Impact of music on consumer behaviour

A

a ( e o s f 3 C

K

h t s 2 i e K a s M A a a i

l (

h 0

Journal of Retailing 92 (1, 2016) 83–95

Music Congruity Effects on Product Memory, Perception, and Choice

Adrian C. North a,1, Lorraine P. Sheridan a,2, Charles S. Areni b,∗ a School of Psychology and Speech Pathology, Curtin University, Perth, Western Australia 6845, Australia

b Macquarie Graduate School of Management, 99 Talavera Road, North Ryde, New South Wales 2113, Australia

Available online 26 June 2015

bstract

Music congruity effects on consumer behavior are conceptualized in terms of cognitive priming of semantic networks in memory, and oper- tionalized as congruent with a product’s country of origin (Experiment 1), or congruent with the utilitarian (Experiment 2) or social identity Experiments 2 and 3) connotations of a product. Hearing a specific genre of music (e.g., classical) activates related concepts in memory (e.g., xpensive, sophisticated, formal, educated), which influences the memory for, perception of, and choice of products. Consistent with this account f music congruity effects, three laboratory experiments show that playing music of a specific genre during initial product exposure improved ubsequent recall of conceptually related (i.e., congruent) products compared to unrelated products (Experiment 1), affected product choice in

avor of congruent products (Experiment 1), and affected how much participants were willing to pay for congruent products (Experiments 2 and ). rown Copyright © 2015 Published by Elsevier Inc on behalf of Society affiliation: New York University. All rights reserved.

a m a a p p a O

o t c t f

eywords: Atmospherics; Music; Retailing; Congruity; Recall; Choice

Introduction

Laboratory and field experiments over the last four decades ave demonstrated various effects of what has been referred o as ‘background’, ‘atmospheric’, or ‘in-store’ music, on con- umer perception and behavior (Bruner 1990; Garlin and Owens 006; Turley and Milliman 2000). Various objective character- stics of music like volume (Kellaris and Rice 1993; Morrison t al. 2011), tempo (Caldwell and Hibbert 2002; Kellaris and ent 1994; Milliman, 1982, 1986; Oakes 2003), key (Kellaris

nd Kent 1992, 1994), texture (Kellaris and Kent 1994), and ubjective aspects of music like arousal (Dube, Chebat, and

orin 2007; Mattila and Wirtz 2001), familiarity (Bailey and reni 2006; Yalch and Spangenberg 2000), liking (Caldwell

nd Hibbert 2002; Dube, Chebat, and Morin 2007; Hui, Dube, nd Chebat 1997; Morin, Dube, and Chebat 2007), complex- ty (North and Hargreaves 1998), and perceived mood (Alpert

∗ Corresponding author. Tel.: +61 02 9850 9085. E-mail addresses: [email protected] (A.C. North),

[email protected] (L.P. Sheridan), [email protected] C.S. Areni). 1 Tel.: +61 08 9266 7867. 2 Tel.: +61 08 9266 9279.

c – t e c N a o a U

ttp://dx.doi.org/10.1016/j.jretai.2015.06.001 022-4359/Crown Copyright © 2015 Published by Elsevier Inc on behalf of Society

nd Alpert 1990; Alpert, Alpert, and Maltz 2005) have been anipulated and shown to influence a wide range of perceptions

nd behaviors, including actual purchases, purchase intentions, ctual shopping time, actual waiting time, actual dining time, erceived shopping time, perceived waiting time, specific store erceptions, overall store evaluations, customer–employee inter- ctions, and salesperson effectiveness (Garlin and Owens 2006; akes 2000; Turley and Milliman 2000). But perhaps the most influential aspect of music, in terms

f its effect on retail sales, is what has been variously referred o as the ‘genre’, ‘format’, or ‘style’ of the music (i.e., classi- al vs. hip hop vs. country, etc.). A common finding has been hat congruity between the connotations of the musical genre, ormat or style and certain products can influence sales and per- eption of the latter. One particular musical genre – classical

has received a considerable amount of attention in the litera- ure. For instance, Baker, Levy, and Grewal (1992) and Grewal t al. (2003) found that playing classical music resulted in per- eptions of higher quality service and merchandize. Likewise, orth and Hargreaves (1998) found that classical music cre-

ted a greater willingness to pay for each of fourteen items

n sale at a cafe, compared to pop music, easy listening, and

no music control condition; the mean total amounts were K£17.23, UK£16.61, UK£14.51, and UK£14.30 respectively.

affiliation: New York University. All rights reserved.

8 of Ret

N c i n o c t c p W i p a a e d p

m c a o ‘ 1 p t c a a 1

e b “ t r a “ a K m ( ( l r (

t e a a s t

c t w c e fi t ( i a t t A

a n w s r l m t f s s i O c v m c a l w c e d n o c

R i i f r h J e D h o i

4 A.C. North et al. / Journal

orth, Shilcock, and Hargreaves (2003) similarly found that lassical music in a restaurant was associated with higher spend- ng per head (UK£32.52) than either pop music (UK£29.46) or o music (UK£29.73). Areni and Kim (1993) reported effects f even greater magnitude; classical music led to wine cellar ustomers spending a mean of US$7.43, whereas pop music led o a mean of only US$2.18. Collectively, these studies show that lassical music induces consumers to spend more money, and to urchase more expensive items, than other genres of music (c.f., ilson 2003). However, because these experiments took place

n actual retail settings, they cannot fully specify the cognitive rocesses by which classical music induces greater spending rel- tive to other musical genres. This research presents a theoretical ccount of the cognitive processes involved in musical congruity ffects, and reports the results of three laboratory experiments emonstrating the effects of musical congruity on recall, product erceptions, and choice.

Music Congruity, Cognitive Priming and Spreading Activation

Although it is typically thought of as an objective aspect of usic, like tempo or texture, the genre or format of a musi-

al composition is actually a subjective judgment (Aucouturier nd Pachet 2003). Consumers essentially amalgamate a range f objective musical features into a categorization of a song as light classical’ versus ‘instrumental jazz’ (Rose and Wagner 995). Moreover, two consumers listening to the same com- osition might classify it into different genres depending on heir idiosyncratic conceptions of what makes a piece ‘classi- al’ versus ‘jazz’. However, when a composition is perceived s belonging to a specific genre, the cultural meanings associ- ted with that category become associated with the piece (Fiske 993; White 1997).

Scott (1990) was perhaps the first to identify this potential ffect of music in a marketing context, describing music as eing “meaningful, language-like” and called for research that draws on notions of culture, rhetoric, and symbolic action” o understand persuasive effects of music (p. 233). Subsequent esearchers elaborated on this aspect of music. MacInnis nd Park (1991) referred to the “fit” of music as reflecting consumers’ subjective perceptions of the music’s relevance or ppropriateness to the central ad message” (p. 162). Likewise, ellaris, Cox, and Cox (1993) proposed that music “enhances essage reception when the music evokes message-congruent

vs. incongruent) thoughts” (p. 114). Zhu and Meyers-Levy 2005) identified multiple aspects of music, one of which was abeled “referential meaning, which is context dependent and eflects networks of semantic-laden, external world concepts” p. 333).

These descriptions of music can be embedded into a broader heoretical framework which conceptualizes music congruity ffects in terms of cognitive priming of concepts and spreading

ctivation along related semantic networks in memory. Collins nd Loftus (1975) argue that concepts can be represented as emantic nodes within a network in memory, and the proper- ies of those concepts can be represented as relational links

s i

t

ailing 92 (1, 2016) 83–95

onnecting one node to another. Each of the nodes that are linked o a given node will have links to other concept nodes, and so on, hich creates a vast semantic, interconnected network of con-

ept nodes. When one node in memory is initially activated via xternal stimulation, the spread of activation expands outwards, rst to all the nodes linked to the initially activated node, then

o all the nodes linked to each of those, and so forth. Anderson 1983) believed that the key process of memory performance s the retrieval function, which is performed through spreading ctivation in the interconnected semantic network contained in he long-term memory. He developed a general theory of cogni- ion, which focused on the processes of memory, known as the daptive Control of Thought (ACT) theory. The ACT theory represents knowledge in a semantic network,

nd focuses on how memory processes are defined within this etwork. Anderson also developed the concept of cognitive units hich consist of units of memory involving a unit node and a

et of elements within this node. Concepts, images and tempo- al strings of words can all be thought of as cognitive units. The inks between nodes are not of the same strength, however. All

emory traces or links between cognitive units/nodes have a ype of signal strength associated with them. The first success- ul trial or stimulation will establish the memory trace with the trength of one unit, and each subsequent trace will increase the trength by one additional unit. The strength of a memory trace s important in determining its probability and speed of retrieval. nce traces are formed between units they are not lost, but they

an decay if they are not regularly activated. If traces are acti- ated regularly, they will strengthen, and spreading activation ay occur more rapidly. This effect of this process is that we

an more readily retrieve cognitive information about regularly ssociated concepts. In summary, our experiences throughout ife establish a vast interconnected network of semantic nodes, hich are connected by links of varying strength. Each node

an be thought of as a cognitive unit, which can be activated by xternal stimuli. Levels of activation of subsequent nodes are ependent on the degree of association to the initially activated ode. The present research will address some of the implications f musical congruity in the context of these well-established oncepts.

