Scenting the Supermarket and Mixing the Music
The effect of atmospheric cues on consumer behavior
Jolein Hondsmerk, 11884916 Master’s Thesis, University of Amsterdam
Graduate School of Communication Research Master’s Communication Science
Supervisor:
Dr. M.H.C. Meijers (University of Amsterdam) June 24th, 2019
2
Abstract
Using music or ambient scent in a shopping environment is becoming increasingly popular among retailers, but the effects on consumer behavior are largely unknown. Especially the combined effect of music and ambient scent has been largely ignored in atmospheric research. This study addressed this gap by conducting an experiment (N = 87) in a controlled environment with an online supermarket that was specifically developed for this study. A website was build that resembled the webpage of the Dutch supermarket Jumbo, in which one could navigate like in a real online supermarket. A stimulating environment (i.e., stimulating music and ambient scent) was compared with a relaxing environment (i.e., relaxing music and the absent of scent) in their effect on consumers’ time spend, product choices, and overall satisfaction. Two underlying mechanisms for these effects were
proposed: self-control and arousal, which were innovatively measured with both implicit and explicit measures. The findings showed that a stimulating consumer environment does not result in different time spend, product choices, and overall satisfaction compared to a relaxing consumer environment. No support was found for arousal and self-control as underlying mechanisms. Suggestions for future research into atmospheric marketing are made.
3 With the increasing competition in marketing, retailers need to find creative ways to stand out from their competitors. Therefore, unique store atmospheres are created, using atmospheric cues like color, lighting, music, and scent (Turley & Milliman, 2000). Travel agency Thomson, for example, disperses a coconut scent in their stores to persuade consumers to book their holiday (Roberts, 2008). Abercrombie & Fitch plays exciting in-store music that relates to their young public and reflects a carefree attitude (Morrison & Beverland, 2003). Such atmospheric cues can affect consumers’ shopping behavior (Turley & Milliman, 2000).
Music was shown to influence several behavioral elements, like product choice (North, Hargreaves, & McKendrick, 1997), purchase intention (Michel, Baumann, & Gayer, 2017), and consumers’ emotions (Lee, Henderson, & Shum, 2004). Ambient scent has been researched noticeably less (Leenders, Smidts, & El Haji, 2019). No consensus has been reached about the effects of ambient scent, as some researchers reported strong effects, for example on product evaluations (Bosmans, 2006), while others showed mixed results or no measurable effects at all, for example on retention time and spending (Teller & Dennis, 2012).
Remarkably, the combined effect of music and ambient scent on consumers’ shopping behavior has largely been ignored (Morrison et al., 2011). As even minor changes in retail atmospheres are capable of influencing consumer behavior (Turley & Milliman, 2000), major changes like applying both music and scent are important to examine. To address this current gap, an experiment combining music and ambient scent is conducted in a controlled
environment using a specially developed online supermarket. Three behavioral elements are studied: consumers’ product choices (i.e., healthy or unhealthy), their time spent in the supermarket, and their overall satisfaction with the shopping experience. Two underlying mechanisms are proposed to mediate the effect of music and ambient scent on these variables: self-control and arousal. Music and scent are expected to lower self-control through ego-depletion and increase arousal. It is predicted this will lead to a longer time spend, a higher
4 satisfaction, and the purchase of more unhealthy products. While previous research often relies on one type of measurement, this study uses implicit measures for arousal and self-control, and behavioral measures for time spend and product choice. This strengthens the methodology, as biases like social desirability and memory bias are avoided (Hefner et al., 2011). This, combined with the novel conjunction of two important atmospheric cues, gives this study a high scientific relevance.
This study can also benefit retailers, as traditional marketing tools might nowadays not be enough to gain a competitive advantage. Gaining insights into the advantages of
atmospheric cues can help retailers create a unique store atmosphere, benefiting their market position. Good examples are Harrods and Bloomingdales, who specifically developed
ambient scents for this purpose (Doucé & Janssens, 2013). Lastly, the results of this study can empower consumers in their product choices. As consumers are often unaware of the
influence of ambient cues on their shopping behavior, making them aware of the processes at play can help consumers make decisions without being influenced by ambient cues.
In sum, this study examines the influence of music and scent on consumer behavior. The accompanying research question is: “What is the influence of music and ambient scent on consumers’ time spend in a supermarket, their product choices and their overall satisfaction with the shopping experience?”.
Theoretical background Explaining the Effects of Atmospheric Marketing
Atmospheric marketing is not new and has been successfully used by retailers for decades (Doucé & Janssens, 2013). The usage of atmospheric cues can help create the desired shopping environment (e.g., relaxing environments for expensive stores or stimulating
environments for stores with a high flow of customers). Sound and scent are powerful cues for this, as customers perceive a store’s atmosphere through their senses (Doucé & Janssens,
5 2013; Bäckström & Johansson, 2006). Furthermore, research shows that customers’ behavior is influenced by store atmospherics, even if they are not actively aware of their exposure to it (Morrin & Ratneshwar, 2000). Creating a pleasurable consumption environment through atmospheric marketing is therefore an important marketing strategy for retailers and can be considered a key factor to success (Turley & Milliman, 2000; Pan & Zinkhan, 2006).
In explaining the effects of atmospheric cues on consumers’ (shopping) behavior, scholars primarily rely on the principles of environmental psychology. This framework sees human behavior as a holistic response to all the stimuli in the environment (Mattila & Wirtz, 2001). Applied to the retail context, relations between stimuli in the shopping environment and consumers’ behavioral responses are central (Vida, Obadia, & Kunz, 2007). The S-O-R model explicates how such stimuli can influence human behavior (Mehrabian & Russell, 1974). In this model, the stimulus (S) is the store’s atmosphere, the organism (O) are the internal evaluations of consumers, and the response (R) are the behavioral responses to these evaluations (Doucé & Janssens, 2013). Thus, consumers’ internal evaluations (i.e., emotional states, like arousal) are considered to be mediators between environmental stimuli and
consumer behavior (Mehrabian & Russell, 1974).
In line with the S-O-R model, the present study considers arousal as a mediator, which refers to the extent that someone is stimulated by the atmospheric cues (Vida et al., 2007). Furthermore, self-control is introduced as a mediator for the effect of atmospheric cues on consumer behavior.This study thus applies the S-O-R model in the following way: it examines the effects of music (S) and ambient scent (S) in a supermarket environment on consumers’ arousal (O) and self-control (O) and their behavioral responses to it, namely time spend (R), product choice (R), and overall satisfaction with the shopping experience (R).
Previous research into in-store music examined several characteristics of music, like its emotional valence (Mattila & Wirtz, 2001). The current study focusses on the physical
6 dimension (i.e., tempo, volume, and rhythm; Bruner, in Mattila & Wirtz, 2001). Research on ambient scent generally considers two types of scents: those inherent to a product (e.g., scented candles) and ambient scents (Doucé & Janssens, 2013). The latter is the focus of this study and can be defined as “a scent not emanating from a particular object, but present in the environment” (Spangenberg, Crowley, & Henderson, 1996, p.67). An example is the smell of pine, cinnamon, and spiced cider in stores during the Christmas holiday season (Spangenberg, Grohmann, & Sprott, 2005). This study focusses on the effect of the absence versus presence of ambient scent.
