Music influences on mood and purchase intentions

ABSTRACT

Mood and emotional elements of advertising and consumer response have re-ceived increasing attention from practitioners and scholars in recent years. Amongthe various sources of mood induction is the frequently used practice of accom-panying commercial messages with background music. This paper extends mar-keting's discussion of the role of music in influencing audiences, by using musictheory to analyze and investigate the effects of music's structural profiles onconsumers. The experiment's results indicate tha! music may have significantimpact on audience moods and purchase intentions, without necessaniy affectingintervening cognitions.

Music Influences onMood and PurchaseIntentions

Judy I. AlpertSt. Edwards UniversityMark I. AlpertThe University of Texas at Austin

This article examines the relationship of music and consumers' moods,attitudes and behaviors. A seminal paper in the Journal of Marketing byGorn (1982) studied music's influence from a classical conditioning per-spective and sparked renewed interest in music and other "background"elements of commercials and stores (e.g., Gardner and Hill, 1988). Thisarticle will discuss, integrate and build upon the work of Gorn and otherswho have provided theoretical and empirical insight into the ways in whichmusic may influence consumer responses.

The purposes of this study are to: (1) review major conceptual basesrelevant to discussing music and other nonverbal influences on mood, aswell as theory regarding the roles of informational and cognitive versusnoncognitive means of influencing buyer behavior; (2) expand upon ex-istmg research to investigate whether background music can influence

Psychology & Marketing Vol. 7(2): I()9-L13 (Summer 1990)© I99<) John Wiley & Sons, Inc. CCC 0742-6O46/90/02Ul()9-U25$(M.O0

MUSIC INFLUENCES

moods and common measures of advertising effectiveness; (3) begin toderive principles for predicting these effects, given analysis of the musicalcontent of an ad; (4) provide some suggestions for the construction ofeffective musical influences on emotions and product orientations; and (5)indicate directions for future investigations of musical content of advertisingand testing of its influence.

Theoretical Base and Literature Review*

In a recent paper (Gardner, 1985), mood was defined as a fleeting,temporary feeling state, usually not intense, and not tied to a specifiablebehavior. Moods can be positive or negative, such as cheeriness, peace-fulness or guilt and depression. According to Clark and Isen (1982), moodsoccur frequently and do not usually interrupt ongoing behavior. Moodsare distinguished from emotions, such as fear or rage, which are usuallymore intense, obvious and are tied to a specifiable behavior.

A number of studies have shown that mood has an impact on attitudesand behavior. We shall summarize major works in this stream. The presentstudy seeks to analyze the relationship between the profile of structuralelements in background music and its effects on audiences' mood. Giventhat mood is relevant and of increasing interest in consumer behavior, itssensitivity to the influence of background music in commercials is worthexamining. Finally, studies which have used music as an independent vari-able will be cited to discuss the current state of knowledge concerningmusic effects, leading to a consideration of the contributions to that knowl-edge proposed in the present study. Before addressing these literatures, itmay be useful to consider some of the underlying theories that are relevantto the integration of affect and emotional influences in communication.One useful framework is that of central and peripheral information pro-cessing.

Central and Peripheral Processing

Petty, Cacioppo, and Schumann's review of research in psychologyand consumer psychology (1983) concludes that neither central nor pe-ripheral information processing alone can explain the diversity of attitudechange results observed. The critical feature of the central route viewsattitude change as resulting from diligent consideration of information aperson views as central to the merits of an issue or product. Research

For brevity, this review is meant lo be representative but not exhaustive. Only a few keypapers in each section will be noted here. Additional references and a compar;iiive exhibitfor the music studies are available from the authors, as are more detailed statistical analysesand tables.

"no SUMMER 1990

MUSIC INFLUENCES

following this route focuses on the familiar topics of cognitive consistency,cognitive algebra, perception, learning, retention and a series of relatedissues. Petty, Cacioppo, and Schumann have presented research and ar-guments supporting the relevance of central route processing of informationunder conditions of high cognitive involvement.