Berkowitz’s ‘priming effects’ perspective (Berkowitz and ogers 1986) represents a bridge between ACT and processes

nvolving music in commercial contexts. It holds that present- ng mass media stimuli with a particular meaning (e.g., a movie eaturing a violent battle) primes concepts that are semantically elated (e.g., aggressive thoughts), thereby increasing the likeli- ood that the related concepts are activated and employed (see o and Berkowitz 1994). The activated mental constructs influ- nce how a person perceives and evaluates other constructs (see omke, Shah, and Wackman 1998). Cognitive structures that ave recently been activated are said to be advanced to the top f a mental set of files, making them easily accessible for a lim- ted time period (see Wyer and Srull 1989). If applicable, these

chemata will be employed in evaluations taking place in the mmediate future.

Schubert, Hargreaves, and North (2014, p. 1) applied his framework to musical likes and dislikes, arguing that,

of Ret

“ p p a t e i f M a t t t

p ( i i w A u p i f t m t

r r p p c b a t t m i fi o

i e t v ( l s q c i r t i H l

d c

a “ e e e o t i M m s p

r a v r s r K d t c c

H n n a

H n g n

M

P

s 1 t c a c e

D

A.C. North et al. / Journal

Spreading activation through previously associated networks is roposed as a pleasurable outcome of musical engagement”. Of articular relevance to the present research, Schubert et al. also rgue that this process of spreading activation includes responses o music, “stemming from the characteristics of the individual ngaging in the musical activity (whether listener, composer, mproviser, or performer), the situation and contexts (e.g., social actors), and the music (e.g., genre, structural features)” (p. 1).

ore practically, the model provides a framework in which all spects of the musical milieu interact in creating meaning, so hat although Schubert et al. address overtly esthetic responses, he same mechanism applies with equal relevance to responses o products and services experienced in the context of music.

Although prior research has accumulated evidence that music rimes semantically related concepts in an advertising setting Oakes 2007), there is an important difference between music n advertising and atmospheric music in a retail setting. Music n advertising is part of the focal stimulus. It is interspersed ith product information and it is likely to capture attention. tmospheric music, on the other hand, is a background stim- lus. It is external to the product information consumers are rocessing visually and aurally, and recalling from memory, n evaluating products and making choices. It is not yet clear rom previous research that music influences product evalua- ion and choice by priming conceptually relevant concepts in

emory, when it is a background stimulus external to the focal ask.

This omission is at least partly explained because previous esearch on music congruity effects has taken place in actual etail settings, where measures of the internal processes of shop- ers are difficult to obtain. If direct evidence that music affects roduct evaluation and choice by priming music-congruent con- epts in memory could be collated, however, it would have two enefits. First, it would place this growing body of evidence into

theoretical context that would guide future research. Second, here would be clear commercial implications; by understanding he theoretical basis of the effect, retailers would be able to use usic to improve financial performance by deliberately select-

ng atmospheric music that activates concepts relevant to specific nancial objectives (i.e., playing classical music to increase sales f premium brands in a store).

The three experiments described here address three gaps n the literature on musical congruity. First, there is no direct vidence substantiating the claim that music primes activa- ion of product-related nodes, which could be demonstrated ia enhanced recall of products related to a musical prime Experiment 1). Second, if the processes outlined here under- ie musical congruity, the effect should vary depending on the trength of the relationship between the music and products in uestion, so that use of music to activate the concepts ‘edu- ated’, ‘prestigious’, and ‘sophisticated’ should have greater nfluence on purchase decisions for ‘social identity’ products ather than ‘utilitarian’ products (Experiment 2). Third, given

hat cognitive load has repeatedly been shown to increase biased nformation processing (Ford and Kruglanski 1995; Gilbert and ixon 1991; Newman 1996), greater cognitive load should also

ead to musical congruity having a stronger effect on consumer

t c l

ailing 92 (1, 2016) 83–95 85

ecision-making than when those decisions are taken under omparatively light cognitive load (Experiment 3).

Experiment 1: Product Memory and Choice

Research indicates that certain styles of music are associ- ted with specific nations, such that music can be identified as French”, “Chinese”, “Indian”, etc. even by listeners with oth- rwise limited knowledge of the corresponding culture (Boer t al. 2013; Folkestad 2002). Moreover, national music styles can voke concepts and images congruent with cultural stereotypes f that country, so playing German music might make consumers hink of beer and bratwurst, whereas French music might evoke mages of wine and the Eiffel Tower (North, Hargreaves, and

cKendrick 1997, 1999; Tekman and Hortacsu, 2002). Experi- ent 1 capitalizes on these possible congruities between musical

tyles and nations to examine the effect of musical primes on roduct choice and recall.

If specific songs and musical genres prime semantically elated concepts in memory, then these concepts should remain ctive even when the music stops playing. Moreover, the acti- ated concepts should facilitate the encoding and retrieval of elated information in memory. Although prior research has hown that music influences the subsequent recall of concur- ently presented, congruent information (Kellaris and Rice 1993; ellaris, Cox, and Cox 1993), this memory effect has not been irectly linked to product choice. Specifically, in addition to esting whether national music can enhance recall of related oncepts, Experiment 1 focuses on whether it can also influence hoice, as specified in the following hypotheses.

1. Participants are more likely to choose a product of a given ational origin if the music playing concurrently is of the same ational origin compared to instances in which the music is of

different national origin.

2. Participants are more likely to recall a product of a given ational origin if the music playing is of the same national ori- in compared to instances in which the music is of a different ational origin.

aterials and Methods

articipants Participants were 120 Scottish undergraduate psychology

tudents, 60 males and 60 females, mean age 20.52 years (range 7–25 years, SD = 1.98 years). All received course credit for aking part in the study and were told that the study concerned hoices made in restaurants. Each successive participant was llocated to each condition in turn (i.e., participant 1 was allo- ated to condition 1, participant 2 was allocated to condition 2, tc.).

esign and Procedure

Employing a between-subjects design, participants were

aken to one of four research rooms. Each room contained four hairs and had a table in a corner on which sat an mp3 player and oudspeaker, which played the appropriate music at 70 decibels.

8 of Retailing 92 (1, 2016) 83–95

E ( w s 1 N u i p c w a

i t 3 d o m p b t w d o

a m m w w a a t t w

R

c w w = i i a c

o t t o t A o p

Table 1 Choice likelihood by type of music and type of cuisine.a

Music Frequency Menu choice

American food Chinese food Indian food

American Observed 27 2 1 Expected 18.5 5.8 5.8

Chinese Observed 12 15 3 Expected 18.5 5.8 5.8

Indian Observed 12 3 15 Expected 18.5 5.8 5.8

No music Observed 22 3 5 Expected 18.5 5.8 5.8

i t r A h ( i ( fi s

o t o e s s s o a w

i o ing, likeable, and familiar than others), Experiment 2 involves a very different manipulation of music genre, and a different

Table 2 Mean scores for recall of food items by type of background music played.a

Music Cuisine Menu items recalled

American American food 4.63 Chinese food 3.10 Indian food 3.70

Chinese American food 3.76 Chinese food 5.02 Indian food 3.17

Indian American food 3.63 Chinese food 3.23 Indian food 5.23

No music American food 3.46

6 A.C. North et al. / Journal

ach of the four rooms featured one of three types of music American, Chinese, Indian) or no music (see Appendix 1), hich played on a continuous loop. This mirrors music pre-

entation techniques used in numerous studies (Areni and Kim 993; North 2012; North, Hargreaves, and McKendrick 1997; orth, Shilcock, and Hargreaves 2003). A pilot study asked ten ndergraduate students to sort the nine candidate musical pieces nto three groups, but provided no further guidance: each of the ilot participants grouped the musical pieces according to the ountry each represented as per Appendix 1, indicating that this as their most salient characteristic when contrasted with one

nother directly. Participants were given a menu showing 30 main course food

tems (10 USA, 10 Chinese, 10 Indian). The menu was simple: he word ‘menu’ was printed at the top and it then listed the 0 items along with clear color photographs of each item. No escriptions were added and four random orderings of the items n the menus were created to avoid serial position effects. The ain course dishes included hamburger, hot dog, Chicago-style

izza, dim sum, noodle soup, chicken chow mein, vegetable iryani, chicken tikka masala, and rogan josh. For each of the hree experimental conditions, a number of travel and culinary ebsites were examined and the ten most commonly mentioned ishes were included on the menus employed. Equal numbers f nonmeat dishes were included in each condition.