Atmospheric Research into Behavioral Responses
Three indirect behavioral responses to the presence of music and ambient scent are examined in this study. Firstly, the time consumers spent in a shopping environment. This is of utmost importance for retailers as a longer shopping time gives more opportunities to sell their products (Morrison et al., 2011). Secondly, consumers’ product choices, with the focus on the purchase of healthy or unhealthy products, which is important as obesity numbers keep rising (Leefstijlmonitor, 2018). Thirdly, consumers’ overall satisfaction, which can be defined as “the favorability of the individuals’ subjective evaluation of the various outcomes and experiences associated with using or consuming it [the service]” (Shin & Elliott, 2001, p.4). This is important for retailers as a satisfying experience is crucial in guaranteeing consumers’ return (Chang, 2000; McQuitty, Finn, & Wiley, 2000).
Existing atmospheric research gives support for the notion that music and scent can influence these three variables. For the effect on time spend, Milliman (1982) found that consumers tend to adjust their shopping pace to the music tempo, which influences their time spent and their money spent (Sullivan, 2002; Wilson, 2003). Research also suggests that the presence of fruity or flowery ambient scents increase consumers’ time spend (Knasko, 1989; Guéguen & Petr, 2006; Stöhr, 1998), but it is not clear why this effect occurs.
7 Secondly, an influence on product choice was found by North et al. (1999). They found that in-store German (vs. French) music increases the sales for German (vs. French) wines. Other studies show that the presence of ambient scent lead to a higher variety seeking in products (Madzharov, Block, & Morrin, 2015), and the purchase of products that are congruent with the scent (Spangenberg et al., 2006).
Thirdly, consumers’ overall satisfaction is also expected to be influenced by music and scent. Studies showed a higher satisfaction among patients (Cruise et al., 1997) and visitors of restaurants (Wansink & Ittersum, 2012) when soft music was played, as the environment was perceived as more relaxing. Other studies showed an increase in overall satisfaction when a pleasant ambient scent was present in the environment (Spangenberg et al., 1996), but a decrease in overall satisfaction when it was perceived as unpleasant (Bone & Ellen, 1999).
Overall, research suggests that music and scent can influence time spend, product choice, and overall satisfaction. However, results are quite fragmented, inconclusive, and sometimes even contradicting (see meta-analysis by Roschk, Loureiro, & Breitshol, 2017). Moreover, research mostly focusses on the separate effects of either music or ambient scent. The combined effect of music and ambient scent has barely been examined, and it is not clear how this combination can influence consumers’ behavioral responses.
The present study suggests that the combination of music and ambient scent in a shopping environment can either lead to a stimulating environment (i.e., if stimulating music and a stimulating scent are present) or a relaxing environment (i.e., if relaxing music is present and scent is absent). This is expected to influence two important mental states of consumers: self-control and arousal. Self-control and arousal are therefore considered to be mediators for the effects of music and ambient scent on time spend, product choice, and overall satisfaction, which will be further explained in the following paragraphs. For an illustration, see the conceptual model (Fig. 1).
8
Self-control as a Mediator
The first proposed underlying mechanism for the effect of music and scent on consumer behavior is self-control. In short, it is suggested that the presence of music and scent creates a relaxing or stimulating environment, which influences consumers’ self-control. This is consequently expected to influence consumers’ time spend and product choices.
Self-control is an important human process as it helps to adhere to rules and
regulations, and can be defined as “the ability to control or override one’s thoughts, emotions, urges, and behavior” (Gailliot et al., 2007, p.325). Successfully regulating oneself has been linked to multiple desirable consequences, like better mental health and healthier lifestyles, while failing in self-control has been associated with negative life outcomes, like
uncontrollable cravings and impulsive buying behavior (Schmeichel & Vohs, 2009; Mattila & Wirtz, 2008).
Interestingly, self-control is a limited resource, and humans can deplete it. Research suggests that engaging in an act of control depletes the internal energy needed for self-control, such that following efforts at self-control are impaired (Galliot et al., 2007). This is also called ‘ego-depletion’ (Schmeichel & Vohs, 2009). To illustrate, scholars showed that participants who had to resist eating a cookie, had more trouble at the subsequent task that required self-control than participants who did not have to resist eating the cookie
(Baumeister et al., 1998). Even small acts of self-control can thus deplete the energy resource. In this study, the presence of stimulating music and ambient scent is suggested to diminish consumers’ amount of self-control. This is due to the fact that stimulating music and ambient scent create a very stimulating shopping environment (Mattila & Wirtz, 2008). Research showed that when an environment is perceived as very stimulating or over-stimulating (i.e., more over-stimulating than desired, for example when music and scent are perceived to be incongruent with each other (Spangenberg et al., 1996)), consumers
9 experience a higher level of excitement (Mattila & Wirtz, 2008). This requests an initial act of self-control, which consequently leads to a momentary loss of self-control during shopping (Mattila & Wirtz, 2008). Thus, a stimulating environment (i.e., with stimulating music and scent) is expected to deplete ones’ self-control. On the other hand, relaxing store
environments (i.e., with relaxing music and the absence of scent) are not perceived as over-stimulating. These environments do therefore not lead to higher levels of excitement and are not expected to deplete consumers’ self-control (Mattilla & Wirtz, 2008).
The loss in self-control due to a stimulating environment is expected to consequently influence consumers’ time spend in the store, as a loss in self-control diminishes the ability to think orderly and rationally about decisions (Tice, Batslavsky, & Baumeister, 2001). This will cause consumers to have a longer shopping time as they have to think longer about their purchases (Tice et al., 2001). Also, previous research showed that a highly stimulating
shopping environment leads consumers to engage in approach behaviors, like lingering in the store (Roehm & Roehm, 2005). Therefore, it is expected that consumers will spend more time in the store in which stimulating music and ambient scent is present, compared to a store with relaxing music and the absence of scent:
H1. Stimulating (vs. relaxing) music combined with the presence (vs. absence) of
ambient scent will deplete consumers’ self-control, which will consequently lead to more time spend in the supermarket.
Furthermore, it is suggested that a depleted state of self-control can also influence one’s product choices. Earlier research showed that a depleted state of self-control influences consumers’ impulsive buying or eating behavior. For example, participants on a diet who regulated themselves not to eat a cookie, could in a subsequent task regulate themselves to a lesser extent and ate more than people who were not on a diet, due to their depleted self-control after the first task (Vohs & Heatherton, 2000). Also, Mattila and Wirtz (2008) showed
10 that consumers in a stimulating environment engage in more impulsive buying behavior than consumers who are not in a stimulating environment, as the latter does not lead to a depleted self-control. A stimulating environment thus leads to a depleted self-control, which enhances the probability of impulsive eating or impulse purchases. Earlier research shows that
impulsive buying behavior most often concerns the purchase of unhealthy products. According to Verplanken et al. (2005), consumers who engage in impulsive buying are particularly vulnerable to unhealthy products as they are easily available and often portrayed in a way that triggers a quick purchase. A consumer with a depleted self-control is extra susceptible to these triggers (Verplanken et al., 2005).
In line with the literature, it is expected that consumers in a highly stimulating environment (i.e., in the presence of stimulating music and ambient scent) buy a higher amount of unhealthy products than consumers in a relaxing environment (i.e., in the presence of relaxing music and no ambient scent):
H2. Stimulating (vs. relaxing) music combined with the presence (vs. absence) of
ambient scent will deplete consumers’ self-control, which will consequently lead to the purchase of a higher amount of unhealthy products.
Arousal as a Mediator
The second underlying mechanism that is proposed for the effect of music and scent on consumer behavior is arousal. The presence of music and ambient scent is again expected to lead to a relaxing or stimulating environment, and it is suggested this will influence consumers’ arousal. This is consequently expected to influence consumers’ time spend and overall satisfaction.
