1
The effect of positive arousal on memory retrieval
of advertisements:
A neuromarketing approach
Óscar Picallo
Master’s Thesis, 15 ECTS Master´s Program in Cognitive Science, 60 ECTS
Spring Term 2018
Supervisor: Tiziana Pedale
2
Acknowledgements I would like to thank my supervisor, Tiziana Pedale, for her infinite patience with my constant mistakes.
3
Abstrakt Målet med den här neuromarketing-studien var att undersöka effekten av positiva emotionella kontexter vid återkallning av neutrala stimuli. I föreliggande fMRI-experiment ingick 26 ryska deltagare, varav 13 var kvinnor och 13 män, med en medelålder på 23.4 år i ett åldersspann mellan 18 och 35 år. Deltagarna scannades medan de utförde igenkänningstest bestående av två faser: inkodnings-fas och återkallnings-fas. I inkodningsfasen visades olika snacks-produkter på en skärm tillsammans med emotionella scener (scener med hög och låg upphetsningsgrad) och neutrala scener. Senare, i återkallningsfasen, blev deltagarna tillfrågade om huruvida de kom ihåg produkten eller inte. De erhållna resultaten var oväntade, givet att två områden relaterat till minne, så som amygdala och hippocampus, inte visade ökade aktivitet när deltagarna framgångsrikt återkallade neutrala stimuli som visats i en emotionell kontext. Å andra sidan, andra områden som relaterar till episodiska emotionella minnen, så som temporala cortex, insula och cingulate cortex, visade på ökad aktivitet. På beteende-nivån hittade vi inte någon förbättring av minnes-prestation för den emotionella betingelsen jämfört med den neutrala betingelsen. Den här studien visar på det stora antalet hjärnområden som är involverade i återkallningsprocessen och begränsningarna om emotionella kontexter har när det kommer till att förbättra minnes-prestation.
4
Abstract The objective of this neuromarketing study is to investigate the effect of positive emotional contexts in the recall of neutral stimuli. In the present fMRI experiment, 26 Russian participants of whom, 13 were women, and 13 were men, with an average age of 23.4 years and within a range of age between 18 and 35 years, were scanned while performing a recognition memory test composed of two phases: encoding phase and retrieval phase. In the encoding phase, different snacks products were shown on the monitor accompanied by emotional scenes (high arousal and low arousal scenes) and neutral scenes. Later, at the retrieval phase, the participants were shown the same products plus the same number of unseen products again. The participants had to answer if they remember the product or not. The results obtained showed that the two areas closely related to memory, such as the amygdala and the hippocampus, did not show greater activation when the subjects successfully recalled the neutral stimuli that had been exposed in emotional contexts. On the other hand, other areas related to episodic emotional memory, such as the temporal cortex, the insula or the cingulate cortex, did show greater activation. On the behavioural level, I did not find an improvement of the memory performance on the emotional condition compared to the neutral condition. This study shows a large number of brain areas involved in the retrieval process and the limitations that emotional scenarios have when it comes to improving memory performance.
5
The effect of positive arousal on memory retrieval of advertisements:
A neuromarketing approach
Background
Emotions are defined as a temporary change in the psychophysiological, cognitive and behavioral state driven by an internal/external stimulus, which is salient from an affective point of view (Damasio, 1998; Cabanac, 2002). According to the dimensional approach (Hamann, 2012), emotional states can be described through two orthogonal dimensions: the valence (i.e., the degree of pleasantness/unpleasantness of the emotional state) and the arousal (i.e., the the degree of excitement of the stimuli, generally represented as a single axis increasing from "most-calm" to "most-exciting"). The dimensional approach is applicable across multiple domains to the entire spectrum of emotions: In fact, dimensional theories are focused on measurable details of every emotional experience (Hamann, 2012) and are very flexible: The same emotional label can vary depending on different levels of arousal and valence. For instance, happiness can be produced by a sunset or a smiling baby (both stimuli could have the same level of positive valence), but the arousal produced by the sunset is probably more moderate than that produced by a smiling baby.
James Russell, the father of the circumplex model of emotion, claimed that emotions are organized in a bidimensional circle, where the dimensions are valence and arousal (Russell, 1980). Valence and arousal would constitute the horizontal and vertical axis respectively. The center of this circle would represent a medium degree of arousal and a neutral valence. The circumplex model claims that emotional states can be portrayed through all the possible combinations of arousal and valence levels. Usually, this model has been used to test affective states or emotional stimuli (Remington, Fabrigar & Visser, 2000).
From an evolutionary perspective, it makes sense that the functional properties of memory evolved following the pressure of natural selection. The fact we remember an emotional stimulus, easier than a neutral stimulus, is a consequence of the relevance for survival or reproductive success of the emotional stimuli (Nairne, Thompson & Pandeirada, 2007).
Moreover, the effect that arousal has on memory has been proven by a large number of studies over several decades (Christianson, 1986;De Quervain 2015; Eysenck, 1976; Jeong, & Biocca 2011; Lang, Greenwald, Bradley, & Hamm, 1993; Leventon, Camacho, Ramos Rojas, & Ruedas 2018). The stimuli considered as highly arousing are remembered in better and in a higher number of occasions than the stimuli considered as neutral or low arousal. From a purely cognitive perspective; it was argued that the improvement in the memory of the stimuli with high arousal be due to more significant attention to the stimulus which resulted in a greater elaboration of the same, provoking, ultimately, better memory of the stimulus. (Neisser & Harsch, 1992). “What we call arousal-biased competition (ABC) is the notion that arousal (whether elicited by external stimuli, internal thoughts or stress hormones) modulates the strength of competing for mental representations, enhancing memory for items that dominate the contest for selective attention.” (Mather & Sutherland 2011). A large number of studies proved the trade-off effect of negative arousal stimuli on memory performance: They tend to increase the perceptual processing of central task-relevant elements, with the consequence to increase memory for the central emotional element and reduce memory for other simultaneously presented neutral stimuli (Gable & Harmon-Jones, 2010; Mather & Sutherland 2011). On the contrary, the effect of positive arousal on memory for associated neutral elements is still debated. Some results have suggested that positive arousing emotional element could attract
6
all the attention resources, thus impairing the memory for other neutral elements in the scenes (Mather at al., 2006; Waring & Kensinger, 2009, 2011). On the contrary, in other studies it seems that emotionally positive arousing events improved recognition memory for neutral information showed earlier. (Nielson & Powless, 2007; Nielson & Bryant, 2005). Similarly, other studies have found proof that exposure to positive high arousing stimuli expands cognitive processes, including the range of attention (Fredrickson & Branigan, 2005), which result in increased perceptual processing, with the consequence to enhance the processing of other peripheral elements of the emotional event. These results challenge the assumptions of the arousal-biased competition model, the idea that arousal regulates the competition between the mental representations, improving the memory recall for items that attract our attention.