Participants perused the menu for 5 min and were then given distractor task that involved solving anagrams of the names of arine creatures. They were then asked to recall as many of the enu items as they could and write them as a list. Participants ere then asked to state which single dish from the menu they ould choose to order. In debriefing sessions, participants were

sked to state which country was represented by the music, and ll identified this correctly. Participants were also asked why they hought that music was playing. Any participant suggesting that he music was intended to influence their choice from the menu as eliminated from the data analysis.

esults and Discussion

A Chi-square test was carried out to determine whether the ountry of origin of the single dish selected was associated ith the country of origin of the music played whilst the menu as examined. The relationship was significant χ2 (6, N = 120), 5.62, p < .001). The observed and expected frequencies shown

n Table 1 indicate that, consistent with Hypothesis 1, partic- pants were more likely to select menu items associated with

given country when the music playing was from the same ountry compared to when it was from a different country.

A mixed ANOVA was carried out to determine whether type f music (American, Chinese, Indian, no music) was related to he country of origin of the products recalled. The results showed hat recall for the menu items was influenced by the country of rigin of the music played F(6, 232) = 6.71, p < .001, such that

he recall results seem related to those concerning food choice. s shown in Table 2, consistent with Hypothesis 2 the number f dishes from each country recalled was higher when the music laying was associated with the same country compared to when

a Congruent conditions are in bold font.

t was associated with a different country. Tukey tests showed hat there was no difference between the music conditions in ecall of American items, although the greatest recall was in the merican music condition; that recall of Chinese food items was igher in the Chinese music than the American music condition p < .01) or the Indian music condition (p < .05), and also higher n the no music condition than in the American music condition p < .01) or the Indian music condition (p < .05). As such, the ndings concerning the no music conditions appear to indicate pecifically that lack of musical congruity inhibits recall.

These results suggest that the type of cuisine being recalled r selected was contingent upon the music genre. It is only when he national connotations of the music are congruent with those f the cuisine being considered that music congruity effects merge. However, music genre was manipulated by including everal pieces of music from a catalog of appropriately labeled elections. Although this is standard experimental practice in tudies manipulating genre (reference), this approach does leave pen the possibility that the observed effects are due to other spects of the music (i.e., liking, arousal, familiarity) correlated ith the genre manipulation. Although this seems unlikely given that manipulating genre

n this way would introduce a large amount of error variance n other music dimensions (i.e., some songs are more arous-

Chinese food 5.10 Indian food 5.87

a Congruent conditions are in bold font.

of Ret

t I o c

w a t s p m n p v e ( m c s p

e e t a 2 l a r L

p e t m e o t w s

H p c

H i m p

H p c

H i m p

M

P

d a 1

D

v t w p c o h u n a c m l p t p t

M

i p b u b o p ( c t i 1 2 t c t

A.C. North et al. / Journal

ype of product connotation to examine music congruity effects. t is based on the ‘sophisticated social identity’ connotations f classical music in contrast with the ‘pragmatic, utilitarian’ onnotations of country music.

Experiment 2: Cognitive Priming Using Multiple Music Genres

As the literature review indicates, the effects of congruity ould be expected to be wide-ranging, and so Experiments 2

nd 3 tested other implications of the phenomenon suggested by his literature. Several studies have indicated that the upmarket tereotype of classical music is associated with customers being repared to pay more for the same products than when other usical styles or no music are played. According to the cog-

itive priming/spreading activation explanation, these premium ricing congruity effects would appear to be based on the acti- ation of semantically related concepts like ‘wealthy’, ‘better ducated’, and ‘sophisticated’. Given this, we would expect that i) operationalizations of musical congruity employing classical usic would have a stronger effect if an upmarket stereotype,

onsistent with the music, is primed; and that (ii) the upmarket tereotype of classical music would have a stronger effect on roducts that share the upmarket connotations of the music.

If priming of related beliefs is the cause of musical congruity ffects leading to customers being prepared to pay more, how- ver, then we would not necessarily expect the effect to be limited o classical music. For instance, country music is well-known for

focus on everyday life and a self-sufficient worldview (see Ellis 010). Moreover, research on the sociology of music has estab- ished that country music fosters a utilitarian mindset, prompting

focus on pragmatic aspects of domestic life indicative of lower ather than higher socio-economic status (Blair and Hyatt 1992; ewis 1999).

So playing country music might lead to customers being repared to pay more for utilitarian products, as it might be xpected to lead to greater activation of information concerning he practicality of those products and so make the latter appear

ore valuable. Moreover, we would expect such effects to be xacerbated when these practical, self-sufficient connotations f country music are primed; and similarly we would expect he premium pricing effect of classical music to be exacerbated hen the upmarket connotations of the music are primed, as

tated in the following research hypotheses.

3a. Participants are prepared to pay more for utilitarian roducts when country rather than classical music is playing oncurrently.

3b. The effect of music genre on willingness to pay for util- tarian products is greater when the utilitarian connotations of

usic have been primed compared to when they have not been rimed.

4a. Participants are prepared to pay more for social identity roducts when classical rather than country music is playing oncurrently.

c u a l

ailing 92 (1, 2016) 83–95 87

4b. The effect of music genre on willingness to pay for social dentity products is greater when the utilitarian connotations of

usic have been primed compared to when they have not been rimed.

aterials and Methods

articipants Participants were 180 Scottish undergraduate university stu-

ents, comprising 90 females and 90 males. The number of males nd females was counterbalanced throughout. Age ranged from 8 to 25 years, with a mean age of 21.18 years (SD 2.21 years).

esign A three-way, mixed factor design was employed. Participants

iewed slides of ten social identity products and also ten utili- arian products, and stated the maximum monetary amount they ould be prepared to pay for each product. While doing so, inde- endent groups heard either classical music, which has upmarket onnotations, country music, which has utilitarian connotations, r no music. Of those exposed to each of the two types of music, alf were first exposed to a prime which highlighted either the pmarket connotations of classical music or the utilitarian con- otations of country. It was expected that the results would show n interaction between type of music and type of product, so that lassical music should lead to participants being prepared to pay ore for social identity products, whereas country music should

ead to participants being prepared to pay more for utilitarian roducts. A three-way interaction was also anticipated, in that hese effects should be magnified by the use of the relevant prime rior to presentation of the products. No priming took place in he no music condition.

aterials and Procedure

The ten utilitarian products were a bottle of aspirin tablets, ce, cough syrup, handheld fan, orange juice, toothbrush, dis- osable ball point pen, washing up liquid, white towel, and light ulb; and the ten social identity products were a greetings card, nisex friendship bracelet, a graduation portrait photograph, pin adge, unisex spray cologne, gold stud earring, university logo rnament, hair styling product, name stickers, and a pack of tem- orary tattoos. Products were partly taken from work by Shavitt 1989,1990) and Shavitt, Lowrey, and Han (1992), and were hosen so that they had just one attitudinal function. (A mobile elephone or a handbag, for instance, would have both social dentity and utilitarian functions – see Shavitt, Lowrey, and Han 992.) A pilot study asked a panel of three persons to sort the 0 products into utilitarian and social identity categories, and hese categorizations matched those of the experimenters in all ases. A second pilot study asked 20 participants (drawn from he same population as the final sample) to state whether first

lassical music and then country music were associated with an pmarket or a utilitarian stereotype, and all participants associ- ted classical music with the former and country music with the atter.

8 of Retailing 92 (1, 2016) 83–95

s 2 c t a p P w s

p t i e s s r m t o i p

1 p p m t e m i w m w p

R

v u t p b a s p w a w o e t i d

Table 3 Means (United Kingdom pounds sterling) for primed versus nonprimed utilitar- ian and social identity products, split according to background music condition.a

Product type

Utilitarian Social identity

Condition M SD M SD

Classical music Primed 2.90 (−.17) 1.70 5.09 (.15) 2.23 Not primed 3.43 (−.17) 1.38 4.50 (.15) 1.43 Country music Primed 4.76 (.45) 2.05 3.83 (−.03) 1.42 Not primed 3.09 (.45) 1.15 3.91 (−.03) 1.35 No music No prime 4.22 1.73 4.21 1.57

v

p c

t u t a c t t t o m w c i

F s e m t a c p i R o m o

a fi c t t

8 A.C. North et al. / Journal

Individual participants were played one of five pieces of clas- ical music or one of five pieces of country music (see Appendix ), with the specific piece heard by each participant selected by ycling through the possible alternatives in sequential order. Par- icipants were seated at a table and the music was played through

pair of headphones attached to a laptop that also presented the ictures of the products. Each product was presented on a single owerPoint slide, and each picture was of at least 100 KB and as downloaded from the internet. Beneath each picture was a

imple statement of the product shown (e.g., “A white towel”). Immediately before data collection began, half of the partici-

ants who listened to classical music were primed, as were half he participants who listened to country music. Primed partic- pants in the classical music condition were played 5 × 1-min xcerpts of classical music (see Appendix 2) while they were hown a sequence of images (selected via a pilot study) that are tereotypically upmarket (e.g., sports cars, royal palaces, luxu- ious holiday destinations). Primed participants in the country usic condition were played five x one-minute excerpts of coun-

ry music (see Appendix 2) while they were shown a sequence f images (also selected via a pilot study) concerning utilitar- an stereotypes (e.g., men chopping wood, cars being cleaned, eople using hygiene products).