Arousal is the extent to which someone is stimulated, for example by the store
environment (Vida et al., 2007). Music and ambient scent are expected to positively influence consumers’ arousal. Earlier research showed that the presence of stimulating music (i.e.,
11 music with a high tempo) increases the amount of arousal among consumers (Kellaris & Kent, 1991). According to Mehrabian and Russell (1974), this is due to the fact that music with a high tempo heightens the complexity of the environment, which leads to a higher amount of information that needs to be processed by the consumer. This consequently heightens the arousal one experiences. It is suggested that adding ambient scent to the environment heightens the complexity of it even more, as consumers receive another cue they need to process. Therefore, it seems logical to assume that an environment with stimulating music and ambient scent is highly arousing (i.e., a stimulating environment). An environment with relaxing music and no ambient scent on the other hand, is not expected to be arousing (i.e., a relaxing environment).
The present study thus suggests that stimulating music and ambient scent positively influence consumers’ arousal. Consequently, this is expected to positively influence the time consumers spend in the shopping environment. This expectation is in line with the S-O-R-model (Mehrabian & Russell, 1974), which predicts that a pleasantly arousing environment produces approach behaviors, like lingering longer in the store (Spangenberg et al., 2006; Roehm & Roehm, 2005). On the other hand, an unpleasant arousing environment can produce avoidance behavior (Spangenberg et al., 2006). Thus, consumers should perceive the
environment as pleasantly arousing to increase their time spend. Research shows that arousal is only perceived as pleasant if the music and ambient scent are congruent with each other (Mattila & Wirtz, 2001). There should thus be a fit between the music and scent to positively influence consumers’ time spend in the store (Roehm & Roehm, 2005).
To take this into account, this study kept music and ambient scent congruent in terms of their arousal for both conditions, by combining stimulating music with a stimulating ambient scent and relaxing music with the absence of ambient scent. The music and scent are thus congruent and therefore both conditions are expected to be pleasantly arousing.
12 However, this effect is expected to be stronger for the stimulating condition, as stimulating music and scent produce more arousal than relaxing music and no scent (Kellaris & Kent, 1991). It is suggested that arousal is an underlying mechanism of the effect of music and ambient scent on time spend:
H3. Stimulating (vs. relaxing) music combined with the presence (vs. absence) of
ambient scent will increase consumers’ arousal, which will consequently lead consumers to spend more time in the supermarket.
Finally, the amount of arousal is also expected to positively influence the overall satisfaction with the shopping experience. Again, it is important that consumers perceive the arousal as pleasant in order to produce positive responses (Spangenberg et al., 2006). As explained above, a congruence between music and ambient scent is therefore necessary. Earlier research indeed found that arousal only has a positive effect on overall satisfaction if the music and scent were congruent in terms of their arousing qualities (Mattila & Wirtz, 2001). To illustrate, a highly arousing scent (grapefruit) combined with fast tempo music led to higher satisfaction levels than a combination of a low arousing scent (lavender) and fast tempo music, which was attributed to the congruency between the scent and music (Mattila & Wirtz, 2001). Other studies on overall satisfaction confirmed this, showing that pleasurable arousal (i.e., congruency between music and scent) enhances consumers’ evaluations of the store (Vida et al., 2007; Bigné, Andreu, & Gnoth, 2005).
Congruency between music and scent is thus important to positively effect consumers’ overall satisfaction. As explained above, the present study kept the music and ambient scent congruent in both conditions. Both conditions are therefore expected to lead to a higher overall satisfaction, but this effect is expected to be stronger for the stimulating condition as consumers will experience more arousal than in the relaxing environment. To conclude, the
13 amount of arousal is suggested to be an underlying mechanism of the effect of music and ambient scent on overall satisfaction:
H4. Stimulating (vs. relaxing) music combined with the presence (vs. absence) of
ambient scent will increase consumers’ arousal, which will consequently lead to a higher overall satisfaction with the shopping experience.
Fig. 1 Conceptual model of the study, illustrating the proposed effects.
The Importance of Implicit Measures and Behavioral Measures
This study uses an experiment with two different measures: a combination of explicit measures (i.e., reports) and implicit measures or behavioral measures. Merely using self-reports is considered to be insufficient, as they are susceptible to a lot of biases. An example of this is social desirability, as participants want to present themselves in a positive way (Hefner et al., 2011). Also, participants might have insufficient insight into their own behavior or cognitive processes to answer a self-report correctly or a memory bias might distort their answers (Wilson & Dunn, 2004). To avoid these biases and strengthen the methodology of this study, effort was put into developing implicit measures. These implicit measures were
14 used for the variables self-control and arousal, respectively by measuring participants’
persistence on a self-developed unsolvable numerical puzzle (Appendix 4b) and by measuring arousal with the SAM-test (Lang, 1985, Appendix 4a). Behavioral measures were used for the variables product choice and time spend, respectively by using a specifically developed, realistic supermarket in which participants could choose products and add to their shopping cart, and by measuring the shopping time in seconds. To conclude, the combination of implicit and explicit measures make this study circumventing above problems and gives it a strong methodology.
Method
In this thesis, three studies were performed. First, a pre-test was conducted to choose a suitable ambient scent and music (i.e., stimulating and relaxing). Second, a pilot-study was conducted to test the self-developed online supermarket on usefulness and health perceptions of the products. Then, the main study was performed. All studies are described below. Pre-test
In order to decide which music and scent were suitable for the purpose of this study, a pre-test was performed. The specific aim was to find songs that were either stimulating or relaxing, and to find a scent that was perceived as highly stimulating. Initially, 16 participants joined the study. One response was unreliable as the answers were constantly the same and removed from the data, leading to a final sample of 15 participants (N = 15, Mage= 31.53,
SDage= 9.30, 66.7% female). They were approached face to face and asked to participate. Six
songs were tested, of which three were expected to be highly arousing, and three to be less arousing.1 Furthermore, three scents were tested on their perceived arousal.
A small experiment was conducted and combined with a questionnaire in Qualtrics.
1 Based on the tempo and rhythm. The volume of music can also influence shopping time, as studies showed that
loud music leads to less time spent (Smith & Curnow, 1966), but a higher perceived shopping time (Kellaris, Mantel, & Altsech, 1996). To make sure the effect of volume does not influence the results or the mediators (e.g., arousal and self-control), the volume of the music (the stimuli) is kept constant among participants.
15 After giving informed consent and providing their demographic details, participants were asked to listen to a music fragment and answer four questions to measure arousal: ‘This music activates me’, ‘This music makes me calm’, ‘This music stimulates me’, and ‘This music gives me energy’. These questions were asked for all six music fragments and answered on a seven-point Likert scale (1 = ‘Strongly disagree’, 7 = ‘Strongly agree’). Of each song, the first verse and chorus were played. The scale was reliable, Cronbach’s = .73.
After this, participants were asked to smell three scents and answer the same questions as mentioned above. The Cronbach’s alfa was = .36, indicating an unreliable scale. By removing the item ‘This scent makes me calm’, the reliability was heightened to = .83. Afterwards participants were thanked for their participation.
With a paired samples t-test, the responses for music were analyzed (Table 2a). The results showed that Bob Sinclar’s ‘love generation’ was perceived as most arousing (M = 5.65, SD = 0.43), while Jack Johnson’s ‘better together’ was perceived as least arousing (M = 3.07, SD = 0.35). The perceived arousal differed significant among these songs: t(14) = 19.12, p < .001, Cohen’s d = 6.58. The following most and least arousing songs were respectively
‘rather be’ of Clean Bandit ft. Jess Glynne’s (M = 4.78, SD = 0.90) and Maroon 5’s ‘Sunday morning’ (M = 3.65, SD = 1.06). These songs also differed significant on perceived arousal: t(14) = 4.33, p = .001, Cohen’s d = 1.15. From these results it was decided to create a playlist
with ‘love generation’, ‘rather be’, and several comparable songs for the stimulating condition, and ‘better together’, ‘Sunday morning’, and several comparable songs for the relaxing condition. The duration of the final playlist was respectively 33 and 32 minutes, allowing participants to complete the survey without hearing a repeated song (see appendix 2a and 2b for the final playlists).