Since the appearance of neuroimaging techniques, the objective has been not only to confirm this improvement in memory performance once again but also to find the neuronal substrates responsible for this effect.
The possibility that the processing of highly arousal stimuli required the activation of specific neuronal substrates seems plausible. Different neuroimaging techniques have identified the amygdala as a critical area facilitating memory enhancement of highly arousing emotional stimuli (McGaugh et al., 2000). Hamann's review, (2001) it is an extensive review on the role of the amygdala in the storage of emotional memory and several conclusions can be extracted as a summary suggesting that the amygdala is the main responsible for the relation between memory and emotion, without it, there are not emotional effects on memory. Moreover, the amygdala can modify explicit memory by regulating or improve the activity of other brain regions involved in memory. High levels of stress hormones can improve emotional arousal, due to its relationship with the amygdala. Finally, the regulatory effect of emotional arousal through the amygdala takes effect concretely on consolidation processes in memory zones like the hippocampus.
The left amygdala, the inferior prefrontal cortex and the left hippocampus are areas related to the successful recall of arousing stimuli (e.g. words; Kensinger & Corkin, 2004). Moreover, several other brain regions have been associated with the successful recall of emotional stimuli (e.g. words), such as the temporal, medial temporal lobe, medial prefrontal cortex, angular gyrus, posterior cingulate cortex, middle occipital gyrus, cerebellum, insula and middle temporal gyrus (Maratos et al., 2001; Hayama, Vilberg, & Rugg, 2012; Rugg & Vilberg, 2012; Hongkeun, 2016).
There are neuroimaging studies, where they compared patterns of brain activation associated with the retrieval of previously studied emotional and neutral pictorial material, that support a stronger activation of frontal, temporal and parietal areas with the successful recognition of items that were shown in high arousal contexts compared to the recognition in low arousal contexts (Dolan, Lanne, Chua, & Fletcher, 2000). Furthermore, items that were remembered from emotional contexts provoked BOLD activation at the retrieval in limbic areas like the amygdala, insula and cingulate cortex (LaBar, & Cabeza, 2006; Maratos & Rugg, 2001).
According to Magids, Sorfas & Leemon (2015), emotions are crucial to the relation between a company and their customer, emotions are crucial to the relation between a company and their customers.The development of an emotional bond with consumers is the key to a grand marketing strategy as well as substantial financial gain (Magids, Zorfas & Leemon, 2015). Traditional marketing research has historically been based on techniques such as interviews, questionnaires and focus groups. From these methods, they tried to predict the behaviour of the participants when buying products and services in real life. However, the effectiveness of these techniques has repeatedly been criticised. Castellion and Markham (2013) found that the rate of new products that were launched on the market and failed is around 40%. The low predictive power of traditional market research techniques is what has led to the
7
emergence of neuromarketing. This discipline is born from the union of marketing with psychology and neuroscience, and its objective is to link the knowledge of these fields. Its goal is to discover the neural processes underlying the thoughts, emotions and memories affecting the consumer during the purchase process, and relate this knowledge to the marketing concepts, such as branding, customer loyalty or subliminal advertising(Katarzyna, 2014).
The most characteristic example of the conjunction of these areas of study is the use of neuroimaging techniques, which allow us to know the reactions, at the brain level, of the consumer to different scenarios (Agarwal, 2015; Çakir, Çakar, Girisken & Dicle, 2018; Goodman et al., 2017; Daugherty, Hoffman, & Kennedy, 2016; Kenning & Plassmann, 2008; Leanza, 2017; Levy, Lazzaro, Rutledge & Glimcher, 2011; Pozharliev, Verbeke, & Bagozzi, 2017; Reimann, Zaichkowsky, Neuhaus, Bender & Weber, 2010; Smidts et al, 2014; Venkatraman, Clithero, Fitzsimons, & Huettel, 2012, Welber, 2013). These techniques will come to improve the techniques that have been used up to the present such as questionnaires, interviews and focus group. Several types of research have claimed the utility of neuromarketing techniques like fMRI, “Five features render fMRI experiments as more beneficial and reliable than the traditional approach of customer behavioural studies. These features are the minimisation of recall bias, reduction of cognitive bias, integration of the psychological and cognitive processes, assessment of human behaviour during unawareness and ability to distinguish between similarities and differences in the neurocognitive system” (Hsu & Cheng, 2018, p. 201).
These methods would be fundamental to understand, at a neuronal level, the effect of emotional stimuli in memory. However, the research that was being done in the marketing discipline was with a delay of more than ten years. For example, one of the first studies of the effect of emotional announcements and their subsequent recall dates back to the late 1980s, when it was tested the recall of super bowl’s ads in different arousal conditions, concluding recall of ads was negatively related to emotional intensity (Pavelchak, Antil & Munch, 1988). Starting in the 1990s, while neuroimaging techniques were about to revolutionize our perception of mental states, marketing would continue to be anchored in the purely cognitive paradigm (Lang, Dhillon & Dong, 1995; Ambler & Burne, 1999). It would not be until after 2009 when the first meaningful studies, which used neuroimaging techniques, appeared to answer questions specific to the discipline of marketing, correctly, how ads with high emotional content are remembered. (Langleben et al., 2009; Bakalash & Riemer, 2013). In this experiment "message sensation value" was measured across different ads to test how the level of "sensation" affects the posterior recall of the ads. The results showed greater activation in the amygdala for the successful retrieved ads compared to the not recognitized ads.