They contrast this approach with a more peripheral route, in whichattitude change is due to the presence of simple positive or negative cues,or simple decision rules, and attitudes are less affected by issue-relevantarguments. In both routes, information or stimuli may be conveyed visu-ally, verbally, or in source/message characteristics. What distinguishes theroute seems to be the amount of conscious information processing, weigh-ing of evidence, and the like. It is believed that peripheral processing islikely under conditions of low cognitive involvement, as lower motivationto process information via the central route may evoke simpler heuristicsand cues to attitude formation (Petty, Cacioppo, and Schumann, 1983).They note that one of the more common approaches (there are others) toinfluence via peripheral processing is classical conditioning, which will besummarized next.

Affect and Behavior Conditioning and Peripheral ProcessingA stream of conditioning research in which direct transfer of affect

(or liking) results from message execution tactics such as music, humor,visual imagery, color and sex has been shown to influence consumers'feeling states (e.g., Gorn. 1982; Watson and Rayner, 1920). This researchsuggests that pairing a conditioned stimulus (a brand) with an uncondi-tioned stimulus (e.g., music, humor) produces emotional responses whichmay then be associated with the brand. Here, attitude change is allegedto have occurred due to the presence of simple positive or negative cues,without the necessity of logically intervening cognitive reactions. In fact,many have argued for and/or demonstrated behavioral change due toconditioning stimuli, even without attitude or preference change (Kroeber-Riel, 1984; Allen and Madden, 1985; Staats and Staats, 1957; Zajonc,1968; Zajonc, Markus, and Wilson, 1974). This may be relevant to non-informational, low involvement ads, where there is minimal motivation forcognitive processing, and the goal is to leave consumers with a favorable(but not necessarily conscious) "feeling" toward the product. When a prod-uct does not possess objective advantages, and is a simple product withfew attributes, persuasion may be more successful by using backgroundfeatures such as visual imagery or music (Kotler, 1974; Batra and Ray,1983). In addition, visual and other nonverbal aspects of an ad fit in wellunder low involvement conditions because they are effective in generatingfeelings, and because they are more easily and quickly processed thanverbal stimuli (Zajonc, 1980; Paivio, 1971).

On the other hand, there is diversity of opinion whether feelings

PSYCHOLOGY & MARKETING/VOL 7, NO 2 VvT

MUSIC INFLUENCES

automatically transfer between stimuli (affective conditioning), or if affec-tive states can influence attention and perception by affecting audiences'moods and prompt cognitive activity. Even affective conditioning mayinvolve some cognitive activity, although conscious links between feelingsand product evaluations may be absent. At this point no definite conclu-sions can be made regarding the type of cognitions that are included inaffective reactions, nor the extent to which classical conditioning may beinvoked to "explain" the process. The assertion that cognitive participationis not necessary for the occurrence of affect has been made (Zajonc, 1980;Zajonc and Markus, 1982; and Kroeber-Riel, 1984) and countered byothers (Lazarus, 1982, 1984; Tsal, 1985).

In addition, subjects may not normally be aware of how their feelingstates influence their evaluations and behavior (Clark and Isen, 1982; John-son and Tversky, 1983). Where feelings are concerned, there can be ar-guments supporting their effects coming through central as well asperipheral processing. Indeed, both processing routes may be involved toone degree or another, leading to variations in the resulting patterns of"stimulus—perceptions—beliefs—attitude—behavioral intention—behav-ior." A view of this phenomenon from a mood perspective is presentednext.

Mood Effects on Evaluation and Behavior

Research has shown that mood states have an important influenceon behavior, evaluation and recall (Gardner, 1985; Gardner and Vander-steel, 1984). While this general conclusion may not hold in all cases, Gard-ner notes that mood states appear to bias evaluations and judgments insimilar directions to mood, and she reviews studies detailing this process(1985).

The association between mood states and affective responses, judg-ments and behavior can be seen as both direct and indirect. A directaffective reaction may be viewed as a conditioned response when there aredirect linkages in associations in memory between mood states and affectivereactions (Griffitt and Guay. 1969), and mood states and behavior (seeGardner, 1985; Aaker, Stayman, and Vezina, 1988, for references). In-direct associations between feeling states and affective responses and/orbehavior include the influence of mood on information processing, or cog-nitive activity. Weiner notes that behavior may reflect interaction amongcognitions, feelings, and resulting expectations or future outcomes (1979).Mood may also affect evaluations by evoking mood-congruent thoughtsand affect the performance of the behavior by increasing the accessibilityof positive associations to the behavior (Bower. 1981; Clark and Isen, 1982;Goldberg and Gorn, 1987; Isen era/., 1978). To the extent that associations

112 SUMMER 1990

MUSIC INFLUENCES

are direct and involve little conscious processing of the relationship betweenthe advertisement's information and attaining one's goals, mood's effectsmay be seen as via the peripheral route. Indirect associations may operatewhen other issue-relevant cues are processed to yield attitudes and behav-iors in a manner affected by mood.