Each participant was tested on an individual basis between 0.00 am and 9.30 pm over a two week period. Testing took lace in a quiet study room in a University library. As the music layed, participants viewed the 20 pictures of the products. The usic was played at a constant volume which was considered

o be sufficient to be heard clearly. Immediately after viewing ach product, participants wrote down (using pen and paper) the aximum monetary amount they would be prepared to pay for

t. In debriefing sessions, participants were asked whether they ere aware of the music playing, and if so, why they thought that usic was playing. Any participant suggesting that the music as intended to influence how much they would be willing to ay was eliminated from the data analysis.

esults and Discussion

A 3 (classical music, country music, no music) × 2 (prime s. no prime) × 2 (social identity product vs. utilitarian prod- ct) mixed ANOVA was performed on the maximum amount hat participants would be prepared to pay for each of the 20 roducts. A significant three-way interaction effect was found etween type of music, type of product and presence versus bsence of the prime (F(1, 175) = 12.75, p < .001). The means hown in Table 3 indicate that, as predicted by Hypothesis 3 articipants were prepared to pay more for utilitarian products hen country as opposed to classical music was played, and

s predicted by Hypothesis 4, more for social identity products hen classical as opposed to country music was played. More- ver, consistent with Hypotheses 3b and 4b, these effects were

xacerbated by the use of the relevant prime prior to presenta- ion of the products. Note also that, for both utilitarian and social dentity products, no music led to higher maximum prices than id incongruent music, suggesting that music incongruent with

l n i S

a The bracketed values are estimates of effect size r for comparisons of primed ersus nonprimed responses. Effect size estimates are duplicated to aid clarity.

roduct image can lead to a reduction in the maximum prices onsumers are prepared to pay.

Tukey tests confirmed this pattern. Specifically, they showed hat in the case of utilitarian products, when the utilitarian and pmarket connotations of country and classical music, respec- ively, were not primed, music had little or no difference in the mount that participants were willing to pay, whereas when the onnotations of country and classical music were primed coun- ry music led to higher willingness to pay for utilitarian products han did classical music (p < .001). In the case of social iden- ity products, when the utilitarian and upmarket connotations f country and classical music, respectively, were not primed, usic had little or no effect on the amount that customer were illing to pay, whereas when the connotations of country and

lassical music were primed classical music lead to higher will- ngness to pay than did country music (p < .05).

The pattern of results supports both experimental hypotheses. irst, the classical music premium pricing effect was obtained for ocial identity products (but not utilitarian products), and was xacerbated by the inclusion of priming; and second, country usic appeared to have a premium pricing effect on utili-

arian products, which was exacerbated by priming. Priming nd spreading activation offer a good explanation of music- ongruent consumer behavior. It is not simply the case that laying classical (or country) music in a retail environment pos- tively impacts on the prices that consumers are prepared to pay. ather the key factor is the correspondence between the type f product (in this case, social identity or utilitarian) and the usic in question (which, in this case, expressed social identity

r utilitarian values). This observed music genre x product type interaction effect

lso makes experimenter demand a less likely account of the ndings. Although demand effects might account for why classi- al music prompted participants to pay more for products related o social identity, there are three reasons why the effect of coun- ry music on willingness to pay for utilitarian products is not

ikely to be due to experimenter demand. First, as noted above, o participants indicated a possible link between the music play- ng and their willing to pay judgments in debriefing sessions. econd, it is unlikely that the distinction between ‘social image’

of Ret

a g d i s b m h I e t

m c p w n m d s s a K a p p p t S a o s T

H i t

M

P

t c a

D

w 2 w t m s

b fi w p l t f w h o l

P

i f p e w fi a t

M

c A s t t s s t P w s

P

E 9 a p m t p m i

R

c

A.C. North et al. / Journal

nd ‘utilitarian’ products would have occurred to participants iven that the products presented to them varied on so many imensions. Finally, it is unlikely that the observed pattern of nteraction would occur to participants (i.e., that country music hould induce them to pay more for half the products presented, ut less for the other half). The connection between country usic and pragmatism is not as obviously related to indicating

ow much one should be willing to pay for certain products. ndeed, a naïve participant trying to guess the research hypoth- sis might have presumed that the country music was intended o influence them to pay less for all the products presented.

Experiment 3: Music Congruity and Cognitive Load

The final experiment tested two hypotheses. As per Experi- ent 2, if priming offers an explanation of the impact of music on

onsumer evaluations, then Hypothesis 4b predicts that partici- ants who have received a ‘classical music is upmarket’ prime ill be willing to pay more for social identity products than onprimed participants. Moreover, we might further expect that usic should be particularly influential in cases where consumer

ecisions are made under high cognitive load (i.e., when con- umers can devote very little thought to the decision). Several tudies indicate that decisions made under high cognitive load re more susceptible to influence from cognitive biases (Ford and ruglanski 1995; Gilbert and Hixon 1991; Newman 1996). Ford

nd Kruglanski, employing a priming paradigm that involved resenting either the word ‘reckless’ or ‘adventurous’, had their articipants make judgments about an ambiguous situation. The rotagonist was described by participants in a manner consis- ent with the presented adjectives when cognitive load was high. imilarly evidence in a musical context is provided by Hansen nd Krygowski (1994) who found that creating a high degree f physiological arousal narrowed attentional focus and caused chema-consistent, sexist interpretations of rock music videos. his suggests the following hypothesis.

5. The effect of music genre on willingness to pay for social dentity products is greater when cognitive load is high compared o when it is low.

aterials and Methods

articipants Participants were 360 university undergraduate students in

he United Kingdom who were asked to take part in a study on onsumer products. There was an even gender split and mean ge was 18.8 years (SD = 1.10).

esign A three-way, between-subjects design was employed in

hich 360 university students were shown Powerpoint slides of 0 social identity products and asked to state the maximum they

ould be prepared to pay for each product. The first factor was

he music played, which was classical music or easy listening usic or no music. The second factor was whether the ‘clas-

ical music as upmarket’ stereotype was primed. Accordingly,

A p r e

ailing 92 (1, 2016) 83–95 89

efore the experiment began, half of the participants were played ve x one-minute excerpts of classical music (see Appendix 3) hile they were shown a sequence of images (selected via a ilot study) that are stereotypically upmarket (e.g., royal palaces, uxurious holiday destinations, race horses) or mundane (e.g., erraced houses, British holiday destinations, cows). The third actor was the cognitive load imposed upon participants, which as operationalized as how much time participants had to assess ow much they would pay for each product, namely five sec- nds in the high load condition versus one minute in the low oad condition (see Yeoh and North 2009).

ilot Study To ensure that the music was suited to the needs of the exper-

ment, a pilot study was undertaken within the same population rom which the main sample derived. All 20 participants in the ilot study labeled the pieces as ‘classical’ or ‘easy listening’ as xpected. The time gap between the presentation of each slide as also tested via the pilot study, and it was concluded that a ve second time frame was the minimum that would nonetheless llow participants to decide how much they would be prepared o spend on each of the 20 products.

aterials Each participant was played no music or one of five pieces of

lassical music or one or five pieces of easy listening music (see ppendix 3), with the specific piece heard by each participant

elected by cycling through the possible alternatives in sequen- ial order. Participants sat at a table and the music was played hrough a pair of headphones attached to a laptop that also pre- ented the pictures of the products. The products represented the ame ten social identity products as in Experiment 2, but utilized wo exemplars of each. Each product was presented on a single owerPoint slide, and each picture was of at least 100 KB and as downloaded from the internet. Beneath each picture was a

imple statement of the product shown (e.g., “A white towel”).

rocedure A quiet study room was reserved in a University library.

ach participant was tested individually between 10.00 am and .30 pm over a two week period. Participants put on headphones nd a Powerpoint file was opened on a laptop. Whilst the music layed, participants viewed the 20 pictures of the products. The usic was played at a constant volume which was considered

o be sufficient to be heard clearly. After viewing each product, articipants immediately wrote down (using pen and paper) the aximum monetary amount they would be prepared to pay for

t.

esults and Discussion

A 3 (no music, easy listening music, classical music) × 2 (low ognitive load, high cognitive load) × 2 (primed, nonprimed)

NOVA was performed on the maximum monetary amounts articipants indicated they would pay for the 20 products. This evealed a significant interaction effect and two significant main ffects. As regards the two main effects, the music played (or

90 A.C. North et al. / Journal of Ret

Table 4 Means (United Kingdom pounds sterling) for primed versus nonprimed, and high versus low cognitive load, according to background music condition.a