The goal of this study is to test the influence of positive emotional advertisements with different level of arousal in affecting memory for the products that they are promoting. I will investigate whether the high arousing positive advertisements would increase memory for the associated neutral products and identify the neural structures that support theses process as opposed to low arousing positive and neutral advertisements.
An fMRI experiment was designed, in which I measured the effect of high-arousing positive, low arousing positive and neutral advertisements in promoting the successful subsequent retrieval of different brands of snacks, which were embedded in them. The experiment was divided into an encoding phase and a recognition phase, where participants had to answer if they remember the products, different snacks well-known for the participants, that were shown in the encoding phase. My goal was to analyze if high arousing positive advertisements will facilitate the memory for the products compared with neutral and low-arousing advertisements. Moreover, I focused on what part of the brain were activated when the subjects were able to successfully remember the products. I explored if there was any
8
difference in the brain activation during the successful retrieval depending on the kind of ads in which the product was embedded (high-arousing positive, low arousing positive or neutral).
At the behavioral level, I expect that the positive scenes will improve memory performance because the participants will look at the full scene. On the other hand, if arousal will capture all the attention resources, reducing the exploration of the scene as a whole, the recall of ads will not increase with positive high arousing positive ads, but only with low-arousing positive ones. At the neuronal level, I expect that the successful recall of products embedded in positive emotional ads will produce increased activity in limbic areas (e.g. the amygdale; Hamann et al., 1999); and brain areas connected with the processing of reward (e.g. medial orbital prefrontal cortex, nucleus accumbens and caudate; Colibazzi et al., 2010; O'Doherty et al., 2003). As for the comparison of the products accompanied by high-arousal scenes versus low arousal positive scenes, I could expect greater activations in brain regions involved in sensory processing such as the temporo-occipital brain areas (e.g., Kirchhoff, Shapiro, & Buckner, 2005). The involvement of sensory areas during the encoding of high-arousing stimuli would be consistent with extensive evidence that high-arousing stimuli can enhance perceptual processing and tend to capture bottom-up attention largely than low-arousing stimuli (Anderson, 2005). The activation observed during the retrieval of products paired with high arousing scenes ads would be useful to clarify if the possible perceptual improvement with high arousing positive stimuli will facilitate or impair the memory for the peripheral products.
Methods and materials
Participants The subjects recruited were 26 Russian people, of whom, 13 were women, and 13 were men, with an average age of 23.4 years and within a range of age between 18 and 35 years. All the subjects were right-handed and did not suffer from any psychiatric or neurological disease, and all of them had normal vision. Before initiating the experiment, the participants gave their written consent. Due to technical problems, and the excessive movement of the participants during the scanner, three subjects were eliminated, so the final number of participants was 23 subjects (12 = females, 11 = males, mean age = 24.1, age range = 18-35). Stimuli
All the scenes selected for this study belong to the International Affective Pictorial System database, which is a “database of pictures designed to provide a standardised set of pictures for studying emotion and attention that has been widely used in psychological research (IAPS, Lang, Bradley & Cuthbert, 2008). The IAPS was developed by the National Institute of Mental Health Center for Emotion and Attention at the University of Florida. In 2005, the IAPS comprised 956 color photographs ranging from everyday objects and scenes − such as household furniture and landscapes − to extremely rare or exciting scenes − such as mutilated bodies and erotic nudes.”. From this database 180 pictures were selected: The selected scenes were divided into three types, 60 positive high arousing pictures, (Valence: M = 7.11, SD = 0.59, Arousal: M =6.11, SD = 0.54), 60 positive low-arousing pictures (Valence: M = 7.11, SD =0.63; Arousal: M = 3.75, SD = 0.56) and 60 neutral low-arousing pictures (Valence: M = 5.01, SD = 0.44, Arousal: M = 3.16 SD = 0.44). Given the special effect that human faces have on memory, each type of scenes contained the same number of pictures representing human
9
beings. Regarding the products that were accompanied by the scenes, 180 different types of snacks were used. All of them were Russian brands widely known in the country. Procedure
Encoding phase: Inside the fMRI scanner 90 (30 high arousal, 30 low arousals, 30 neutral) ads were shown to the participants. The ads shown consisted of the product superimpose on a scene. The size of the product in the image was 300x300 pixels, while the contextual scene had a size of 1280x1024 pixels. The products were randomly displayed in one of the four corners of the scene. The matching between the scenes and the products occurred randomly so that all the products were accompanied by all the three conditions (positive-high arousal, positive-low arousal and neutral). Each trial started with a warning signal for 500ms. Next, the product and the scene in which it was imposed, was presented during 3000ms. At the end of this period, a question was displayed on the screen "It is a good idea for a future advertisement?". The subjects answered this question by pressing one of the two response keys with the index finger of both hands, right for "yes" and left for "no". The subjects had a period of 3000 ms to answer the question. Between each image shown, a black cross on a grey background as inter-trial interval was. The inter-trial interval ranged from 8000 to10000 ms (see fig. 1).
 Fig. 1 (Example of the pictures that were shown to the participants during the encoding phase)
Recognition phase: This phase consisted of two fMRI sessions, each lasting 25 minutes with an interval of 5 minutes between the two sessions. During the two sessions, 180 images were shown, 90 images per session. These 180 images consisted of the 90 products that had been exposed during the encoding phase intermixed with 90 new products. Each trial began with the presentation of a fixation cross for 500ms. Next, the product (300x300 pixels) was shown in the center of the image, surrounded by a grey background, for 3000ms. During this period, a ”yes or no” question had to be answered by the participants about the fact if the had seen the product or nor. Next, the question “Are you sure?” was showed for 3000 ms and had to be answered with a “yes or no" answer. Between each trial, an inter-trial grey background was shown during a period ranging from 8000 to 10000 ms (see Fig.2).