The likelihood that a host of behaviors may be performed appear tobe enhanced by positive moods (Gardner, 1985). Negative moods' effectson behavior may be more complex than the effects of positive moods (Isen,1984; Donnerstein, Donnerstein, and Munger, 1975; Cialdini and Kenrick,1976). For example, helping may be enhanced by some negative moodstates such as sadness (Baumann, Cialdini, and Kenrick, 1981) and not byothers such as frustration. This may be due to some evidence that negativemood states are not as homogeneous as positive ones (Isen, 1984), andthat behaviors seen to reverse unpleasant mood states (e.g., helping) mayovercome tendencies to enact mood-congruent behavior (e.g., with-drawal).

Moods can be affected by many different variables. Gardner andVanderstee! (1984) discuss studies of independent variables found to inducemood states. Although much work has been accomplished in the study ofmood induction, there remains a need for more theory regarding the rea-sons why communications (and other inducers) influence moods.

In view of the fact that music is a common element in commercials,and one which has a long history of mood inducement in a variety ofcontexts, the next section will focus on how music has been used as anindependent variable to effect moods, as well as other dependent variablesof interest to marketers.

Music EffectsMusic has been used in consumer behavior research, as well as com-

munications, psychology and music therapy research to determine its effectson behavior, preference and mood. Research investigating music effectsmay be divided into those which analyzed and/or manipulated the struc-tural and the sound elements of music (structural elements) and those whichdid not analyze nor manipulate structural sound elements (nonstructuralelements). Structural elements refer to the properties making up musicalsound such as melody, rhythm, harmony, major or minor modality, andtempo.

This distinction is an important one for this study because we areconcerned with the relationship between musical structure and mood ef-fects. While music has been used as an independent variable in someconsumer research, no causal effect has been explored between the musicalelements and the subject's mood. In studies where music was used fornonstructural purposes, it appeared that the musical properties themselveswere not of major interest, since they were not analyzed or manipulated.

PSYCHOLOGY & MARKETING/VOL 7, NO 2 rT3~

MUSIC INFLUENCES

hlonstructural Musical StudiesSome of the nonstructural literature indicates that music does influ-

ence what products consumers will select. Gorn (1982) suggests that pe-ripheral influences such as background music used in commercials maybecome associated with the advertised product (in memory, even if notconsciously), and influence product choice through classical conditioning.Mere exposure did not lead to liking, which apparently depended onwhether the target product, a pen, was presented with liked versus dislikedmusic.

The second experiment by Gorn (1982) provided support for hishypothesis that when subjects were not in a decisionmaking mode, thecommercial's impact appeared to be more influential in its appeal whenpresented with musical background as opposed to product information. Heconcluded that through classical conditioning, the product becomes asso-ciated with the positive feelings of liked music.

Bierley, McSweeney, and Vannieuwkerk (1985) extended Gorn'sstudies. Preference ratings for stimuli that "predicted" (preceded) pleasantmusic were significantly greater than preference ratings for stimuli thatpredicted the absence of music. In another extension of Gorn's work,researchers questioned the theory of affective conditioning and suggestedthe mood position of Bower (1981) and Isen (1984) as a possible explanation(Allen and Madden, 1985). Results indicated that there may have been aninteraction between subjects' thought processes and the moods invoked bythe "background"* stimulus in the ad (in their case, liked versus dislikedhumor). Music in advertising's possible effects on audience moods thusmay complicate the effects of "simple" conditioning by the music.