Presentation time

1 min 5 s

Condition M SD M SD

Classical music Mundane images 3.78 1.17 4.01 1.37

(.37) (.36) (*.43) (*.43)

Upmarket images 4.25 1.72 4.45 1.83 (.37) (.36) (*.35) (*.35)

Easy listening music Mundane images 2.99 1.13 3.23 1.25

(.45) (.37) (*.36) (*.36)

Upmarket images 3.78 1.49 3.31 1.36 (.45) (.37) (*.41) (*.41)

No music Mundane images 4.20 1.46 3.13 .90

(.47) (.36) (*.36) (*.36)

Upmarket images 3.38 1.21 3.84 1.27 (.47) (.36) (*.35) (*.35)

a The bracketed values are estimates of effect size r for comparisons of primed versus nonprimed responses. The bracketed values with an asterisk are estimates o e

n t ( s t l H i T t u u i e n h j η

T I p l ( e t m (

l t

b m e p m s a (

l b c i m l d w c a K i p

t e p e

p d r d 2 S ( t a e t a

Although prior studies have demonstrated music congruity

f effect size r for comparisons of high versus low cognitive load responses. All ffect sizes are duplicated to aid clarity.

ot played) had a significant impact on how much money par- icipants stated they were willing to pay for the 20 products F(2) = 10.4, p < .001). Table 4 (and subsequent Tukey tests) hows that, consistent with Hypothesis 4a, participants who lis- ened to classical music were willing to pay more than those who istened to easy listening music (p < .001) or no music (p < .05). owever, counter to Hypothesis 4b, the music × cognitive prime

nteraction effect was not significant (F(2, 348) = 1.36, p = .258). he other significant main effect related to priming, such that

hose participants who had been primed with stereotypically pmarket images were prepared to pay more for the 20 prod- cts than the nonprimed participants who had viewed mundane mages (F(1) = 3.6, p < .001). There was no significant main ffect of cognitive load on maximum price judgments. A sig- ificant three-way interaction effect (F(1, 119) = 3.7, p < .001), owever, indicated that music, priming and cognitive load were ointly related to participant responses (F(1) = 5.3, p < .001, 2 = .05). Means and effect sizes concerning this are shown in able 4. All estimates of effect size were of medium strength. n the case of the one minute presentation times and mundane riming images, Tukey tests showed that easy listening lead to ower willingness to pay than either classical (<.05) or no music p = .001); whereas the upmarket priming image led to no differ- nces between the conditions in willingness to pay. In the case of

he 5 s presentation times and mundane priming image, classical usic lead to higher willingness to pay than either easy listening

p < .05) or no music (p < .05); and the upmarket priming image

e a a

ailing 92 (1, 2016) 83–95

ead to classical music being associated with higher willingness o pay than easy listening music (p < .05).

The results indicate that listening to classical music in the ackground led to the student participants being prepared to pay ore for the 20 social identity products, that this effect was

nhanced when the ‘classical music is upmarket’ schema was rimed, and when the product information was presented in a anner that led to higher cognitive load. As such, both hypothe-

es were supported, suggesting that (i) persons who are primed ccordingly are willing to pay more for consumer products, and ii) high cognitive load exacerbates this effect.

The effects of the prime and the higher level of cognitive oad demonstrate that the semantic associations of music can e activated, and also that these effects are magnified under onditions in which participants experience constraints on their nformation processing as a consequence of the environment. We

ight therefore expect to see these effects under a variety of real ife in-store conditions in which consumers experience greater emands on their information processing (e.g., when crowded, hen short of time, when they have a long list of items to pur-

hase). Given that schemas become stronger and more readily ccessible the more they are primed (Bargh 1984; Higgins and ing 1981), we would expect that brand identity effects and will-

ngness to pay more would be increased if similar music were layed each time customers entered the store.

It should also be noted that the moderating effect of cogni- ive load casts further doubt on experimenter demand as a rival xplanation for the findings. Demand effects are the result of articipants’ cognitive construal of an experimental setting. As xplained by Zizzo (2010):

“. . .there is a decision or set of decisions that is in the hands of subjects and which is not fully defined by the experimenter or by peers. Subjects then try to make sense of the unfamiliar and incompletely defined experimental environment based on the instructions, cues and feedback they receive: that is, they have a job of task construal. . .” (p. 77)

Hence, any manipulation that restricts the ability of partici- ants to mentally construe an experimental task should inhibit emand effects. The results reported here are also consistent with esearch finding that high cognitive load limits various forms of eceptive behavior by research participants (Duffy and Smith 014; Schulz et al. 2012; van ’t Veer, Stel, and van Beest 2014). ince experimenter demand is essentially a form of deception i.e., making the researcher think the experiment “worked”), he less time participants have to think about what is happening round them, the less their behavior should be driven by demand ffects. Yet, in Experiment 3 the effects of music on willingness o pay actually increased when participants had less time to think bout their decisions.

Conclusions

ffects on product choice, and have identified aspects of music ssociated with its cultural meaning, the research reported bove is the first to show that atmospheric or background

of Ret

m c p a

e e K i I s a e t i a m i c u p T o w t r

i l e p t g r a a c f g a

a f e u g a v o f p a m p

a E

t i p a s s i s b o e w e

i t p r p i S o c f

d t e C e l b c i d

n v L p t o c c o l i T d l o t a

A.C. North et al. / Journal

usic influences product choice via its ability to prime related oncepts in memory. Once primed, these concepts influence roduct perception and guide choice to be congruent with the ctivated semantic network.

Experiment 1 showed that atmospheric music influenced the ncoding of menu items such that congruent items were more asily recalled at a later point in time. This result is similar to ellaris and Rice (1993) and Kellaris, Cox, and Cox 1993 find-

ng in an advertising setting, but with an important distinction. n their study, music was clearly a component of the advertising timulus that participants were paying attention to rather than n irrelevant background feature of the information processing nvironment. The results reported here suggest that music facili- ates the encoding of semantically congruent material even when t is not directly associated with the stimulus to which consumers re paying attention. Experiments 2 and 3 extend the effects of usic congruity to judgments of a product’s value, measured

n terms of consumers’ willingness to pay. Music that primes oncepts associated with an upmarket positioning of a prod- ct increases the amount consumers are willing to pay for that roduct, relative to music that connotes an economy positioning. his theoretical account of music congruity effects has numer- us practical implications concerning the circumstances under hich musical congruity might operate, and there is a clear need

o test and understand these further. Some directions for future esearch in these areas are discussed below.

In all three experiments, music genre was manipulated by ncluding several pieces of music from a catalog of appropriately abeled selections. Although this method is often employed in xperimental studies manipulating genre, it does leave open the ossibility that the observed effects are due to other aspects of he music (i.e., liking, arousal, familiarity) correlated with the enre manipulation. However, this seems unlikely for at least two easons. First, manipulating genre in this way would introduce

large amount of error variance on other music dimensions, nd so work against finding significant results. In short, some ountry songs, for instance, are more arousing, likeable, and amiliar than others. Choosing several songs to represent each enre makes it unlikely that the resulting effects are due to these spects of music.

Second, the moderating effects observed in Experiments 2 nd 3 are inconsistent with the effects of variables like arousal, amiliarity and liking. Specifically, in Experiment 2 the mod- rating variable of utilitarian versus social identity products is nlikely to be due to these variables (participants might find one enre of music to be more familiar, likeable, or arousing than nother, but why would these effects reverse for social identity ersus utilitarian products? Likewise, in Experiment 3 the effects f the priming manipulation and the cognitive load manipulation urther rule out the effects of these variables (i.e., why would the riming manipulation magnify the results if arousal, pleasure, nd liking were the underlying mechanisms?). Even in Experi- ent 1, the effects of genre depended on what kind of cuisine

articipants were choosing or trying to recall. These same interaction effects rule out experimenter demand

s an alternate explanation for the findings. In particular, in xperiment 2 the relationship between the country music and

a

t p

ailing 92 (1, 2016) 83–95 91

he judgment task would more easily be interpreted by partic- pants as demanding that they decrease their willingness the ay. It is unlikely that the distinction between ‘social image’ nd ‘utilitarian’ products would even occur to them given that o many products, differing on so many dimensions, were pre- ented to them. It is even less likely that the observed pattern of nteractions would occur to participants (i.e., that country music hould induce them to pay more for half the products presented, ut less for the other half). Likewise, in Experiment 3 the effects f music genre were stronger when cognitive load was high, ven in the absence of the cognitive priming manipulation. Why ould giving participants less time to identify possible demand

ffects (i.e., 5 s) increase the effects of music? More generally, while a certain rival hypothesis like famil-

arity might be a plausible account of some of the findings, he introduction of different manipulations of genre and multi- le moderating variables over three experiments systematically ules out these rival hypotheses. Only cognitive priming can lausibly account for all of the findings across the three stud- es. In this sense, we have adopted the comparative approach of ternthal, Tybout, and Calder (1987) and the Bayesian approach f Brinberg, Lynch, and Sawyer (1992) for testing a theory by onducting multiple studies to rule out alternative explanations or the findings.