Fig. 2 (Example of the pictures that were shown to the participants during the retrieval phase)
10
Magnetic resonance imaging
Neuroimaging data were collected using ascending interleaved slice acquisition with gradient echo (GRE)1 T2 * -weighted echo-planar imaging (EPI)2The sequence in a 3.0 Tesla Magnetom Verio equipped with a 32-channel head coil (Siemens; Erlangen, Germany). Scanning protocol parameters were as follows: TE3 = 30 ms, flip angle4 = 80 °, TR5 = 2280 m,
slice thickness = 3 mm, no gap, slice matrix = 64x64, number of axial slices = 35, FoV6 = 192 mm, Voxel resolution = 3x3d Magnetization-Prepared Rapid Gradient Echo 7(MP-RAGE) Sequence. High-resolution structural MRI data acquisition used a T 1 –weighted MP-RAGE sequence. Parameters were as follows: TE=2.47 ms; flip angle=9°; TR=1900 ms; slice thickness=0.5 mm; slice matrix=512x512x176; number of slices=176; FoV=256 mm; Voxel resolution=0.508x0.508x1mm. fMRI data analysis
All neuroimaging data were preprocessed and analysed using SPM12 (Statistical Parametric Mapping, Welcome Department of Cognitive Neurology, London, UK, implemented in MATLAB (Matlab 2015a, The Mathworks, Inc., Natick, MA, USA). Each subject underwent three fMRI sessions (one encoding session and two retrieval sessions for products), each consisting of 650 volumes. First four volumes of each session were discarded to allow for equilibrium effects, and all images for each session performed motion correction for head movements. Before preprocessing, a manual quality inspection using in-house software was done. Preprocessing was done in the following order; slice-timing correction to the first slice using a Fourier phase- shift interpolation method, head-motion correction with unwarping of B0 distortions, DARTEL normalisation (Ashburner 2007) using a 12-parameter affine transformation model to MNI anatomical space, and an 8mm FWHM Gaussian smoothing. DARTEL normalisation and smoothing was applied on the contrast images after intrasubject model estimation. For intrasubject modelling, a General Linear Model (GLM) with restricted maximum likelihood estimation was used.
I performed an event-related analysis on a voxel-by-voxel basis, for each subject, only on the retrieval phase, I measured the memory performance of the participants, according to their answers, hits and misses, in the three kinds of conditions, high positive arousal, low positive arousal and neutral. The model consisted of the following eight regressors of interest: the combination of the three levels (high positive arousal, low positive arousal, neutral) and the possible recognition or not of the product, plus the false alarms and correct rejections and separated regressor for the missing responses. All regressors were convolved with the
1GRE= A signal echo produced by reversing the direction of a magnetic field gradient or by applying balanced pulses of magnetic field gradient before and after a refocusing RF pulse so as to cancel out the position-dependent phase shifts that have accumulated due to the gradient. 2EPI= A single-shot gradient echo or spin echo imaging technique that collects acomplete 2D image data set with Cartesian k-space coverage from a single excitation. 3TE= Echo time. Time between the middle of exciting (e.g., 90°) RF pulse and middle of spin echo production.4Flip angle= Amount of rotation of the macroscopic magnetization vector produced by an RF pulse, with respect to the direction of the static magnetic field. 5TR= Repetition time. The period of time between the beginning of a pulse sequence and the beginning of the succeeding (essentially identical) pulse sequence. 6FoV= The rectangular region superimposed over the human body over which MRI data are acquired. 7MP-RAGE = Echo-train or RARE techniques speed image acquisition by applying a different phase-encoding to each echo, speeding image acquisition, but blending echoes with different T2 weightings into a single image set.
11
“canonical" hemodynamic response function as defined in SPM12 and the head movements were included as covariates. The high-pass filter had a cut-off at 128 s, and the autocorrelation model was global AR(1). Model estimations from each were taken to second-level random-effects analyses (one-sample t-tests) to account for interindividual variability. Statistical inferences were made on same areas with P ≤ 0.005, uncorrected for multiple comparisons.
Results Behavioural results
I have measured the performance of memory through the hit rate (the percentage of correct recognition of previously exposed products and the false alarm the percentage of false recognition of the new products that have not appeared during the encoding phase. The statistical analyses were carried out through the IBM SPSS software program, using the scores of the hit rate minus false alarms. A one way within subjects ANOVA with three levels (high positive arousal, low positive arousal and neutral) was performed.
Table 1. Successful memory retrieval on three conditions. Pos_HA (Positive high arousal), Pos_LA (Positive low arousal) and NEU (Neutral)
Conditions Corrected Scores (Hit rate – False alarm) Standard Error
Pos_HA 46.28 4.06 Pos_LA 44.39 4.50
NEU 42.65 3.70 Pos_HA (Positive high arousal), Pos_LA (Positive low arousal) and Neu (Neutral)
Although the number of products remembered successfully shows a trend for the products with high emotional content, the ANOVA analysis failed to reveal any statistically significant differences between the three levels (high arousing positive, low arousing positive and neutral scenes; [F (1,654, 36,383) = 0.833, p = 0.423]) suggesting that the products were not differentially remembered depending on the context in which they were embedded. Neuroimaging results
The fMRI analyses tested for areas showing an effect of successful memory during the retrieval phase. I compared the different brain activity for correct recognized products paired at the encoding paired with high arousing ads (POS HA), low arousing ads (POS LA) and neutral ads (NEU).