Park and Young (1986) extended this work by examining the impactof music versus no music on attitude toward the brand, the ad, and be-havioral intention under conditions of high cognitive, high affective andlow involvement towards the advertising situation. Under high cognitiveinvolvement, they found music to be a distraction, lowering these depen-dent variable scores, because it was unrelated to attribute-based messagecontents. In the low involvement condition, they found that music (whichhad been preselected as popular and liked) was associated with more pos-itive attitudes towards the brand than was no music. Under high affectiveinvolvement, the expected positive effect of music on brand attitude wasnot found, probably because the music selected did not really fit the imageof the product and affective theme. As Park and Young note, the music,"The Tide is High" by "Blondie" may have been incongruent with thestylish classic beauty of Ingrid Bergman and the consumers' intended self-concept stressed in the hair shampoo commercials for which this musicappeared as background.

Since many commercials are viewed in situations which involve con-sumers who are interested in the programs, and not in the commercials,

114 SUMMER 1990

MUSIC INFLUENCES

the audience may be largely comprised of potentially uninvolved, nonde-cisionmaking consumers rather than cognitively active problemsolvers. Inthis context, emotionally arousing components such as music, colors orlighting may exert strong but subtle influence on viewers' product attitudesand choices. Some of this impact may come via associations conditioned(by advertisements and other cultural learning) and hnked to the advertisedproducts. Others may come through an indirect route resulting from music'sinfluence (for example) on respondents' mood and other emotional re-sponses, which in turn affect information processing.

An illustration of music's power to affect subjects' emotional re-sponses was reported in a study by Rohner and Miller (1980), where sed-ative music showed a trend to decrease anxiety. Another study dealt withpersuasion, among other variables. Subjects had greater affective arousal,persuasion affect and attidudinai acceptance of the song's message withguitar accompaniment than without guitar accompaniment (Galizio andHendrick, 1972). Thus changes in the presentation of music influencedsubjects' responses.

Structural Music StudiesThe above studies have provided some insights into the effects of

liked music on brand attitudes under some conditions. However, moreappears to be involved than music familiarity and popularity in eliciting amood congruent with the commercial used (Park and Young, 1986). Infact a recent replication of the Gorn study (1982) by Keilaris and Cox(1987) failed to reproduce the positive effect of liked versus disliked music,after controlling for musical structural elements and possible demand ef-fects. They call for research on the influence of music's structural char-acteristics on cognitive and affective responses (such as consumer mood)toward the ad and the product. Thus we turn to a review of research intothe relationship between structural elements of music and audience re-sponses.

A study by Milliman (1982) suggests that slow tempo of instrumentalbackground music can significantly slow the pace of in-store traffic flow ofsupermarket customers, as opposed to fast tempo. In a follow-up studydiners stayed longer and consumed more alcoholic beverages when slowtempo instrumental background music was playing than when fast tempoinstrumental music was used (Milliman, 1986). An early study dealing withthe question of music's effect on shopping behavior found that significantlyless time was spent in the stores when the music was loud compared towhen it was soft (Smith and Curnow, 1966). These studies are informativeregarding possible shopping influence of tempo and loudness, although noinformation was given concerning the actual music used nor the levels ofother musical elements in the selections.

In research by Holbrook (1981) and Holbrook and Corfman (1985),

PSYCHOLOGY & MARKETING/VOL 7, NO 2 r i5~

MUSIC INFLUENCES

care was taken to analyze and manipulate elements of musical structure,including tempo, rhythm, dynamics, phrasing, timbre and speed. The de-pendent variable of interest was preference for the music itself, as affectedby the variations in musical elements, rather than variables such as moodand attitudes towards products, with the music taken as background. How-ever, the approach to music's structure was rigorous and could be extendedto study its influence on other dependent variables beyond preferencestowards the music.

Manipulation and analysis of musical material in a study by Infanteand Berg (1979) investigated the effects of using two identical melodies,(composed for this research) one in a major mode, and one in a minormode, on perceptions of communications.* Results indicated that majormodality had the greatest positive effect on viewers' perceptions when facialexpressions were sad or neutral, and when a situation was unpleasant.Music modality did not affect perception of a happy facial expression norhow favorable a pleasant situation was perceived by viewers.