Similarly, the responses for scent were analyzed (Table 2b). Three scents were tested: vanilla-aroma, citrus detergent, and flowery toilet spray. The results showed that citrus
16 detergent was perceived as most arousing (M = 4.76, SD = 1.16), and significantly more arousing than the toilet spray (M = 3.60, SD = 1.42; t(14) = 3.39, p = .004, Cohens d = 0.89), though not significantly more arousing than vanillaaroma (M = 4.42, SD = 0.60, t(14) = -1,00, p = .334, Cohens d = 0.75). Nevertheless, citrus detergent was selected for the experiment as a highly arousing scent.
Pilot Study
A fictive online supermarket looking like the Dutch Jumbo supermarket was specifically developed to use in the main study (Appendix 3). On the website
(www.jouwjumbo.nl), visitors could click on five categories in the header (e.g., ‘bread and cereals’) and choose between forty different products. The product of their liking could then be placed in the shopping cart. To test the usefulness of the online supermarket, a pilot study was performed (N = 8, Mage = 27, SDage = 10.18, 87.5% female). Participants were asked to
look around in the supermarket and choose the products they liked. The aim was to test whether the online supermarket was easy to navigate.
Next, as the main study will examine the amount of unhealthy product chosen, the products had to be categorized beforehand as healthy or unhealthy. Most products available in the supermarket were clearly healthy (e.g., grapes) or unhealthy (e.g., donuts), but six
products were a bit more ambiguous.2 To be able to categorize these products as healthy or unhealthy, participants were asked to classify them. Three products (Petit pains, Coco Pops, and Hero B’tween) were unanimously classified as unhealthy, and one product as healthy (Brinta) (Table 3). For the remaining two non-unanimously categorized products, a Chi-square analysis was performed. For ‘Bolletje Krokante Ontbijtgranen’, six out of eight participants classified it as healthy. The results of the Chi-square showed no significant association between the participants and their perception: χ2 (7) = 8.997, p = .253. However,
17 since the majority classified the product as healthy, it was classified as healthy. For ‘Jumbo Schoteleierkoeken’, six out of eight participants classified it as unhealthy. The results of the Chi-square showed no significant association between the participants and their perception: χ2 (7) = 8.997, p = .253. However, since the majority classified the product as unhealthy, it was classified as unhealthy. Lastly, participants were asked to give any additional comments they might have on the website. From the pilot-study, it became clear the online supermarket was easy to navigate. All participants thought the online supermarket looked realistic.
Main Study
The main study tested for two underlying mechanisms of the combined effect of music and ambient scent on product choice, time spend, and overall satisfaction: self-control and arousal.
Participants
Adolescents and adults, both men and women, were invited to participate in the study. Participants should be at least 18 years old, no upper age limit was set. They were approached via face-to-face contact, lab.uva.nl, social media, and word-of-mouth promotion. In total, 87 participants joined the study (N = 87), of which 20 men and 67 women (77% female). The participants were between 18 and 55 years old (M = 23.68, SD = 7.03). Among the
participants, the lowest form of completed education was MAVO (1.1%) and the highest form of completed education was a Masters’ degree (33.3%).
Design
The main study used a between-subject design. It was mainly conducted in the UvA-laboratory, and consisted of an experiment and a questionnaire in Qualtrics. The experimental part concerned shopping in the supermarket. Participants were randomly assigned to either the condition with ambient scent and stimulating music (stimulating condition, N = 42) or the condition without ambient scent and relaxing music (relaxing condition, N = 45). The
18 dependent variables were participants’ product choices, their time spent in the supermarket, and their overall satisfaction with the shopping experience. The mediators were self-control and amount of arousal.
Procedure
Participants were asked to come to the laboratory at the University of Amsterdam.3 After giving informed consent, participants were seated in front of the computer and were explained the procedure. They were then told the cover story that the study examined the effects of different website lay-outs. Depending on the condition they were in, participants were either exposed to highly stimulating music and a stimulating citrus scent (stimulating condition) or relaxing music without ambient scent (relaxing condition). When starting the survey, participants were first asked to shop for groceries. An online supermarket resembling the Jumbo supermarket website was especially developed for this study, so that consumers’ product choices could be monitored. Participants were able to choose between five food categories and were asked to put six products of their liking in the shopping card (see pilot study and appendix 3). When finished, participants’ overall satisfaction with the shopping experience was measured with six statements. After this, participants’ arousal and self-control was measured, with both an implicit and explicit measure. When completed, participants were asked three control questions. The survey was concluded with several demographic questions, concerning sex, age, and highest education level. By clicking on the arrow, participants were thanked for their participation, informed about the real aim of the experiment, and referred to an e-mail address for any questions they might have left. They could then call the instructor
3 As only a limited amount of participants could come to the laboratory (N = 78), it was decided to also conduct
this study outside of the laboratory (N = 9, 10.34% of total sample, stimulating condition: N = 5, relaxing condition: N = 4). When conducted outside the laboratory, it was made sure the conditions were exactly similar and no distractions were present. Participants were between 22 and 54 years old (M = 25.78, SD = 5.34), of which 3 men and 6 women (66.67% women). The highest form of completed education was a Masters’ degree (55.56%).
19 and leave the room.4 The instructor then filled in which products were chosen and submitted the questionnaire. No comments were made at the participants about the presence of music and (if applicable) ambient scent.
Product choice. In the supermarket, forty products could be chosen, sorted into five categories (bread and cereals, fruit and vegetables, chips and candy, cookies and pastry, coffee, tea, soda; appendix 3). Products were categorized beforehand as either healthy (e.g., grapes) or unhealthy (e.g., donuts; see also the pilot test). Participants were told they could choose six products of their liking while shopping. The amount of unhealthy products chosen by the participants was summed, so the dependent variable ranged from 0-6.
Time spend. The amount of time participants spent in the supermarket was measured with the help of Qualtrics. A timer started when participants started shopping, and ended when they confirmed to be done with the shopping. This variable was measured in seconds (M = 144.11, SD = 60.66).
Overall satisfaction. Participants’ overall satisfaction with shopping in the online supermarket was measured by a combination of two measures. Participants were asked six semantic differential items on a seven-point scale. The first three were dissatisfied vs. satisfied, displeased vs. pleased, and unfavorable vs. favorable (Jones & Shu, 2000). Then,
participants were asked to answer ‘I truly enjoyed coming to the supermarket, ‘I am satisfied with the supermarket’, and ‘I would return to this supermarket’ (adapted from Westbrook & Oliver, 1981). A principal component factor analysis with direct oblimin rotation showed the scale indeed consists of one factor (i.e., overall satisfaction), explaining 74.07% of the variance (Table 4). The reliability of the measurement was good (Cronbach’s = .93), indicating a reliable scale.
Arousal. Participants’ arousal was measured with implicit and explicit measures.
4 This study was conducted in collaboration with another study. Participants took part in both studies and were
randomly assigned to start with one of the studies. They performed a distracting task in-between to minimize the possible mutual influence.