The results of the functional magnetic resonance, revealed, in the successful retrieval of products, if I compare the conditions of positive contexts against neutral contexts (Figure 3a) (POS>NEU), higher activity patterns in superior temporal gyrus, cerebellum, mid-temporal gyrus and occipital fusiform gyrus. These activations suggest visual processing in retrieval task for emotional items (Mather et al., 2006) and recovery of memories that had been conditioned in the past by highly emotional stimuli (Mueller & Pizzagalli, 2015). If the comparison is made between high arousal against low arousal conditions (Figure 3d) (HA>LA), I see higher activity
12
patterns in brain areas such as paracentral gyrus, precuneus gyrus, lateral occipital cortex, postcentral gyrus, Supplementary Motor Cortex and Insula. These activations suggest a process of recognition memory (Henson et al., 1999a, Konishi et al., 2000; Maratos et al., 2001; Rugg et al., 2002; Smith, Henson, Doland & Rugg, 2004), a successful recall of a stimuli, (Maratos et al., 2001; Hayama, Vilberg, & Rugg, 2012; Rugg & Vilberg, 2012; Kim, 2016) and the supervision of the information retrieval process. (see Table 2)
On the other hand, if the comparison is made between the conditions of a positive context of high arousal against a neutral context (Figure 3b) (HA>NEU), higher activity patterns are revealed in brain areas such as mid-temporal gyrus, superior temporal gyrus, supramarginal gyrus, occipital fusiform gyrus. These activations suggest the influence of emotion over episodic memory (Maddock, 1999) and recovery of memories that had been conditioned in the past by highly emotional stimuli (Mueller & Pizzagalli, 2015). Finally, if the comparison is made between the conditions of a positive context of low arousal against a neutral context (Figure 3c) (LA>NEU), greater activation is found in areas such as temporary mid gyrus, superior temporal gyrus, lateral occipital gyrus and gyrus calcarine. These activations suggest the influence of emotion over episodic memory (Maddock, 1999) and the influence of the amygdala, related to arousal, on the hippocampus (Ahs et al., 2011) (see Table 2).
a. b. c.
d. e. Fig. 3 Pictures of the regions activated during the successful retrieval of products. Fig 3a = Superior temporal gyrus (POS>NEU). Fig. 3b = Fusiform gyrus (HA>NEU). Fig. 3c = Cingulate gyrus (LA>NEU). Fig. 3d = Insula (HA>LA). Fig. 3e = Parahippocampal gyrus (LA>NEU).
13
Table2. Regions corresponding with the successful retrieval of products for each contrast of interest
Contrast Region Hemisphere MNI Coordinates
Cluster Size
T
POS>NEU
Superior temporal gyrus
Left -72
-26
-2 17 5.22
Cerebellum Left -48
-76
-34
54 4.09
Occipital fusiform gyrus
Right 26 -86
-18
9 3.50
Mid temporal gyrus Left -66
-28
-16
10 3.1
HA>LA
Gyrus paracentral Right 12 -10
46 42 3.99
Gyrus precuneus Left -10
-52
64 183 3.96
Lateral occipital cortex
Left -40
-90
-24
17 3.96
Postcentral gyrus Right 26 -34
48 160 3.64
Supplementary motor cortex
Left -14
0 44 29 3.50
Insula Right 42 -14
24 24 3.59
HA>NEU
Cingulate Gyrus Left -10
-16
32 7 3.23
Occipital fusiform gyrus
Left -28
-82
-4 2 3.07
Occipital pole Right 28 -94
32 1 2.92
Occipital fusiform gyrus
Left -30
-78
-2 1 2.92
LA>NEU
Cerebellum Left -26
-46
18 30 3.43
Parahippocampal gyrus
Left -14
-28
-18
8 3.27
Cingulate gyrus Left -12
-14
32 3 3.10
P < 0.005, uncorrected, voxel cluster extent = 0, L = left; R = right; POS = positive; HA = high arousal; LA = low arousal; NEU = neutral.
14
Discussion
The objective of this study was to clarify the influence that an emotional context of positive valence could have on the memory of a product that was embedded in that context. Different studies had tried to answer this question with very different results, while some studies showed an improvement in the recall of neutral products (Nielson & Powless, 2007; Nielson & Bryant, 2005), in other studies this effect did not occur, or it was even counterproductive (Buchanan & Lovallo, 2001; Cahill, Gorski, & Le, 2003; Smeets, Otgaar, Candel, & Wolf, 2008).
Regarding the behavioural results, no significant differences were found between the conditions of the experiment. There is a greater recall in the favourable high arousal condition, but it does not represent a statistically significant increase compared to the other conditions. This result has been surprising since the increase of memory for neutral stimuli in emotional contexts is an effect that has been replicated in numerous studies (Christianson, 1986, Bradley, Greewald, Petry & Lang, 1992, Lang, Dhillon & Dong, nineteen ninety-five). It was expected to obtain results in the line marked by the work of Fredrickson (2001), in which the positive emotion had a better global or expansive effect on the neutral stimulus, while I also test, this effect with the high and low arousal levels.
Among the possible reasons that could be responsible for the absence of this effect, three possible causes could be able to explain our lack of results. First, the size of the product compared to the size of the scene was small and in the phase of encoding, the product, instead of appearing in the middle, appeared randomly in one of the four corners of the screen. All this, added to the limiting conditions of a study with fMRI, could have affected our study.
Second, there are studies (LaBar and Cabeza, 2006, LaBar and Phelps, 1998) which affirm that, if there is an increase in the recall of a neutral stimulus due to the influence of an emotional context, a period of at least one day must pass between the encoding phase and the recognition phase. It seems that the effect needs some subsequent processing that does not arise immediately after the encoding phase, but later (Cunningham et al., 2014). Although arousal responses are often thought to account for emotional enhancement in long-term memory, these findings suggest that both an arousal response to the encoding and a subsequent period of sleep are needed to optimise selective emotional memory consolidation. One could argue that this period of 24 hours of which would be related to the critical role that the dream plays in the processing of memory.
Finally, for the effect of memory increase of a neutral stimulus by the influence of an emotional context to occur, there must be congruence between the context and the neutral stimulus. Otherwise, this effect does not occur (Madan et al., 2012). Since I have used conceptual scenes that were not related to snack brands that are popular with Russian consumers, it is possible that they perceived incongruity between the products they already knew and the scenes that served as context.