The key basic research relating musical elements to emotional re-sponses was reported by Hevner (1935), who presented subjects with iden-tical pieces, controlling for all elements but major and minor modes. Sheconcluded that all of the historically affirmed characteristics of the twomodes were confirmed in her study. In later research, she also reportedassociations between musical elements such as fast tempo, loud dynamics,lively and varied rhythm, and high register with perceptions of the musicas happy, merry, graceful, playful. Musical elements such as slower tempo,quiet dynamics, unvaried rhythm, and low register were reported to besad, dreamy, and sentimental (Hevner. 1935, 1936). She noted that, al-though mode is never the sole factor which determines the way music isperceived, it is the most stable, generally understood and influential of anyof the elements in expressing the affective mood of music.

Meyer's (1956) theory of deviations from expectations in music sup-ports Hevner's findings. He explains that expectations of more regular andnormative musical progressions occur in the major modes, and thereforeare associated with the more normal human emotional states such as con-tentment, joy and calm. The minor mode is characterized by more forceful,complex departures from tones found in the major scales, and these de-viations have become associated in western culture with feelings of sadness.anguish and suffering. He states that although there are exceptions, minormodes have tended to be played in slower tempi than major modes because

Readers may recall that major musical modes are characterized by distances betweensuccessive notes or chords (intervals) that are one-half step (halflone) greater than thecorresponding minor intervals. Listening to examples cited in Table 1 may reinforce thedescriptions of the pieces.

T m SUMMER 1990

MUSIC INFLUENCES

the disjunct melodies with unusual skips were technically more difficult toplay and sing rapidly.

HYPOTHESES

The basic design for the present study involves exposing subjects toa range of musical "mood" selections and a range of mood-evoking prod-ucts within a given category. We test for the effect on subjects' moods andproduct evaluations of presenting products accompanied by music withvarying musical structural profiles. A mixed factorial design was used, inwhich each group was exposed to three greeting cards, randomly matchedwith happy, sad or no music playing, while the cards were viewed onbrightness-controlled slide pictures. Each music condition was presentedtherefore with each card, one third of the time (and vice versa for cards),leading to null hypotheses that average responses to each type of music,across a variety of dependent variables, should (in the absence of musicstructure effects) be equal. However, following from the music researchreviewed above, it is expected that:

H,: All else equal, music whose structural profile is "happy" will in-fluence listener moods to become more positive than music ana-lyzed a priori as "sad."

In addition, we believe that music evokes emotional responses, whichmay be more important than cognitive processing in the context of an adwith little objective product information. In this instance developing atti-tudes might not require much focused mental processing. In fact, it hasbeen argued and shown that behavior may emerge without the necessityfor clearly developed attitudes. Indeed, from a conditioning perspective,we might hypothesize that music structure, if it has impact on subjects'mood, would be more likely to cause variability in purchase intentions,and to a lesser extent, on perceived attributes and evaluations of the cards.Therefore:

Hi: Variations in musical structure may not necessarily influence per-ceptions of the "happiness" or "sadness" of the greeting cards.

Hv Variations in musical structure may not necessarily influence over-all attitude towards the greeting cards.

H4: Variations in musical structure will influence behavioral intentionstowards the greeting cards.

The subject would be induced, in effect, into a happy or sad moodwhile looking at cards and hearing happy or sad music in the background.

PSYCHOLOGY & MARKETING VOL 7, NO 2 ri7"

MUSIC INFLUENCES

If this mood is appropriate to the one which s/he feels when thinking aboutcommunicating with a friend who is away, then the subject may FEEL thiscard as appropriate in expressing feelings, even if the card itself is notconsciously, cognitively processed as "happy" or "sad." When responsesappear to be contingent upon stimulus-feeling associations, as in this in-stance, it may reflect a low-cognitive and peripheral processing situation.

Although conditioning theory and a number of studies noted abovemight support H2-H4, the limitations of the empirical study reported hereand the preliminary stage of research into music's impact in advertisingsuggest that these hypotheses be viewed as tentative. Hence the unusualwording of H2 and H:i. The following section discusses the methods em-ployed regarding sample, stimulus choice and analysis, measurements andthe experimental design used to test these notions.

METHODOLOGY

Pilots

Three different friendship greeting cards were sought to be used asconditioned stimuli, one perceived to be happy, one sad and one neutral.Three cards per subject were used in order to improve statistical powerthrough repeated measures, as well as improve generality by sampling fromthe domain of emotional range for greeting cards.