A limitation of the cognitive priming explanation is that it oes not obviously suggest a preference for products or retailers hat are consistent with what has already been primed. The influ- nce of music is essentially a context-effect on product choice. onsumers may develop a preference for selected options after xperiencing it, but the congruity effect does not necessarily ead to a long-term effect on product choice. Hence, if the delay etween exposure to the music and product choice is long, or the onsumer is exposed to stimuli priming very different concepts n memory in the interim period, music congruity effects will iminish.

Another limitation of this theoretical account is that it does ot specify whether the cognitive priming and spreading acti- ation requires conscious awareness of the music. To illustrate, aran, Dalton, and Andrade (2011) found that, whilst brands roduce priming effects, brand slogans do not. In fact, they found hat slogans produced reverse priming effects (i.e., inducing the pposite behavior implied by the slogans). The authors con- luded that slogans are perceived as a persuasion tactic, and that onsumers do not necessarily accept them, but perception of the verall brand is not influenced by this. Does background music, ike a brand, operate outside immediate conscious awareness, or s it more like a slogan, operating within conscious awareness? he spreading activation approach discussed earlier implies that eliberate processing of the music, product, or context should ead to a stronger pattern of activation that increases the salience f concepts and objects stored in memory at some distance from he node originally activated. In contrast, a lack of conscious wareness should lead to weak activation of related nodes, and

relatively narrow spread of subsequent activation. Another direction for future research would be to iden-

ify the specific aspects of music most salient in determining articipants’ responses (e.g., instrumentation, lyrics, etc.). For

9 of Ret

e p i o w w m l g

c t s g o a J G b c 1 R r e

i t a i c i B m p

k R c M t a p d e 2 i a a w p r o

a ( c p

f d t v t s l (

s c t i c e r m t a p i i a

m w r m e t o a o w G r H v t ( v m t v

b c i s i o u c

2 A.C. North et al. / Journal

xample, it is likely that the lyrical content of country music lays a major role in activating pragmatic, utilitarian concepts n memory (Ellis 2010), whereas classical music involves little r no lyrical content in the listeners’ native language, so lyrics ould play a less obvious role in activating concepts associated ith prestige or social identity. Subsequent investigations into usic congruity could adopt a ‘molecular’ perspective in iso-

ating the contributions of various components of each music enre.

In terms of retail management, the importance of selecting ongruent in-store music is suggested a number of ways. First, he formats or channels offered by suppliers of in-store music uggest connections between music genres and product cate- ories implicitly, if not explicitly. For example, one supplier ffers a music format labeled ‘Japanese Traditional’, described s ‘perfect’ for ‘Japanese steak house restaurants and similar apanese-themed settings’. Another genre, labeled ‘Southwest rill’, offers a blend of ‘the old Southwest’ and ‘south-of-the- order Spanish flair’ that is described in terms of Tex-Mex uisine, suggesting music congruity as examined in Experiment . The same supplier offers a genre labeled ‘Luxury Upscale etail’ described as ‘elegant and sophisticated, ideal for luxury

etail stores’ (Anonymous 2015), suggesting music congruity as xamined in Experiments 2 and 3.

So companies that supply in-store music to retail chains and ndependent stores describe their offerings in terms of the seman- ic associations of each music format. The research reported bove identifies the specific process by which these intuitive deas actually influence retail sales. By evoking semantic asso- iations relevant to a specific country, music enhances recall and ncreases the likelihood of choosing items from that country. y activating concepts associated with a desired social identity, usic increases how much consumers are willing to pay for

roducts relevant to that social identity. Unlike many other aspects of music (i.e., tempo, volume,

ey), music congruity can influence the products shoppers buy. etailers have multiple motives for promoting the sale of spe- ific products, including mannaging inventory levels (Achabal, cIntyre, and Smith 1990), taking advantage of trade promo-

ions (Haans and Gijsbrechts 2011), and countering competitor ctivity (Bolton and Shankar 2003), and they can use a variety of romotional tools (i.e., price discounts, in-store signage, special isplays, in-store coupons, etc.) to achieve this end (Ailawadi t al. 2009; Bolton and Shankar 2003; Kamakura and Kang 007). But unlike these promotion tools, congruent music is less ntrusive, so it can be used to influence purchases of wine from

specific region (North, Hargreaves, and McKendrick 1997) or t a specific price point (Areni and Kim 1993) in much the same ay that special displays (Areni, Duhan, and Kiecker 1999) or rice discounts (Adams and Beenstock 2012) do, without a cor- esponding intrusion into merchandizing display space or a loss f margin, respectively.

Moreover, the dimensions on which musical genres differ

re remarkably similar to ‘generic’ brand positioning strategies Blankson and Kalafatis 2004), suggesting that specific musi- al genres could be used to trigger purchases (a) at different rice points, (b) for different usage occasions, (c) of products

r p

e

ailing 92 (1, 2016) 83–95

rom different countries/regions of origin, (d) by consumers with iffering levels of expertise, and (e) of brands with specific posi- ions in the minds of shoppers (i.e., friendly vs. serious, urban s. rural, liberal vs. conservative, etc.), which is perhaps one of he reasons why commercial providers of in-store music have o many parameters in the search engines used to sort through arge electronic music collections to generate play lists for clients Areni 2010).

In addition to relying on sophisticated search engines, these uppliers provide music consultants who advise large retail hains as to the search engine dimensions most relevant for heir stores. This level of sophistication suggests that identify- ng congruent music is beyond the expertise of even large retail hains, or conversely, that retailers who do not rely on such xpertise may inadvertently play incongruent music. Indeed, esearch indicates that retailers sometimes play inappropriate usic in their stores. Experimental research in actual retail set-

ings often involves introducing a ‘test’ condition in the form of different genre of music, which improves sales relative to that reviously played in stores (Areni and Kim 1993). So, confirm- ng that music congruity drives retail sales performance holds mportant implications for managerial decisions regarding the ppropriate store environment.

To illustrate how different types of retailers might apply usic congruity effects, it might be useful to consider selling ine in various types of retailer formats. A retailer with a nar-

ow mix of wines from France might play stereotypically French usic (North, Hargreaves, and McKendrick 1997), which would

licit congruent concepts in the minds of shoppers and, poten- ially generate incremental sales relative to playing irrelevant r incongruent music. Even in the absence of a sales effect, ctivating congruent associations would enhance authenticity f the retailer (Areni 2010). However, if the retailer featured ines from multiple countries the music could be varied (i.e., erman, Spanish, Italian, etc.) depending on which wines the

etailer wanted to promote in-store in a given period (see North, argreaves, and McKendrick 1997). If the merchandize also aried by price point, then classical (i.e., expensive) versus con- emporary pop (i.e., inexpensive) selections could be played Areni and Kim 1993); and of course, if the wine assortment aried by region and price point, it would be possible to pro- ote more specific ranges of merchandize by combining these

wo dimensions of music congruity (i.e., classical German music ersus contemporary French music).

In limited line mass merchants like ABC Liquor, with a much roader mix of alcoholic beverages, additional aspects of music ongruity might be considered. For example, bourbon whisky s associated with southern states like Kentucky and Tennessee, o playing blue grass or country music might trigger purchases n that product category (Sweet 1996). For supermarkets, which ften sell alcoholic beverages among an even wider mix of prod- ct categories, the number of dimensions for considering music ongruity would increase dramatically, which may be another

eason for the number of search dimensions used by commercial roviders of in-store music.

But by verifying the process by which music congruity influ- nces sales, the present research provides additional direction for

of Ret

f o t i e e l f t u c s r

e p m c h t j ‘ a

t p t s p l d T r

o t a p a u a c w c p f f q

A

A

A

A

A

A

A

A A

A

A

B

B

B

A.C. North et al. / Journal

uture studies with relevance to retail managers. For example, ne interesting avenue for further research would be to explore he relationship between music congruity and product familiar- ty. On one hand, the cognitive priming and spreading activation xplanation implies that music congruity would have weaker ffects for unfamiliar products for which semantic networks are ess well-developed and inter-connected compared to those for amiliar products. On the other hand, consumers are more likely o have knowledge and attitudes toward more familiar prod- cts. If music congruity affects choice simply by suggesting a ontext-specific basis for comparing products, this may not off- et a more established product preference in memory. Future esearch would do well to compare these rival hypotheses.

Retailers may be able to exacerbate congruity effects by mploying a prime similar to that presented to participants in the resent study. For instance, businesses wishing to prime the pre- ium pricing effect of classical music on social identity products

ould use upmarket in-store pictures similar to those employed ere. Interestingly, no music led to higher values being assigned o the products than did incongruent music which indicates that, ust as the ‘right’ music has a premium pricing effect, so the wrong’ music has an effect that is worse than playing no music t all.