The recovery of emotional information has been associated with areas such as the amygdala and the orbitofrontal cortex. (LeDoux, 2000). Also, retrieval of items in emotional context, compared to neutral contexts, was associated with other brain areas that played a general role in memory recovery, such as the hippocampus. However, this has not happened in our experiment. Despite this fact, other brain areas involved in the processing of emotions have been activated during the experiment, such as the temporal cortex (Dolan, Lane, Chua & Fletcher, 2000), cingulate cortex (Fredrikson, Wik, Fischer & Andersson, 1995) and insula (Buchel, Dolan, Armony & Kriston, 1999).
Regarding the cingulate, several experiments (Maddock, 1999) have defended the idea that this brain region bridges the gap between emotion and episodic memory, this fact makes match our experiment in the conditions of high and low arousal compare to the neutral one. It
15
seems that although I did not obtain any activation of the amygdala, there was an emotional process affecting the episodic memory. Regarding the activation of the insula, is interpreted as the area in charge of the operations involved in the processing of salient stimuli (Henson et al., 1999). In our study, this area was activated at the high arousal context compare to the low arousal context. This fact could be interpreted asa tendency of improved performance of retrieval memory for the stimuli encoded in the high arousal salient context. Moreover, the activation of fusiform gyrus has been related to the recovery of memories that had been conditioned in the past by highly emotional stimuli (Mueller & Pizzagalli, 2015). This fact matches our results since the activation of the fusiform gyrus appeared in the favourable high arousal condition compare to the neutral condition, but it was not activated in the low arousal condition compare to the neutral condition. Regarding the activation of the parahippocampal gyrus, this area has an important role in the retrieval process (Maratos et al., 2001). However, is essential to say that this activation happened in the low arousal condition compared to the neutral condition, but strangely, there was no difference between the neutral condition and the high arousal condition.. Finally, the activation of areas such as, precuneus gyrus, gyrus postcentral, paracentral gyrus match to the activated areas other relevant event-related fMRI studies of recognition memory (Henson et al., 1999a, Konishi et al., 2000; Maratos et al., 2001; Rugg et al., 2002; Smith, Henson, Doland & Rugg, 2004).
In short, I can say that the recovery of information stored in high arousing emotional contexts, activated these brain areas to a greater extent than the recovery of information from neutral contexts, except to the hippocampus that was involved solely during the retrieval of product paired with low arousing positive scenes.Regarding the non-activation of the amygdala, there are studies which claim the activation of the amygdala is directly related to the amount of accurate detail recognised (Kensinger & Schacter, 2007). As a consequence, it is conceivable that the small size of the product's image shown, compared to the contextual scene, would provoke a low the level of detail remembered, and as a result, no activation of the amygdala.. On the other hand, the majority of studies on emotional memory have linked the activation of the amygdala with high arousing negative stimuli whereas in our study I have only used positive arousing stimuli (Murty et al. 2010).
The conclusions that can be derived from this study are several. First, the results I have obtained compared to the existing literature on the subject highlights the complexity of the effect that the valence and arousal of a stimulus have on a neutral stimulus. There are no formal rules in this influence, and many factors come into play when trying to replicate this effect, so it is essential to take into account any aspect of the study that is carried out, since what may seem a minor aspect at the beginning, maybe fundamental at the end of the experiment. The period between encoding and recognition is an excellent example of what can happen.
On the other hand, the role of positive stimuli and neurological correlates is also questioned, especially if I try to find a relationship of equality concerning negative stimuli, although both are emotional stimuli, it is possible to activate different neural circuits independent of each other. An example of this is the low proportion of studies that demonstrate the activation of the amygdala in the memory of positive stimuli compared to a large number of studies that prove its activation in the memory of negative stimuli.
From the perspective of marketing, none of these results can be strange, the factors that influence a successful advertising campaign are as numerous as those that can influence academic research, and that is why it would be interesting to measure "true" ads. With this, I would generate a greater congruence in the emotional context/neutral stimulus relationship, in addition to making neuromarketing not only a branch of applied neuroscience but an area of knowledge by itself.
16
References Agarwal, S. (2015). Introduction to Neuromarketing and Consumer Neuroscience, Journal of
Consumer Marketing, 32, 302-303.
Åhs, F., Palmquist, M., Pissiota, A., Appel, L., Frans, Ö., Liberzon, I., Furmark, T., &
Fredrikson, M. (2011). Arousal modulation of memory and amygdala-
parahippocampal connectivity: A PET-psychophysiology study in specific phobia.
Psychophysiology, 48, 1463-1469.
Ambler, T., & Burne, T. (1999). The Impact of Affect on Memory of Advertising. Journal of
Advertising Research, 39, 25–34.
Bakalash, T., & Riemer, H. (2013). Exploring Ad-Elicited Emotional Arousal and Memory
for the Ad Using fMRI. Journal of Advertising, 42, 275-291.
Bradley, M. M., Greenwald, M. K., Petry, M. C., & Lang, P. J. (1992). Remembering
pictures: Pleasure and arousal in memory. Journal of Experimental Psychology:
Learning, Memory, & Cognition, 18, 379-390.
Braeutigam, S., Rose, S., Swithenby, S., & Ambler, T. (2004). The distributed neuronal
systems are supporting choice-making in real-life situations: Differences between
men and women when choosing groceries detected using magnetoencephalography.
European Journal of Neuroscience, 20, 293-302.
Büchel C., Dolan R.J., Armony J.L., & Friston K.J. (1999). Amygdala-Hippocampal
involvement in human aversive trace conditioning revealed through event-related
functional magnetic resonance imaging. Journal of Neuroscience, 19, 10869-10876.
Cabanac, M. (2002). What is emotion? Behavioural Processes, 60(2), 69-83.
17
Castellion, G., & Markham, S. (2013). Perspective: New Product Failure Rates: Influence of
Argumentum ad Self-Interest. Journal of Product Innovation Management, 30, 976-
979.
Christianson, S. (1986). Effects of positive emotional events on memory. Scandinavian
Journal of Psychology, 27, 287-299.
