Use of fNIRS to assess resting state functional connectivity”, Journal of Neuroscience Methods, Vol. 186 No. 2, pp. 242-249.

McCormick, P.W., Stewart, M., Lewis, G., Dujovny, M. and Ausman, J.I. (1992), “Intracerebral penetration of infrared light: technical note”, Journal of Neurosurgery, Vol. 76 No. 2, pp. 315-318.

Mansor, A.A. and Isa, S.M. (2020), “Fundamentals of neuromarketing: what is it all about?”, Neuroscience Research Notes, Vol. 3 No. 4, pp. 22-28.

Martín-Martín, A., Thelwall, M., Orduna-Malea, E. and López-Cózar, E.D. (2021), “Google scholar, Microsoft academic, Scopus, dimensions, web of science, and OpenCitations’ COCI: a multidisciplinary comparison of coverage via citations”, Scientometrics, Vol. 126 No. 1, pp. 871-906.

Müller-Putz, G.R., Riedl, R. and Wriessnegger, S.C. (2015), “Electroencephalography (EEG) as a research tool in the information systems discipline: foundations, measurement, and applications”, Communications of the Association of Information Systems, Vol. 37 No. 1, p. 46.

Nissen, A. and Krampe, C. (2021), “Why he buys it and she doesn't – exploring self-reported and neural gender differences in the perception of eCommerce websites”, Computers in Human Behavior, Vol. 121.

Noah, J.A., Ono, Y., Nomoto, Y., Shimada, S., Tachibana, A., Zhang, X. and Hirsch, J. (2015), “fMRI validation of fNIRS measurements during a naturalistic task”, Journal of Visualized Experiments, Vol. 100 No. 100.

Ogawa, S., Lee, T.M., Kay, A.R. and Tank, D.W. (1990), “Brain magnetic resonance imaging with contrast dependent on blood oxygenation”, Proceedings of the National Academy of Sciences, Vol. 87 No. 24, pp. 9868-9872.

Okada, E. and Delpy, D.T. (2003), “Near-infrared light propagation in an adult head model. II. Effect of superficial tissue thickness on the sensitivity of the near-infrared spectroscopy signal”, Applied Optics, Vol. 42 No. 16, pp. 2915-2921.

Petersen, S.E. and Dubis, J.W. (2012), “The mixed block/event-related design”, Neuroimage, Vol. 62 No. 2, pp. 1177-1184.

Plassmann, H., O'Doherty, J., Shiv, B. and Rangel, A. (2008), “Marketing actions can modulate neural representations of experienced pleasantness”, Proceedings of the National Academy of Sciences, Vol 105 No. 3, pp. 1050-1054.

Plassmann, H., Venkatraman, V., Huettel, S. and Yoon, C. (2015), “Consumer neuroscience: applications, challenges, and possible solutions”, Journal of Marketing Research, Vol. 52 No. 4, pp. 427-435.

Quaresima, V. and Ferrari, M. (2016), “Functional Near-Infrared spectroscopy (fNIRS) for assessing cerebral cortex function during human behavior in natural/social situations A concise review”, Organizational Research Methods, Vol. 22 No. 1, pp. 46-68.

Quaresima, V. and Ferrari, M. (2019), “Functional near-infrared spectroscopy (fNIRS) for assessing cerebral cortex function during human behavior in natural/social situations: a concise review”, Organizational Research Methods, Vol. 22 No. 1, pp. 46-68.

Scholkmann, F., Kleiser, S., Metz, A.J., Zimmermann, R., Pavia, J.M., Wolf, U. and Wolf, M. (2014), “A review on continuous wave functional near-infrared spectroscopy and imaging instrumentation and methodology”, NeuroImage, Vol. 85, pp. 6-27.

Sergerie, K., Chochol, C. and Armony, J.L. (2008), “The role of the amygdala in emotional processing: a quantitative meta-analysis of functional neuroimaging studies”, Neuroscience and Biobehavioral Reviews, Vol. 32 No. 4, pp. 811-830.

Seth, A.K., Baars, B.J. and Edelman, D.B. (2005), “Criteria for consciousness in humans and other mammals”, Consciousness and Cognition, Vol. 14 No. 1, pp. 119-139.

Shaw, S.D. and Bagozzi, R.P. (2018), “The neuropsychology of consumer behavior and marketing”, Consumer Psychology Review, Vol. 1 No. 1, pp. 22-40.

Shimokawa, T., Suzuki, K., Misawa, T. and Miyagawa, K. (2009), “Predictability of investment behavior from brain information measured by functional near-infrared spectroscopy: a Bayesian neural network model”, Neuroscience, Vol. 161 No. 2, pp. 347-358.

Strangman, G., Culver, J.P., Thompson, J.H. and Boas, D.A. (2002), “A quantitative comparison of simultaneous bold fMRI and NIRS recordings during functional brain activation”, NeuroImage, Vol. 17 No. 2, pp. 719-731.