Likewise, the current findings regarding cognitive load imply hat the effects of music congruity on sales should be more rominent for retail categories where shoppers (1) are under ime pressure, (2) spend little or no time making in-store deci- ions, or (3) make a large number of purchase decisions in a short eriod of time. Presumably, music congruity effects would be ess pronounced in retail categories where shoppers spend time eliberating a relatively small number of purchase decisions. hese ideas can be tested in field experiments involving multiple

etail categories. In the meantime, the present research has outlined a the-

retical basis for musical congruity effects by demonstrating hat appropriate music can improve consumers’ ability to recall nd recognize products; can be applied more readily to some roducts (e.g., social identity products and situations where lternate choices are equally familiar) rather than others (e.g., tilitarian products and situations where one of the competing lternatives is better-known). Such findings have clear impli- ations for retail/leisure premises, allowing more precise use, ith obvious implications for profitability. These effects appear

onsistent with well-developed theoretical models concerning riming, spreading activation, and memory; and the challenge or future research will be to develop this theoretical framework urther in a manner that addresses both theoretical and practical uestions.

Appendix 1. Music Employed in Experiment 1

USA. California girls, Surfin’ USA, Good vibrations. All per-

formed by The Beach Boys. Chinese. The beauty of the Himalayas, Little girl, Dancing and singing in the village. All performed by The Peking Brothers.

B

ailing 92 (1, 2016) 83–95 93

Indian. Aaja nachle, Dhadak dhadak, Dhoom machale. All performed by Sunidhi Chauhan.

Appendix 2. Music employed in Experiment 2

Classical. The four seasons (Spring) – Vivaldi; Fifth symphony – Beethoven; Cello concerto – Elgar; Nessun dorma – Puccini; Habanera – Bizet. Country. Your cheatin’ heart – Hank Williams; Stand by your man – Tammy Wynette; Crazy – Patsy Cline; I walk the line – Johnny Cash; I will always love you – Dolly Parton.

Appendix 3. Music employed in Experiment 3

Classical. The four seasons (Spring) – Vivaldi; Fifth symphony – Beethoven; Cello concerto – Elgar; Nessun dorma – Puccini; Habanera – Bizet. Easy listening. A whiter shade of pale; Granada; Mornings at seven; In the mood; Viva España (All performed by James Last and His Orchestra).

References

chabal, Dale D., Shelby McIntyre and Stephen A. Smith (1990), “Maximizing Profits for Periodic Department Store Promotions,” Journal of Retailing, 66, 383–407.

dams, Jean and Jane Beenstock (2012), “Price Discounts on Alcohol in a City in Northern England,” Alcohol and Alcoholism, 47, 187.

ilawadi, Kusum L., J.P. Beauchamp, Naveen Donthu, Dinesh K. Gauri and Venkatesh Shankar (2009), “Communication and Promotion Decisions in Retailing: A Review and Directions for Future Research,” Journal of Retail- ing, 85, 42–55.

lpert, Judy I. and Mark I. Alpert (1990), “Music Influences on Mood and Purchase Intentions,” Psychology & Marketing, 7, 109–33.

lpert, Mark I., Judy I. Alpert and Eliot N. Maltz (2005), “Purchase Occa- sion Influence on the Role of Music in Advertising,” Journal of Business Research, 58, 369–76.

nderson, John R. (1983), “A Spreading Activation Theory of Memory,” Journal of Verbal Learning and Verbal Behavior, 22, 261–95.

nonymous (2015), Playlists for Commercial Background Music,. http://www.audiosparx.com/sa/playlists/ (accessed May 19, 2015)

reni, Charles S. (2010), “Café tunes,” Café Culture, 23, 40. reni, Charles S. and David Kim (1993), “The Influence of Background Music

on Shopping Behavior: Classical Versus Top-Forty Music in a Wine Store,” Advances in Consumer Research, 20, 336–40.

reni, Charles S., Dale F. Duhan and Pamela L. Kiecker (1999), “Point-of- Purchase Displays, Product Organization, and Brand Purchase Likelihoods,” Journal of the Academy of Marketing Science, 27, 428–41.

ucouturier, Jean-Julien and Francois Pachet (2003), “Representing Musical Genre: A State of the Art,” Journal of New Music Research, 32, 83–93.

ailey, Nicole and Charles S. Areni (2006), “When a Few Minutes Sound Like a Lifetime: Does Atmospheric Music Expand or Contract Perceived Time?,” Journal of Retailing, 82, 189–202.

aker, Julie, Michael Levy and Dhruv Grewal (1992), “An Experimental Approach to Making Retail Store Environmental Decisions,” Journal of Retailing, 68, 445–60.

argh, John A. (1984), “Automatic and Conscious Processing of Social Infor- mation,” In Handbook of Social Cognition, Vol. 3, Wyer Robert S. Jr. and

Srull Thomas K. eds. Hillsdale, NJ: Erlbaum, 1–44.

erkowitz, Leonard and Karen H. Rogers (1986), “A Priming Effect Analysis of Media Influences,” in Perspectives on Media Effects, Bryant Jennings and Zillman Dolf, eds. Hilsdale, NJ: Lawrence Erlbuam Associates, 57–82.

9 of Ret

B

B

B

B

B

B

C

C

D

D

D

E

F

F

F

G

G

G

H

H

H

H

J

K

K

K

K

L

L

M

M

M

M

M

N

N

N

N

N

O

R

4 A.C. North et al. / Journal

lair, M. Elizabeth and Eva M. Hyatt (1992), “Home is Where the Heart is: An Analysis of Meanings of the Home in Country Music,” Popular Music and Society, 16, 69–82.

lankson, Charles and Stavros P. Kalafatis (2004), “The Development and Validation of a Scale Measuring Consumer/Customer-Derived Generic Typology of Positioning Strategies,” Journal of Marketing Management, 20, 5–43.

oer, Diana, Ronald Fischer, Ma L.G. Atilano, Jimena G. Hernández, Luz I.M. García, Socorro Mendoza, Valdiny V. Gouveia, Jason Lam and Eva Lo (2013), “Music, Identity, and Musical Ethnocentrism of Young People in Six Asian, Latin American, and Western Cultures,” Journal of Applied Social Psychology, 43, 2360–76.

olton, Ruth N. and Venkatesh Shankar (2003), “An Empirically Derived Tax- onomy of Retailer Pricing and Promotion Strategies,” Journal of Retailing, 79, 213–24.

rinberg, David, John G. Lynch Jr. and Alan G. Sawyer (1992), “Hypothe- sized and Confounded Explanations in Theory Tests: A Bayesian Analysis,” Journal of Consumer Research, 19, 139–54.

runer, Gordon C. II (1990), “Music, Mood, and Marketing,” Journal of Mar- keting, 54, 94–104.

aldwell, Clare and Sally A. Hibbert (2002), “The Influence of Music Tempo and Musical Preference on Restaurant Patrons’ Behavior,” Psychology & Marketing, 19, 895–917.

ollins, Allan M. and Elizabeth F. Loftus (1975), “A Spreading-Activation Theory of Semantic Processing,” Psychological Review, 82, 407–28.

omke, David, Dhavan V. Shah and Daneil B. Wackman (1998), “Media Priming Effects: Accessibility, Association, and Activation,” International Journal of Public Opinion Research, 10, 51–75.

ube, Laurette, Jean-Charles Chebat and Silvie Morin (2007), “The Effects of Background Music on Consumers’ Desire to Affiliate in Buyer-Seller Interactions,” Psychology & Marketing, 12, 305–19.

uffy, Sean and John Smith (2014), “Cognitive Load in the Multi-Player Pris- oner’s Game: Are There Brains in Games?,” Journal of Behavioral and Experimental Economics, 51, 47–56.

llis, Iain (2010), “Resistance and Relief: The Wit and Woes of Early Twen- tieth Century Folk and Country Music,” International Journal of Humor Research, 23, 161–78.

iske, Susan T. (1993), “Social Cognition and Social Perception,” Annual Review of Psychology, 44, 155–94.

olkestad, Goran (2002), “National Identity and Music,” in Musical Identities, MacDonald Raymond, Hargreaves David J. and Miell Dorothy, eds. New York: Oxford University Press, 151–62.

ord, Thomas E. and Arie W. Kruglanski (1995), “Effects of Epistemic Motiva- tions on the Use of Accessible Constructs in Social Judgment,” Personality and Social Psychology Bulletin, 21, 950–62.

arlin, Francine V. and Katherine Owens (2006), “Setting the Tone With the Tune: A Meta-Analytic Review of the Effects of Background Music in Retail Settings,” Journal of Business Research, 59, 755–64.

ilbert, Daniel T. and J. Gregory Hixon (1991), “The Trouble of Thinking: Activation and Application of Stereotypic Beliefs,” Journal of Personality and Social Psychology, 60, 509–17.

rewal, Dhruv, Julie Baker, Michael Levy and Glenn B. Voss (2003), “The Effects of Wait Expectations and Store Atmosphere Evaluations on Patron- age Intentions in Service-Intensive Retail Stores,” Journal of Retailing, 79, 259–68.

aans, Hans and Els Gijsbrechts (2011), “‘One-deal-fits-all?’ On Category Sales Promotion Effectiveness in Smaller Versus Larger Supermarkets,” Journal of Retailing, 87, 427–43.

ansen, Christine H. and Walter Krygowski (1994), “Arousal-Augmented Prim- ing Effects: Rock Music Videos and Sex Object Schemas,” Communication Research, 18, 24–47.

iggins, E. Tori and Gillian King (1981), “Accessibility of Social Constructs: Information Processing Consequences of Individual and Contextual Vari- ability,” in Personality, Cognition, and Social Interaction, Cantor Nancy

and Kihlstrom John, eds. Hillsdale, NJ: Erlbaum, 69–121.

ui, Michael K., Laurette Dube and Jean-Charles Chebat (1997), “The Impact of Music on Consumers’ Reactions to Waiting for Services,” Journal of Retailing, 73, 87–104.