Cunningham, T. J., Crowell, C. R., Alger, S. E., Kensinger, E. A., Villano, M. A., Mattingly,
S. M., & Payne, J. D. (2014). Psychophysiological arousal at encoding leads to
reduced reactivity but enhanced emotional memory following sleep. Neurobiology of
Learning and Memory, 114, 155-164.
Çakir, M., Çakar, T., Girisken, Y., & Yurdakul, D. (2018). An investigation of the neural
correlates of purchase behaviour through fNIRS. European Journal of Marketing, 52,
224-243.
Damasio, A. (1998). Emotion in the perspective of an integrated nervous system. Brain
Research Reviews, 26, 6-83.
Daugherty, T., Hoffman, E., & Kennedy, K. (2016). Research in reverse: Ad testing using an
inductive consumer neuroscience approach. Journal of Business Research, 69, 3168-
3176.
De Quervain, D. (2015). The impact of emotional arousal on human memory.
Psychoneuroendocrinology, 61, 20.
Dolan, R. J., Lane, R., Chua, P., & Fletcher, P. (2000). Dissociable temporal lobe activations
during emotional episodic memory retrieval. Neuroimage, 11, 203–209.
Dolcos, F., LaBar, K. S., & Cabeza, R. (2004) Interaction between the amygdala and the
medial temporal lobe memory system predicts better memory for emotional events.
Neuron, 42, 855–863.
Eysenck, M. (1976). Arousal, learning, and memory. Psychological Bulletin, 83(3), 389-404.
18
Fredrikson, M., Wik, G., Fischer, H., & Andersson, J. (1995). Affective and attentive neural
networks in humans: A PET study of Pavlovian conditioning. NeuroReport, 7, 97-
101.
Goodman, A., Wang, Y., Kwon, W., Byun, S., Katz, J., & Deshpande, G. (2017). Neural
Correlates of Consumer Buying Motivations: A 7T functional Magnetic Resonance
Imaging (fMRI) Study. Frontiers in Neuroscience, 11, 512.
Hamann, S. (2001). Cognitive and Neural Mechanisms of Emotional Memory. Trends in
Cognitive Science, 5 (9), 394–400.
Hamann, S. (2012). What can neuroimaging meta-analyses really tell us about the nature of
emotion?. Behavioral and Brain Sciences, 35, 150-152.
Harris, J., Ciorciari, J., & Gountas, J. (2018). Consumer neuroscience for marketing
researchers. Journal of Consumer Behaviour, 17, 239-252.
Hayama, H. R., Vilberg, K. L., & Rugg, M. D. (2012). Overlap between the neural correlates
of cued recall and source memory: evidence for a generic recollection network?
Journal of cognitive neuroscience, 24, 1127-1137.
Henson, R., Rugg, M., Shallice, T., Josephs, O., & Dolan, R. (1999). Recollection and
familiarity in recognition memory: An event-related functional magnetic resonance
imaging study. Journal of Neuroscience, 19, 3962-3972.
Hongkeun, K. (2016). Default network activation during episodic and semantic memory
retrieval: a selective meta-analytic comparison. Neuropsychologia, 80, 35-46.
Hsu, M., & Cheng, J. (2018). FMRI neuromarketing and consumer learning theory. European
Journal of Marketing, 52, 199-223.
Huys, Q., & Renz, D. (2017). A Formal Valuation Framework for Emotions and Their
Control. Biological Psychiatry, 82, 413-420.
19
Jeong, E., & Biocca, F. (2011). Are there optimal levels of arousal to memory? Effects of
arousal, centrality, and familiarity on brand memory in video games. Computers in
Human Behavior, 28, 285-291.
Katarzyna, W. (2014). Neuromarketing – its role in building of brand, introduction of
products innovation, and advertising messages. Marketing of Scientific and Research
Organisations, 11, 81-101.
Kenning, P., & Plassmann, H. (2008). How Neuroscience Can Inform Consumer Research.
Neural Systems and Rehabilitation, 16, 532-538.
Kensinger, E. A., & Corkin, S. (2004). Two routes to emotional memory: Distinct neural
processes for valence and arousal. Proceedings of the National Academy of Sciences
of the United States of America, 101, 3310–3315.
Kensinger, E. A., & Schacter, D. L. (2007). Remembering the specific visual details of
presented objects: Neuroimaging evidence for effects of emotion. Neuropsychologia,
45, 2951-2962.
Labar, K., & Cabeza, R. (2006). Cognitive neuroscience of emotional memory. Nature
Reviews Neuroscience, 7, 54-64.
LaBar, K., & Phelps, E. (1998). Arousal-Mediated Memory Consolidation: Role of the
Medial Temporal Lobe in Humans. Psychological Science, 9, 490-493.
Lang, A., Dhillon, K., & Dong, Q. (1995). The effects of emotional arousal and valence on
television viewers' cognitive capacity and memory. Journal of Broadcasting and
Electronic Media, 39, 313-327.
Lang K., Bradley, M., & Cuthbert, B. (2008). International Affective Picture System (IAPS):
Affective Ratings of Pictures and Instruction Manual (Rep. No. A-8). Technical
Report A-8.
20
Langleben, D., Loughead, W., Ruparel, K., Hakun, G., Busch-Winokur, S., Holloway, B.,
Strasser, A., Cappella, N., & Lerman, C. (2009). Reduced prefrontal and temporal
processing and recall of high "sensation value" ads. Neuroimage, 46, 219-225.
LeDoux, J. (2000). Emotion Circuits in the Brain. Annual Review of Neuroscience, 23, 155-
184.
Leanza, F. (2017). Consumer Neuroscience: The traditional and VR TV Commercial.
Neuropsychological Trends, 21, 81-90.
Lee, N., Brandes, L., Chamberlain, L., & Senior, C. (2017). This is your brain on
neuromarketing: Reflections on a decade of research. Journal of Marketing
Management, 33, 878-892.
Leventon, S., Camacho, L., Ramos Rojas, D., & Ruedas, A. (2018). Emotional arousal and
memory after deep encoding. Acta Psychologica, 188, 1-8.