20 First, the Self-Assessment Manikin (SAM) was used (Lang, 1985). Participants were asked to disclose their state of arousal by choosing the puppet that best portrayed their state of being (1 = ’least aroused’, 5 = ’most aroused’). Then, participants subjectively experienced arousal was measured explicitly with the Visual Analogue Scale (VAS; Davey, Barrat, Butow, & Deeks, 2007). They were asked to answer three questions on a scale from 1 (‘not at all’) to 7 (‘a lot’): ‘How nervous do you feel at the moment?’, ‘How fast is your heart beating?’, and ‘How sweaty are your palms?’ (adapted from Miers et al., 2011). A principal component factor analysis with direct oblimin rotation showed the scale indeed consists of one factor (i.e., arousal), explaining 71.96% of the variance. The reliability of the measurement was moderate (Cronbach’s = .80), indicating a reliable scale. Between the implicit and explicit measure of arousal, a high significant positive correlation was found: (r = .68, p < .001, N = 87). This means the measures are coherent.
Self-control. Participants’ amount of self-control was also measured with both explicit and implicit measures. To measure self-control implicitly, participants’ persistence on a numeric puzzle was tested (measured in seconds, M = 248.26, SD = 102.73).5 The goal was to find specific combinations of numbers within a large number grid, of which some were deliberately not included (Appendix 4b). Participants were instructed to take as much time as they wanted to complete it, and that they were allowed to stop at any time. This task was based on previous research (Vohs, Baumeister & Ciarocco, 2005; Schmeichel & Vohs, 2009).
The explicit measure consisted of sixteen items from the Grasmick et al. (1993) scale, one of the most used scales in empirical research into self-control (DeCamp, 2015; DeLisi, Hochstetler & Murphy, 2003). Originally, the scale contained 23 items, but the dimensions ‘risk taking’ and ‘psychical activities’ were omitted as they were deemed unrelevant to this study. Participants were shown statements like ‘When things get complicated, I tend to quit or
5
Although not communicated to the participants, if participants would continue to work on the puzzle for more than twenty minutes, the researcher would interrupt them and make them stop. This did not occur.
21 withdraw’ and ‘I lose my temper pretty easily’ (Appendix 4c). For each statement, they were asked to indicate how much they agreed with it, on a scale from 1 (‘Totally disagree’) to 5 (‘Totally agree’). A principal component factor analysis with varimax rotation showed that the scale of self-control consists of six factors, with one factor explaining 20.26% of the variance (Table 5).6 The reliability of the measurement was moderate (Cronbach’s = .73), indicating a reliable scale. No linear relationship exists between the explicit and implicit measure: a non-significant, negative correlation was found between them (r = -.14, p = .211, N = 87). This can be due to the fact that the implicit measure only used one item or due to the fact that both were measured on different type of scales (Aggerwal & Ranganathan, 2016).
Control variables. Participants were asked three control questions to make sure the data was not influenced by circumstances. These were the following: ‘Are you hungry at the moment?’ (1 = ‘Yes’, 2 = ‘No’), ‘Do you have food allergies?’ (1 = ‘No’, 2 = ‘Yes, namely..’),
and ‘Are you on a diet?’ (1 = ‘Yes’, 2 = ‘No’). These questions were important as they can potentially influence participants’ product choices.
Results Preliminary Analyses
Control variables. Two control variables were considered as possible explanations of differences between participants. These were ‘Are you hungry at the moment?’ and ‘Are you on a diet?’.7 The randomization of both variables was tested with a Chi-square analysis. The
results showed that the number of people being hungry were equally divided over the two conditions: χ2 (1) = 0.22, p = .636, which was also the case for participants on a diet: χ2 (1) =
6
Instead of six, four factors were expected to be found. Two factors that were separately extracted can logically be divided among ne of the other factors (i.e., temper). This contradicting finding can be attributed to the relatively low N.
7
The control variable for food allergies was not taken into account, as only a few allergies were reported which were not related to the available groceries in the supermarket (i.e., shallot and lactose).
22 0.41, p = .522. It is concluded that the results cannot be accounted for by these two variables as they were random divided among participants.
Order-bias. To check whether the order in which the participants performed the two tests (i.e., this or the other study in the laboratory first), an independent samples t-test was performed with order as independent variable and self-control, arousal, product choice, time spend, and overall satisfaction as dependent variables (i.e., the proposed mediators and dependent variables in the conceptual model). Only one significant association was found, between order and the explicit measure of arousal. This can be due to chance or due to the difficulty of accurately answering the explicit measure of arousal, as for example ‘How fast is you heart beating’ is difficult to self-monitor. This could have biased the results. Therefore, no order-bias is assumed or controlled for (Table 6).
Randomization Checks
Gender. To check whether the randomization of the variable gender across conditions was successful, a Chi-square analysis was performed with the variables condition and gender. The results showed that men and women were equally divided among the conditions: χ2 (1) = 0.11, p = .738. Therefore it can be concluded that the randomization for gender was
successful.
Age. To check whether the randomization of the conditions for the variable age was successful, an independent samples t-test was performed, with condition as independent variable and age as dependent variable. The results show different ages are equally distributed among the conditions: t(85) = -0.197, p = .845. The randomization for age was thus
successful.
Hypotheses Testing
23 PROCESS macro tool for SPSS developed by Hayes (2013). All directs effects are presented below (Table 1).
Table 1
The Mediating Role of Self-Control
Effect on time spend in the supermarket. In the first hypothesis, it was proposed that self-control depletion mediates the positive effect of the presence of stimulating music and ambient scent on time spend in the supermarket. For the implicit measure of self-control (i.e., puzzle task), the results were not in line with H1: the effect of stimulating (versus
24 relaxing) music combined with the presence (versus absence) of ambient scent on time spend in the supermarket mediated by self-control depletion was not confirmed (indirect effect = 0.49, SE = 2.95, 95% CI [-3.57; 8.77]. The results showed the stimulating condition versus the relaxing condition did not lead to more depletion of self-control, b = 17.01, SE = 22.09, t = 0.77, p = .444, which consequently did not lead to a significant longer time spent in the supermarket, b = 0.03, SE = 0.06, t = 0.45, p = .666, thereby not providing support for H1
For the explicit measure, the results were also not in line with H1: indirect effect = -0.15, SE = 1.41, 95% CI [-4.33; 1.97]. The results showed that the stimulating condition versus the relaxing condition did not lead to more depletion of self-control, b = 0.02, SE = 0.14, t = 0.16, p = .873, which consequently did not lead to a significant longer time spent in the supermarket, b = -6.91, SE = 10.38, t = -0.67, p = .507.
Effect on product choice. In the second hypothesis, it was proposed that a depletion in self-control mediates the positive effect of the presence of stimulating music and ambient scent on the amount of unhealthy products chosen in the supermarket. For the implicit
measure, the results were not in line with H2: the effect of stimulating (versus relaxing) music combined with the presence (versus absence) of ambient scent on the amount of unhealthy products chosen mediated by self-control was not confirmed (indirect effect = 0.01, SE = .04, 95% CI [-0.03; 0.15]. The results showed that the stimulating condition versus the relaxing condition did not lead to more depletion of self-control, b = 17.01, SE = 22.09, t = 0.77, p = .443, which consequently did not lead to a significant higher amount of unhealthy products, b = 0.00, SE = 0.01, t = 0.51, p = .615.
For the explicit measure, results were also not in line with H2: indirect effect = -0.004, SE = 0.04, 95% CI [-0.13; 0.05]. The results showed that the stimulating condition versus the relaxing condition did not lead to more depletion of self-control, b = 0.02, SE = .14, t = 0.16, p = .873, which consequently did not lead to a significant higher amount of unhealthy
25 products, b = -0.19, SE = 0.21, t = -0.88, p = .382. Supermarkets using stimulating music and ambient scent do not seem to increase consumers’ time spend in the supermarket or their product choices. This effect does not seem to be mediated by the amount of consumers’ self-control. H1 and H2 are thus not confirmed.