Tachtsidis, I. and Scholkmann, F. (2016), “False positives and false negatives in functional near-infrared spectroscopy: issues, challenges, and the way forward”, Neurophotonics, Vol. 3 No. 3.

Telpaz, A., Webb, R. and Levy, D.J. (2015), “Using EEG to predict consumers' future choices”, Journal of Marketing Research, Vol. 52 No. 4, pp. 511-529.

Toronov, V., Walker, S., Gupta, R., Choi, J.H., Gratton, E., Hueber, D. and Webb, A. (2003), “The roles of changes in deoxyhemoglobin concentration and regional cerebral blood volume in the fMRI BOLD signal”, NeuroImage, Vol. 19 No. 4, pp. 1521-1531.

Torricelli, A., Contini, D., Pifferi, A., Caffini, M., Re, R., Zucchelli, L. and Spinelli, L. (2014), “Time domain functional NIRS imaging for human brain mapping”, NeuroImage, Vol. 85, pp. 28-50.

Venkatraman, V., Dimoka, A., Pavlou, P.A., Vo, K., Hampton, W., Bollinger, B., Hershfield, H.E., Ishihara, M. and Winer, R.S. (2015), “Predicting advertising success beyond traditional measures: new insights from neurophysiological methods and market response modelling”, Journal of Marketing Research, Vol. 52 No. 4, pp. 436-452.

Wallis, J.D. (2007), “Orbitofrontal cortex and its contribution to decision-making”, Annual Review of Neuroscience, Vol. 30 No. 1, pp. 31-56.

Wijeakumar, S., Huppert, T.J., Magnotta, V.A., Buss, A.T. and Spencer, J.P. (2017), “Validating an image-based fNIRS approach with fMRI and a working memory task”, NeuroImage, Vol. 147, pp. 204-218.

Yarkoni, T., Poldrack, R.A., Nichols, T.E., Van Essen, D.C. and Wager, T.D. (2011), “Large-scale automated synthesis of human functional neuroimaging data”, Nature Methods, Vol. 8 No. 8, pp. 665-670.

Zaltman, G. (2000), “Consumer researchers: take a hike!”, Journal of Consumer Research, Vol. 26 No. 4, pp. 423-428.

Zhang, H., Zhang, Y.J., Lu, C.M., Ma, S.Y., Zang, Y.F. and Zhu, C.Z. (2010), “Functional connectivity as revealed by independent component analysis of resting-state fNIRS measurements”, NeuroImage, Vol. 51 No. 3, pp. 1150-1161.

Zhou, X., Sobczak, G., McKay, C.M. and Litovsky, R.Y. (2020), “Comparing fNIRS signal qualities between approaches with and without short channels”, Plos One, Vol. 15 No. 12, p. e0244186.

Further reading

Ariely, D. and Berns, G.S. (2010), “Neuromarketing: the hope and hype of neuroimaging in business”, Nature Reviews Neuroscience, Vol. 11 No. 4, pp. 284-292.

Babiloni, F. and Astolfi, L. (2014), “Social neuroscience and hyperscanning techniques: past, present and future”, Neuroscience and Biobehavioral Reviews, Vol. 44, pp. 76-93.

Buettner, R. (2017), “The relationship between visual website complexity and a user’s mental workload: a NeuroIS perspective”, Information Systems and Neuroscience, Springer, Cham, pp. 107-113.

Carrieri, M., Lancia, S., Bocchi, A., Ferrari, M., Piccardi, L. and Quaresima, V. (2018), “Does ventrolateral prefrontal cortex help in searching for the lost key? Evidence from an fNIRS study”, Brain Imaging and Behavior, Vol. 12 No. 3, pp. 785-797.

Cooper, R., Selb, J., Gagnon, L., Philip, D., Schytz, H.W., Iversen, H.K., Ashina, M. and Boas, D.A. (2012), “A systematic comparison of motion artifact correction techniques for functional near-infrared spectroscopy”, Frontiers in Neuroscience, Vol. 6, p. 147.

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.

Dedovic, K., Duchesne, A., Andrews, J., Engert, V. and Pruessner, J.C. (2009a), “The brain and the stress axis: the neural correlated of cortisol regulation in response to stress”, NeuroImage, Vol. 47 No. 3, pp. 864-871.

Dedovic, K., Rexroth, M., Wolff, E., Duchesne, A., Scherling, C., Beaudry, T., Lue, S.D., Lord, C., Engert, V. and Pruessner, J.C. (2009b), “Neural correlates of processing stressful information: An Event-Related fMRI study”, Brain Research, Vol. 1293, pp. 49-60.