S

ailing 92 (1, 2016) 83–95

o, Eunkyung and Leonard Berkowitz (1994), “A Priming Effect Analysis of Media Influences: An Update,” in Media Effects, Bryant Jennings and Zill- mann Dolf, eds. Hillsdale, NJ: Lawrence Erlbaum Associates, 43–60.

amakura, Wagner A. and Wooseong Kang (2007), “Chain-Wide and Store- Level Analysis for Cross-Category Management,” Journal of Retailing, 83, 159–70.

ellaris, James J. and Robert J. Kent (1992), “The Effects of Music on Tem- poral Perception; Does Time Fly When You’re Having Fun?,” Journal of Consumer Psychology, 1, 365–76.

and (1994), “An Exploratory Investigation of Responses Elicited by Music Varying in Tempo, Tonality, and Texture,” Journal of Consumer Psychology, 2, 381–401.

ellaris, James J. and Ronald C. Rice (1993), “The Influence of Tempo, Loudness, and Gender of Listener on Responses to Music,” Psychology & Marketing, 10, 15–29.

ellaris, James J., Anthony D. Cox and Dena Cox (1993), “The Effect of Back- ground Music on Ad Processing: A Contingency Explanation,” Journal of Marketing, 57, 114–25.

aran, Laran, Amy N. Dalton and Eduardo B. Andrade (2011), “The curious case of behavioral backlash: Why brands produce priming effects and slo- gans produce reverse priming effects,” Journal of Consumer Research, 37, 999–1014.

ewis, George H. (1999), “Traps, Troubles, and Social Issues: Country Music in the Social Science Classroom,” Popular Music and Society, 23, 61–82.

acInnis, Deborah and C. Whan Park (1991), “The Differential Role of Char- acteristics of Music on High- and Low-Involvement Consumers’ Processing of Ads,” Journal of Consumer Research, 18, 61–173.

attila, Anna and Jochen Wirtz (2001), “Congruency of Scent and Music as a Driver of In-Store Evaluations and Behavior,” Journal of Retailing, 77, 273–89.

illiman, Ronald E. (1982), “Using Background Music to Affect the Behavior of Supermarket Shoppers,” Journal of Marketing, 46, 86–91.

(1986), “The Influence of Background Music on the Behavior of Restaurant Patrons,” Journal of Consumer Research, 13, 286–9.

orin, Silvie, Laurette Dube and Jean-Charles Chebat (2007), “The Role of Pleasant Music in Servicescapes: A Test of the Dual Model of Environmental Perception,” Journal of Retailing, 83, 115–30.

orrison, Michael, Sarah Gan, Chris Dubelaar and Harmen Oppewal (2011), “In-Store Music and Aroma Influences on Shopper Behavior and Satisfac- tion,” Journal of Business Research, 64, 558–64.

ewman, John (1996), Give: A Cognitive-Linguistic Study, Berlin & New York: Mouton de Gruyter.

orth, Adrian C. (2012), “The Effect of Background Music on the Taste of Wine,” British Journal of Psychology, 103, 293–301.

orth, Adrian C. and David J. Hargreaves (1998), “The Effect of Music on Atmosphere and Purchase Intentions in a Cafeteria,” Journal of Applied Social Psychology, 28, 2254–73.

orth, Adrian C., David J. Hargreaves and Jennifer McKendrick (1997), “In- Store Music Affects Product Choice,” Nature, 390, 132.

, and (1999), “The Effect of Music on In-store Wine Selections,” Journal of Applied Psychology, 84, 271–6.

orth, Adrian C., Amber Shilcock and David J. Hargreaves (2003), “The Effect of Musical Style on Restaurant Customers’ Spending,” Environment and Behavior, 35, 712–8.

akes, Steve (2000), “The Influence of the Musicscape Within Service Envi- ronments,” Journal of Services Marketing, 14, 539–56.

(2003), “Musical Tempo and Waiting Perceptions,” Psy- chology and Marketing, 20, 685–705.

(2007), “Evaluating Empirical Research into Music in Advertising: A Congruity Perspective,” Journal of Advertising Research, 47, 38–50.

ose, Richard F. and Michael J. Wagner (1995), “Eminence Choices in Three

Musical Genres and Music Media Preferences,” Journal of Research in Music Education, 43, 251–60.

chubert, Emery, David J. Hargreaves and Adrian C. North (2014), “A Dynamical Minimalism Model of Musical Experience: Is

of Ret

S

S

S

S

S

S

T

T

v

W

W

W

Y

Y

Z

A.C. North et al. / Journal

Familiarity Everything?,” Frontiers in Psychology, 5,. Retrieved from http://www.frontiersin.org/Journal/10.3389/fpsyg.2014.00038/abstract

chulz, Jonathon, Urs Fischbacher, Christian Thoni and Verena Utikal (2012), “Affect and Fairness: Dictator Games Under Cognitive Load,” Journal of Economic Psychology, 41, 77–87.

cott, Linda M. (1990), “Understanding Jingles and Needledrop: A Rhetorical Approach to Music in Advertising,” Journal of Consumer Research, 17, 223–36.

havitt, Sharon (1989), “Products, Personalities and Situations in Attitude Functions: Implications for Consumer Behavior,” Advances in Consumer Research, 16, 300–5.

(1990), “The Role of Attitude Objects in Attitude Func- tions,” Journal of Experimental Social Psychology, 26, 124–48.

havitt, Sharon, Tina M. Lowrey and Sang-Pil Han (1992), “Attitude Func- tions in Advertising: The Interactive Role of Products and Self-Monitoring,” Journal of Consumer Psychology, 1, 337–64.

ternthal, Brian, Alice M. Tybout and Bobby J. Calder (1987), “Confirma- tory Versus Comparative Approaches to Judging Theory Tests,” Journal of Consumer Research, 14, 114–25.

weet, Stephen A. (1996), “Bluegrass Music and Its Misguided Representation

of Appalachia,” Popular Music and Society, 20, 37–51.

ekman, Hasan G. and Nuran Hortacsu (2002), “Music and Social Identity: Stylistic Identification as a Response to a Musical Style”,” International Journal of Psychology, 37, 277–85.

Z

ailing 92 (1, 2016) 83–95 95

urley, Lawrence.W. and Ronald E. Milliman (2000), “Atmospheric Effects on Shopping Behavior: A Review of the Experimental Evidence,” Journal of Business Research, 49, 193–211.

an ’t Veer, Anna E., Marielle Stel and Ilja van Beest (2014), “Limited Capacity to Lie: Cognitive Load Interferes With Being Dishonest,” Judgment and Decision Making, 9, 199–206.

hite, John Wallace (1997), “Radio Formats and the Transformation of Musical Style: Codes and Cultural Values in the Remaking of Tunes,” College Music Symposium, 37, 1–12.

ilson, Stephanie (2003), “The Effect of Music on Perceived Atmosphere and Purchase Intentions in a Restaurant,” Psychology of Music, 31, 93–112.

yer, Robert S. and Thomas K. Srull (1989), Memory and Cognition in its Social Context, Hillsdale, NJ: Erlbaum.

alch, Richard F. and Eric R. Spangenberg (2000), “The Effects of Music in a Retail Setting on Real and Perceived Shopping Times,” Journal of Business Research, 49, 139–47.

eoh, Joanne P.S. and Adrian C. North (2009), “The Effects of Musical Con- gruity on Choice Between Competing Pairs of Cultural Products,” Empirical Musicology Review, 4, 130–3.

hu, Rui (Juliet) and Joan Meyers-Levy (2005), “Distinguishing Between the

Meanings of Music: When Background Music Affects Product Perceptions,” Journal of Marketing Research, 42, 333–45.

izzo, Daniel John (2010), “Experimenter Demand Effects in Economic Exper- iments,” Experimental Economics, 13, 75–98.