The Mediating Role of Arousal
Effect on time spend in the supermarket. In the third hypothesis, it was proposed that an increase in arousal mediates the positive effect of the presence of stimulating music and ambient scent on time spend in the supermarket. For the implicit measure of arousal, the results were not in line with H3: the effect of stimulating (versus relaxing) music combined with the presence (versus absence) of ambient scent on time spend mediated by arousal was not confirmed (indirect effect = 1.16, SE = 2.13, 95% CI [-1.37; 7.96]. The results showed that the stimulating condition versus the relaxing condition did not lead to a higher arousal, b = -0.15, SE = 0.21, t = -0.71, p = .477, which consequently did not lead to significant more time spent in the supermarket, b = -7.69, SE = 6.70, t = -1.15, p = .255.
For the explicit measure, the results were also not in line with H3: indirect effect = 0.97, SE = 2.20, 95% CI [-1.15; 8.60]. The results showed that the stimulating condition versus the relaxing condition did not lead to a higher arousal, b = -0.21, SE = 0.29, t = -0.72, p = .472, which consequently did not lead to significant more time spent in the supermarket, b = -4.57 SE = 4.84, t = -0.94, p = .348.
Effect on overall satisfaction. In the fourth hypothesis, it was proposed that an increase in arousal mediates the positive effect of the presence of stimulating music and ambient scent on overall satisfaction with the shopping experience. For the implicit measure of arousal, the results were not in line with H4: the effect of stimulating (versus relaxing) music combined with the presence (versus absence) of ambient scent on overall satisfaction mediated by arousal was not confirmed (indirect effect = .003, SE = 0.03, 95% CI [-0.05;
26 0.10]. The results showed that the stimulating condition versus the relaxing condition did not lead to a higher arousal, b = -0.15, SE = 0.21, t = -0.71, p = .477, which consequently did not lead to a significant difference in overall satisfaction, b = -0.02, SE = 0.14, t = -0.17, p = .870.
For the explicit measure, the results were also not in line with H4: indirect effect = -0.02, SE = 0.04, 95% CI [-0.17; 0.03]. The results showed that the stimulating condition versus the relaxing condition did not lead to a higher arousal, b = -0.21, SE = 0.29, t = -0.72, p = .472, which consequently did not lead to a significant difference in overall satisfaction, b = 0.08, SE = 0.10, t = 0.81, p = .422. Supermarkets using stimulating music and ambient scent do not seem to increase consumers’ time spend in the supermarket or their overall satisfaction. This effect does not seem to be mediated by the amount of arousal consumers experience. Therefore H3 and H4 are not confirmed.
Conclusion and Discussion
The aim of this study was to examine the effects of ambient scent and music on consumer behavior in the supermarket. It was expected that music and ambient scent would influence consumers’ time spend, product choice, and overall satisfaction. Two mechanisms were proposed to underlie these effects: self-control and arousal. The results showed that the presence of stimulating music and ambient scent did not lead to less self-control, which consequently did not lead to more time spent or the purchase of more unhealthy products. Furthermore, the presence of stimulating music and ambient scent did not increase
consumers’ arousal, which subsequently did not lead to more time spent or a higher overall satisfaction.
Findings and Implications
Self-control. The first main finding was that stimulating music and ambient scent did not deplete consumers’ self-control. It is thus not the case that the presence of stimulating music and scent in the supermarket lowers people’s control. Also, consumers’
self-27 control did consequently not lead to more time spent in the supermarket. Thus, it is not the case that the amount of self-control influences consumers’ time spend. These findings are in contrast with hypothesis 1 and not in line with studies that showed that an initial act of self-control (i.e., due to stimulating music and scent) leads to less subsequent self-self-control
(Baumeister et al., 1998). It was also not in line with studies showing that a lower self-control diminishes the ability to think orderly and rationally (Tice et al., 2001), which was
consequently expected to heighten consumers’ time spend in the supermarket.
The second main finding was that the presence of music and ambient scent did not lead to the purchase of more unhealthy products through a depletion in self-control. Thus, it is not the case that the amount of self-control influences people’s product choices. This was not in line with the expectations and hypothesis 2, as earlier research showed that people with a depleted self-control have less self-regulation to choose healthy products (Vohs &
Heatherton, 2000) and engage in more impulsive buying behavior (Mattila & Wirtz, 2008), which often leads to the purchase of unhealthy products (Verplanken et al., 2005).
An explanation for the absent effect of music and ambient scent on consumers’ self-control might be that participants in the stimulating condition were not long enough exposed to the music and ambient scent before starting shopping. This could have reduced their perception of the environment as over-stimulating and their consequent level of excitement (Spangenberg et al., 1996). This could explain why the predicted momentary loss in self-control did not occur (Mattila & Wirtz, 2008), and the non-significant difference between the stimulating and relaxing environment. Future research should examine how long consumers need to be exposed to an over-stimulating environment before it leads to a loss in self-control.
Furthermore, the absence of ego-depletion could subsequently explain why
participants’ time spend did not increase, as a diminished ability to think orderly was only expected through ego-depletion (Tice et al., 2001). No relationship was found between a loss
28 of self-control (i.e., ego-depletion) and time spend. The influence of ego-depletion on time spend could be further examined by future research.
The absence of an effect of self-control on the amount of unhealthy product choices might be explained with social desirability, which is “the defensive tendency of individuals to portray themselves in keeping with perceived cultural norms” (Adams et al., 2005, p. 389). Social desirability has been found to influence participants’ reports of diets, for example by participants underreporting their fat intake (Hébert et al., 2001). By using implicit measures, this study tried to minimize the influence of biases like social desirability. However, as participants knew the researcher could see their product choices, they might have chosen less unhealthy products as being healthy is more socially desirable (Hébert et al., 2001).
Lastly, consumers’ self-image might have influenced their product choices. Research shows consumers have a greater preferences for products that are congruent with their self-image or ideal self-self-image (Malhotra, 1988; Sirgy, Grewal, & Mangleburg, 2000). This might explain why the expected increase in unhealthy products did not occur, as participants’ ideal self-image would be conflicting with this. The influence of consumers’ self-image on their product choices could be further explored by future research.
Arousal. The third main finding was that stimulating music and ambient scent did not increase consumers’ arousal. It is thus not the case that the presence of stimulating music and scent in the supermarket heightens people’s arousal. Also, consumers’ arousal did not
consequently increase their shopping time. Thus, it is not the case that the amount of arousal influences time spend. These findings are contradicting to hypothesis 3: the complex
environment with music and scent was expected to heighten participants’ arousal as a lot of information needed to be processed (Mehrabian & Russell, 1974). This arousal was expected to be pleasant due to the congruence between music and scent, and therefore expected to increase consumers’ time spend (Spangenberg et al., 2006).
29 The absence of this effect might be explained with participants’ perception of the arousal. Research shows that congruence between music and scent leads to pleasant arousal (Mattila & Wirtz, 2001) and consequently to approach behavior, while unpleasant arousal leads to avoidance behavior (Spangenberg et al., 2006). Although the music and scent were kept congruent in this study, participants might have unexpectedly still experienced the arousal as unpleasant. This would explain why participants did not spend more time: if the arousal was perceived as unpleasant, it would lead to avoidance behavior (i.e., spending less time). Future research should therefore take participants’ perception of the valence of arousal into account as a control variable.