Dimoka, A., Pavlou, P.A. and Davis, F.D. (2011), “Research commentary – NeuroIS: the potential of cognitive neuroscience for information systems research”, Information Systems Research, Vol. 22 No. 4, pp. 687-702.

Essenpreis, M., Elwell, C., Cope, M., Van der Zee, P., Arridge, S. and Delpy, D. (1993), “Spectral dependence of temporal point spread functions in human tissues”, Applied Optics, Vol. 32 No. 4, pp. 418-425.

Gnewuch, U., Morana, S. and Maedche, A. (2017), “Towards designing cooperative and social conversational agents for customer service”, Proceedings of the International Conference of Information Systems, 33rd, pp. 1-13.

Gravetter, F.J. and Forzano, L.A.B. (2018), Research Methods for the Behavioral Sciences, Cengage Learning.

Homburg, C., Stierl, M. and Bornemann, T. (2013), “Corporate social responsibility in business-to-business markets: how organizational customers account for supplier corporate social responsibility engagement”, Journal of Marketing, Vol. 77 No. 6, pp. 54-72.

Knutson, B. and Bossaerts, P. (2007), “Neural antecedents of financial decisions”, Journal of Neuroscience, Vol. 27 No. 31, pp. 8174-8177.

Koenigs, M. and Tranel, D. (2007), “Prefrontal cortex damage abolishes brand-cued changes in cola preference”, Social Cognitive and Affective Neuroscience, Vol. 3 No. 1.

Kohl, M., Nolte, C., Heekeren, H.R., Horst, S., Scholz, U., Obrig, H. and Villringer, A. (1998), “Determination of the wavelength dependence of the differential pathlength factor from near-infrared pulse signals”, Physics in Medicine and Biology, Vol. 43 No. 6, p. 1771.

Liu, X. and Hong, K.S. (2017), “Detection of primary RGB colors projected on a screen using fNIRS”, Journal of Innovative Optical Health Sciences, Vol. 10 No. 3.

Nissen, A., Krampe, C., Kenning, P. and Schütte, R. (2019), “Utilizing mobile fNIRS to investigate neural correlates of the TAM in eCommerce”, Proceedings of the International Conference for Information Systems.

Pengnate, S.F. and Sarathy, R. (2017), “An experimental investigation of the influence of website emotional design features on trust in unfamiliar online vendors”, Computers in Human Behavior, Vol. 67, pp. 49-60.

Riedl, R., Hubert, M. and Kenning, P. (2010), “Are there neural gender differences in online trust? An fMRI study on the perceived trustworthiness of eBay offers”, MIS Quarterly, Vol. 34 No. 2, pp. 397-428.

Riedl, R. and Léger, P.-M. (2016), Fundamentals of NeuroIS. Studies in Neuroscience, Psychology and Behavioral Economics, Springer, Berlin, Heidelberg.

Rosenbaum, D., Thomas, M., Hilsendegen, P., Metzger, F.G., Haeussinger, F.B., Nuerk, H.C., Fallgatter, A.J., Nieratschker, V. and Ehlis, A.C. (2018), “Stress related dysfunction of the right inferior frontal cortex in high ruminators: An fNIRS study”, NeuroImage: Clinical, Vol. 18, pp. 510-517.

Seeger, A. and Heinzl, A. (2017), “Human vs. Machine: conversational agents”, Information Systems and Neuroscience, pp. 129-139.

Shao, J. and Ünal, E. (2019), “What do consumers value more in green purchasing? Assessing the sustainability practices from demand side of business”, Journal of Cleaner Production, Vol. 209, pp. 1473-1483.

Solnais, C., Andreu-Perez, J., Sánchez-Fernández, J. and Andréu-Abela, J. (2013), “The contribution of neuroscience to consumer research: a conceptual framework and empirical review”, Journal of Economic Psychology, Vol. 36, pp. 68-81.

Weil, M. and Rosen, L.D. (1997), Technostress: Coping with Technology @ Work @Home @Play, Wiley, New York, NY.

Zhao, H. and Cooper, R.J. (2017), “Review of recent progress toward a fiberless, whole-scalp diffuse optical tomography system”, Neurophotonics, Vol. 5 No. 1, p. 1.

Zhao, H., Tanikawa, Y., Gao, F., Onodera, Y., Sassaroli, A., Tanaka, K. and Yamada, Y. (2002), “Maps of optical differential pathlength factor of human adult forehead, somatosensory motor and occipital regions at multi-wavelengths in NIR”, Physics in Medicine and Biology, Vol. 47 No. 12, p. 2075.

Acknowledgements

The author would like to thank Peter Kenning, Nadine Gier (Heinrich Heine University Düsseldorf) and Enrique Strelow (Justus-Liebig-University Gießen) for their valuable and constructive comments and remarks during the course of this research work.

Corresponding author

Caspar Krampe can be contacted at: caspar.krampe@wur.nl

Related articles