Levy, I., Lazzaro, S., Rutledge, R., & Glimcher, P. (2011). Choice from non-choice:
Predicting consumer preferences from blood oxygenation level-dependent signals
obtained during passive viewing. The Journal of Neuroscience, 31, 118-25.
Lin, M., Cross, S., Jones, W., & Childers, L. (2018). Applying EEG in consumer
neuroscience. European Journal of Marketing, 52, 66-91.
Madan, R., Caplan, B., Lau, M., & Fujiwara, E. (2012). Emotional Arousal Does Not
Enhance Association-Memory. Journal of Memory and Language, 66, 695-716.
Maddock, R. (1999). The retrosplenial cortex and emotion: New insights from functional
neuroimaging of the human brain. Trends in Neurosciences, 22, 310-316.
Magids, S., Zorfas, A., & Leemon, D. (2015). The new science of customer emotions.
Harvard Business Review, 76, 66-74.
21
Maratos, E. J., Dolan, R. J., Morris, J. S., Henson, R. N. A., & Rugg, M. D. (2001). Neural
activity associated with episodic memory for emotional context. Neuropsychologia
39, 910–920.
Maratos, E. J., & Rugg, M. D. (2001). Electrophysiological correlates of the retrieval of
emotional and non-emotional context. Journal of Cognitive Neuroscience. 13, 877–
891.
Mather, M., Mitchell, J., Raye, L., Novak, L., Greene, J., & Marcia K. (2006). Emotional
Arousal Can Impair Feature Binding in Working Memory. Journal of Cognitive
Neuroscience, 18, 614-625.
Mather, M., & Sutherland, M. (2011). Arousal-Biased Competition in Perception and
Memory. Perspectives on Psychological Science, 6, 114-133.
Mueller, E., & Pizzagalli, D. (2015). Human remote fear memory recall rapidly activates
fusiform gyrus one year after conditioning. Psychophysiology, 52, S19.
Murty, P., Ritchey, M., Adcock, R., & Labar, K. (2010). fMRI studies of successful
emotional memory encoding: A quantitative meta-analysis. Neuropsychologia,
48(12), 3459-3469.
Nairne, J. S., Thompson S. R., & Pandeirada, J. N. S. (2007). Adaptive memory: survival
processing enhances retention. Journal of Experimental Psychology: Learning,
Memory and Cognition, 33, 263–273.
Neisser, U., & Harsch, N. (1992) Phantom flashbulbs: false recollections of hearing the news
about Challenger E. Winograd, U. Neisser (Eds.), Cambridge University Press.
Nielson, A., & Bryant, T. (2005). The effects of non-contingent extrinsic and intrinsic
rewards on memory consolidation. Neurobiology of Learning and Memory, 84, 42-48.
22
Nielson, A., & Powless, M. (2007). Positive and negative sources of emotional arousal
enhance long-term word-list retention when induced as long as 30 min after learning.
Neurobiology of Learning and Memory, 88, 40-47.
Ohme, R., Reykowska, D., Wiener, D., Choromanska, A., Reimann, M., & Schilke, O.
(2009). Analysis of Neurophysiological Reactions to Advertising Stimuli by Means of
EEG and Galvanic Skin Response Measures. Journal of Neuroscience, Psychology
and Economics, 2, 21-31.
Pavelchak, M. A., Antil, J. H., & Munch, J. M. (1988). The Super Bowl: An Investigation
into the Relationship Among Program Context, Emotional Experience, and Ad Recall.
Journal of Consumer Research, 15, 360-367.
Poldrack, R. (2008). The role of fMRI in Cognitive Neuroscience: Where do we stand?
Current Opinion in Neurobiology, 18, 223-227.
Pozharliev, R., Verbeke, W., & Bagozzi, R. (2017). Social Consumer Neuroscience:
Neurophysiological Measures of Advertising Effectiveness in a Social Context.
Journal of Advertising, 46(3), 351-362.
Reimann, M., Schilke, O., Weber, B., Neuhaus, C., & Zaichkowsky, J. (2011). Functional
magnetic resonance imaging in consumer research: a review and application,
Psychology and Marketing, 28, 608-637.
Reimann, M., Zaichkowsky, J., Neuhaus, C., Bender, T., & Weber, B. (2010). Aesthetic
package design: A behavioural, neural, and psychological investigation. Journal of
Consumer Psychology, 20, 431-441.
Remington, N. A., Fabrigar, L. R., & Visser, P. S. (2000). Re-examining the circumplex
model of affect. Journal of Personality and Social Psychology. 79, 286–300.
23
Reutskaja, E., Nagel, R., Camerer, C., & Rangel, A. (2011). Search Dynamics in Consumer
Choice under Time Pressure: An Eye-Tracking Study. American Economic Review,
101, 900-926.
Rugg, M. D., & Vilberg, K. L. (2013). Brain networks underlying episodic memory retrieval.
Current opinion in neurobiology, 23, 255-260.
Russell, J. A. (1980). A circumplex model of affect. Journal of Personality and Social
Psychology, 39(6), 1161-1178.
Smidts, A., Hsu, M., Sanfey, A., Boksem, G., Ebstein, M., Huettel, A., &Yoon, R. (2014).
Advancing consumer neuroscience. Marketing Letters, 25, 257-267.
Solnais, C., Andreu-Perez, J., Sánchez-Fernández, J., & Andréu-Abela, J. (2013). The
contribution of neuroscience to consumer research: A conceptual framework and
empirical review. Journal of Economic Psychology, 36, 68-81.
Stark, C., & Squire, L. (2000). fMRI activity in the medial temporal lobe during recognition
memory as a function of study-test interval. Hippocampus, 10,329-337.
Venkatraman, V., Clithero, J., Fitzsimons, G., & Huettel, S. (2012). New scanner data for
brand marketers: How neuroscience can help better understand differences in brand
preferences. Journal of Consumer Psychology, 22, 143-153.
Welberg, L. (2013). Neuroimaging: Evaluating ads with fMRI. Nature Reviews
Neuroscience, 14, 379-379.