The last main finding was that stimulating music and ambient scent did not increase consumers arousal, and subsequently did not lead to a higher overall satisfaction. This is not in line with hypothesis 4 and contradicting with earlier research suggesting the presence of music and ambient scent leads to a higher satisfaction if congruent with each other (Mattila & Wirtz, 2001). As music and scent were kept congruent in both conditions of this study, the results are unexpected. It might be that the low number of participants (N = 87) influenced these results, as it decreases the statistical power, making a difference between groups harder to detect (Nayak, 2010). Of course, this is influencing all variables in the study. A more specific explanation for the absent effect on satisfaction could be that the online supermarket only offered forty products, making it sometimes impossible to buy one’s favorite products or brand. As brand loyalty is an important predictor of overall satisfaction (Bloemer & Kasper, 1995), the absence of certain products might have influenced this. Also, by using the look of Jumbo supermarkets for this study, participants might have expected the ability to buy their usual groceries. For future research, it might help to create a fictive supermarket and communicate to participants that it only offers a certain amount of products as their brand strategy (see Mazar & Zhong, 2010 for an example), decreasing these expectations, and
30 thereby the influence on satisfaction.
Although the present study could not demonstrate the hypothesized effects, it does add to existing atmospheric research. Whereas most research only focusses on the effects of music (Morrison et al., 2011), the present study focused on the combined effect of music and
ambient scent in a consumer environment. This is a rather rare approach in atmospheric research (Morrison et al., 2011). Furthermore, this study used both implicit and explicit measures, while existing research often relies on either one. Effort was put into the design of the implicit measures, by developing a numerical puzzle (to measure self-control) and online supermarket (to measure product choice and time spend). This strengthened the methodology of this study, as the measured concept was less obvious to participants which reduced the social desirability in their answers (MacDorman, Vasudevan, & Ho, 2009), and avoided memory biases or biased self-reports. Arousal and self-control were measured with both implicit and explicit measures. The results for both measures were compared and no differences could be found between them.
Lastly, only a few existing studies address the reasons why music or ambient scent can effect consumers, keeping the processes at play largely unknown (Roschk et al., 2017). This study responded to this gap by proposing self-control and arousal as two possible underlying mechanisms. The results could not confirm these concepts as underlying mechanisms for the effects of atmospheric cues on time spend, product choice, and overall satisfaction. Additional research could further examine these mechanisms, and examine whether they are indeed unrelated to the effects of atmospheric cues in the consumer environment.
Limitations and Future Research
The current research has some limitations. Firstly, only the physical dimension of music (i.e., tempo and rhythm) was taken into account. Future research could also examine the emotional and valence dimension of music (Bruner, in Mattila & Wirtz, 2001). Especially
31 valence (i.e., liking the music) can be interesting, as this is a subjective rather than objective dimension like the physical and emotional dimension. Future research could examine whether valence is more or less effective than the physical dimension in influencing consumer
behavior, and whether valence for example influences overall satisfaction. A positive effect on overall satisfaction can be expected, as music valence was already shown to influence consumers’ emotions in a positive way (Vida et al., 2007).
Secondly, the factor analysis for self-control resulted into six factors, while a scale with four factors was used (Grasmick et al., 1993). Two items did surprisingly not load on the existing factor. The discrepancy could be due to the relatively low N, but future research should look into the Grasmick et al. scale (1993) and verify whether the items indeed measure the construct of self-control in an effective way.
For future research, it could be interesting to conduct an actual field study to see if the proposed relationships can be demonstrated in a real-life setting. In addition, future research should look into other possible underlying mechanisms for the effects of music and ambient scent. Creating a better understanding of the effects of atmospheric cues will not only help retailers create a pleasurable consumption environment, but will also help consumers choose their products mindfully by explaining how atmospheric cues – often unconsciously,
influence their behavior, with the ultimate goal of helping them purchase less unhealthy products. As 50.2% of Dutch residents struggle with overweight, this could be an important implication (Leefstijlmonitor, 2018). If future research can demonstrate a way in which consumers can be helped to purchase less unhealthy products through the use of atmospheric marketing, society can become healthier in an innovatively and unique way.
32 Adams, C., Doucé, L., Janssens, W., Vanrie, J., & Petermans, A. (2014). Tasting the smell:
Effects of ambient scent on scent experts’ evaluations of (in) congruent food products. Food quality and preference, 38, 92-97.
Adams, S. A., Matthews, C. E., Ebbeling, C. B., Moore, C. G., Cunningham, J. E., Fulton, J., & Hebert, J. R. (2005). The effect of social desirability and social approval on self-reports of physical activity. American journal of epidemiology, 161(4), 389-398. Aggarwal, R., & Ranganathan, P. (2016). Common pitfalls in statistical analysis: The use of
correlation techniques. Perspectives in clinical research, 7(4), 187.
Baumeister, R. F., Bratslavsky, E., Muraven, M., & Tice, D. M. (1998). Self-control depletion: Is the active self a limited resource? Journal of Personality and Social Psychology, 74, 1252–1265.
Bäckström, K., & Johansson, U. (2006). Creating and consuming experiences in retail store environments: Comparing retailer and consumer perspectives. Journal of Retailing and Consumer Services, 13, 417-430.
Berger, J., & Fitzsimons, G. (2008). Dogs on the street, pumas on your feet: How cues in the environment influence product evaluation and choice. Journal of Marketing
Research, 45(1), 1-14.
Bigné, J. E., Andreu, L., & Gnoth, J. (2005). The theme park experience: An analysis of pleasure, arousal and satisfaction. Tourism management, 26(6), 833-844.
Bloemer, J. M., & Kasper, H. D. (1995). The complex relationship between consumer satisfaction and brand loyalty. Journal of economic psychology, 16(2), 311-329. Bone, P. F., & Ellen, P. S. (1999). Scents in the marketplace: Explaining a fraction of
olfaction. Journal of Retailing, 75(2), 243-262.
Bosmans, A. (2006). Scents and sensibility: when do (in)congruent ambient scents influence product evaluations? Journal of Marketing, 70(3), 32-43.
33 Chang, K. (2000). The impact of perceived physical environments on customers' satisfaction
and return intentions. Journal of Professional Services Marketing, 21(2), 75-85. Cruise, C. J., Chung, F., Yogendran, S., & Little, D. A. (1997). Music increases satisfaction in
elderly outpatients undergoing cataract surgery. Canadian Journal of Anaesthesia, 44(1), 43-48.
Davey, H. M., Barratt, A. L., Butow, P. N., & Deeks, J. J. (2007). A one-item question with a Likert or Visual Analog Scale adequately measured current anxiety. Journal of
Clinical Epidemiology, 60(4), 356-360.
DeCamp, W. (2015). Why so many questions? Measurement issues and the attitudinal self-control scale. Deviant Behavior, 36(6), 463-473.
DeLisi, M., Hochstetler, A., & Murphy, D. S. (2003). Self-control behind bars: A validation study of the Grasmick et al. scale. Justice Quarterly, 20(2), 241-263.
Doucé, L., & Janssens, W. (2013). The presence of a pleasant ambient scent in a fashion store: The moderating role of shopping motivation and affect intensity. Environment and Behavior, 45(2), 215-238.
Gailliot, M. T., Baumeister, R. F., DeWall, C. N., Maner, J. K., Plant, E. A., Tice, D. M., Brewer, L.E., & Schmeichel, B. J. (2007). Self-control relies on glucose as a limited energy source: willpower is more than a metaphor. Journal of personality and social psychology, 92(2), 325.
Grasmick, H. G., Tittle, C. R., Bursik Jr, R. J., & Arneklev, B. J. (1993). Testing the core empirical implications of Gottfredson and Hirschi's general theory of crime. Journal of Research in Crime and Delinquency, 30(1), 5-29.
Guéguen, N., & Petr, C. (2006). Odors and consumer behavior in a restaurant. International Journal of Hospitality Management, 25(2), 335-339.
34 Sorensen, G. (2001). The effect of social desirability trait on self-reported dietary measures among multi-ethnic female health center employees. Annals of
epidemiology, 11(6), 417-427.
Hefner, D., Rothmund, T., Klimmt, C., & Gollwitzer, M. (2011). Implicit measures and media effects research: Challenges and opportunities. Communication Methods and
Measures, 5(3), 181-202.
Kellaris, J. J. & Kent, R. J. (1991). Exploring tempo and modality effects on consumer responses to music, Advances in Consumer Research, 18(1), 243-248
Kellaris, J. J., Mantel, S. P., & Altsech, M. B. (1996). Decibels, disposition, and duration: the impact of musical loudness and internal states on time perceptions. ACR North
American Advances.
Knasko, S. C. (1989). Ambient odor and shopping behavior. Chemical Senses, 14(5), 719. Lang, P. J. (1985). The cognitive psychophysiology of emotion: anxiety and the anxiety
disorders. Hillsdale, NJ: Lawrence Erlbaum.
Lee, D., Henderson, A., & Shum, D. (2004). The effect of music on preprocedure anxiety in Hong Kong Chinese day patients. Journal of Clinical Nursing, 13(3), 297-303. Leefstijlmonitor (2018). “Meer dan de helft van de Nederlanders is te zwaar”. Retrieved via:
https://www.volksgezondheidenzorg.info/onderwerp/overgewicht/cijfers-context/huidige-situatie#bron--node-tabel-bronnen-bij-de-cijfers-over-overgewicht Leenders, M. A., Smidts, A., & El Haji, A. (2019). Ambient scent as a mood inducer in
supermarkets: The role of scent intensity and time-pressure of shoppers. Journal of Retailing and Consumer Services, 48, 270-280.
MacDorman, K. F., Vasudevan, S. K., & Ho, C. C. (2009). Does Japan really have robot mania? Comparing attitudes by implicit and explicit measures. AI & society, 23(4), 485-510.
35 Madzharov, A. V., Block, L. G., & Morrin, M. (2015). The cool scent of power: Effects of
ambient scent on consumer preferences and choice behavior. Journal of Marketing, 79(1), 83-96.
Malhotra, N. K. (1988). Self concept and product choice: An integrated perspective. Journal of Economic Psychology, 9(1), 1-28.
Mattila, A. S., & Wirtz, J. (2001). Congruency of scent and music as a driver of in-store evaluations and behavior. Journal of retailing, 77(2), 273-289.
Mattila, A. S., & Wirtz, J. (2008). The role of store environmental stimulation and social factors on impulse purchasing. Journal of services marketing, 22(7), 562-567.
Mazar, N., & Zhong, C. B. (2010). Do green products make us better people?. Psychological Science, 21(4), 494-498
McQuitty, S., Finn, A., & Wiley, J. B. (2000). Systematically varying consumer satisfaction and its implications for product choice. Academy of Marketing Science Review, 10(1), 231-254.
Mehrabian, A., & Russell, J. A. (1974). An approach to environmental psychology. Cambridge: The MIT Press.
Michel, A., Baumann, C., & Gayer, L. (2017). Thank you for the music–or not? The effects of in-store music in service settings. Journal of Retailing and Consumer Services, 36, 21-32.
Miers, A. C., Blöte, A. W., Sumter, S. R., Kallen, V. L., & Westenberg, P. M. (2011). Subjective and objective arousal correspondence and the role of self-monitoring processes in high and low socially anxious youth. Journal of Experimental Psychopathology, 2(4), 531-550.
Milliman, R. E. (1982). Using background music to affect the behavior of supermarket shoppers. Journal of Marketing, 46(2), 86–91
36 Morrin, M., & Ratneshwar, S. (2000). The impact of ambient scent on evaluation, attention,
and memory for familiar and unfamiliar brands. Journal of Business Research, 49(2), 157-165.
Morrison, M., & Beverland, M. (2003). In search of the right in-store music. Business Horizons, 46(6), 77-82.
Morrison, M., Gan, S., Dubelaar, C., & Oppewal, H. (2011). In-store music and aroma influences on shopper behavior and satisfaction. Journal of Business Research, 64(6), 558-564.
Nayak, B. K. (2010). Understanding the relevance of sample size calculation. Indian journal of Ophthalmology, 58(6), 469.
North, A. C., Hargreaves, D. J., & McKendrick, J. (1997). In-store music affects product choice. Nature, 390(6656), 132.
North, A. C., Hargreaves, D. J., & McKendrick, J. (1999). The influence of in-store music on wine selections. Journal of Applied Psychology, 84(2), 271-276.
Pan, Y., & Zinkhan, G. M. (2006). Determinants of retail patronage: a meta-analytical perspective. Journal of Retailing, 82(3), 229-243.
Roberts, J. (2008). On the scent trail. Brand Strategy, 14-15. Unknown.
Roehm, H. A., & Roehm, M. L. (2005). Revisiting the effect of positive mood on variety seeking. Journal of Consumer Research, 32(2), 330-336.
Roschk, H., Loureiro, S. M. C., & Breitsohl, J. (2017). Calibrating 30 years of experimental research: a meta-analysis of the atmospheric effects of music, scent, and color. Journal of Retailing, 93(2), 228-240.
Schmeichel, B. J., & Vohs, K. (2009). Self-affirmation and self-control: affirming core values counteracts ego depletion. Journal of personality and social psychology, 96(4), 770-782.
37 Shin, D., & Elliott, K. M. (2001). Measuring customers' overall satisfaction: A
multi-attributes assessment. Services Marketing Quarterly, 22(1), 3-19.
Sirgy, M. J., Grewal, D., & Mangleburg, T. (2000). Retail environment, self-congruity, and retail patronage: an integrative model and a research agenda. Journal of Business research, 49(2), 127-138.
Spangenberg, E. R., Crowley, A. E., & Henderson, P. W. (1996). Improving the store environment: do olfactory cues affect evaluations and behaviors? Journal of Marketing, 60(2), 67-80.
Spangenberg, E. R., Grohmann, B., & Sprott, D. E. (2005). It's beginning to smell (and sound) a lot like Christmas: the interactive effects of ambient scent and music in a retail setting. Journal of Business Research, 58(11), 1583-1589.
Spangenberg, E. R., Sprott, D. E., Grohmann, B., & Tracy, D. L. (2006). Gender-congruent ambient scent influences on approach and avoidance behaviors in a retail
store. Journal of Business Research, 59(12), 1281-1287.
Stöhr, A. (1998). Air-Design als Erfolgsfaktor im Handel. Wiesbaden: Deutscher Universitäts Verlag.
Sullivan, M. (2002). The impact of pitch, volume and tempo on the atmospheric effects of music. International Journal of Retail & Distribution Management, 30(6), 323-330. Teller, C., & Dennis, C. (2012). The effect of ambient scent on consumers' perception,
emotions and behaviour: A critical review. Journal of Marketing Management, 28(1-2), 14-36.
Tice, D. M., Bratslavsky, E., & Baumeister, R. F. (2001). Emotional distress regulation takes precedence over impulse control: If you feel bad, do it!. Journal of personality and social psychology, 80(1), 53-67.