Since a primary main effect was beheved to be the feeling the musicevoked, it was felt that the less objective attribute-based and the moreaffect-based the card was, the less it would interfere with the treatmenteffects (Gorn, 1982). Therefore, friendship greeting cards appeared to bea logical choice for an emotionally oriented product. After exploratoryresearch and pretesting, three cards, one happy, one sad and one neutral,were used in the treatments. Thirty-five millimeter black and white slidesof these cards were used to present them to subjects in the main experiment,accompanied by music*

Next, ten classical pieces which were relatively unfamiliar were se-lected. All of the piano pieces were Preludes and were part of the BookI of the Well Tempered Clavier, by J. S. Bach, indicating stylistic uniformity.Following Hevner (1935, 1936) five pieces predicted to have "happy"

Average ratings on a five-point '"sad" (1) to '"happy" (5) scale for Ihe chosen cards were:Happy (4,47, the highest of the pretesled cards); Neutral (3.1); and Sad (2.6. the lowestof the pretested cards). Black and white slides were used to obtain control of looking timeand simplify the matching of relative brightness and emoticinal iniensiiy. Black and whiteslides are often used to simulate advertisements in experimental work (Smith and Hunt.1978).

Ti8 SUMMER 1990

MUSIC INFLUENCES

moods were in the major mode, and the five "sad" ones were in the minor.Since familiarity influences liking (Bradley, 1971), and unfamiliar uncon-ditioned stimuli produce stronger conditioning than familiar unconditionedstimuli (McSweeney and Bierley, 1984), selections were chosen for unfa-miliarity to the general and subject population.

Twenty-four subjects not in the main study listened to 30-secondexcerpts of the ten pieces, which were randomized on a prerecorded tape.Selections were judged as having the Gestalts happy and sad (Biller, Olson,and Breen, 1974), which were included with other adjectives to eliminatedemand effects. Respondents indicated how much they liked the piece,whether they would or would not buy the music, and how familiar themusic was.

Consequently, two pieces rated by respondents as unfamiliar andapproximately equally liked but having the Gestalts of happy and sad wereanalyzed for music structure.* Elements analyzed included the number ofmajor and minor harmonies occurring on the strong beats (Hevner, 1935),fast and slow tempi, loud versus more quiet dynamics, and fast, lively,energetic, versus slower, listless, unvaried rhythms (see Table 1). Theanalysis indicated support for Hevner's earher findings, in that the twocompositions pilot subjects rated as happy versus sad appeared to be thosewhich had internal definiteness and uniformity in harmony, tempo, dy-namics and rhythm.

The overall profile of the elements analyzed in the two compositionswere in strong contrast to each other. These pieces were used to simplifythe design and test whether selections with varying profiles of musicalelements might produce contrasting effects on respondent mood. The im-pact of specific elements (and interactions among them) was not tested inthe present study. Each of the elements was close to ends of the spectrumthat would be associated with happy versus sad music, in a consistent andtypical profile for each type. We felt that if the profile did not influencemood and other dependent variables of interest, analysis of specific ele-ments' effects would be of little relevance. As will be noted below, ob-taining manipulations with the entire profile suggests a number ofsubsequent studies, including those designed to examine relative impactof music elements and their interactions.

Obtaining equally liked background music selections with contrastingstructural profiles was a major purpose of this pilot study. It was somewhatgratifying to note that the pieces analyzed as structurally "happy" versus"sad" were perceived as such when attention was directed solely at the

* Average ratings and significance tests for differences between the pieces chosen to representHappy versus Sad were as follows: Perceived happiness: 4.3 versus 1.8 f = 8 58 p <0.001; Liking: 3.0 versus 3.2, / = 0.88, p = 0.40, n.s.d.; Would buy: 2.5 versus 2.6, f =0.55, p = 0.60, n.s.d.

PSYCHOLOGY & MARKETING VOL 7, NO. 2 m~

MUSIC

Feature

INFLUENCES

Structure

Levels

TABLE 1Profiles of Music Treatmeuts

Description

Harmony Selection 1 53 beats oul of 70 heard were in major, 17 beats in the minorMajor (happy) mode. All occurred on the strong beat of the measure.

(76% in major)Selection 2 29 beats out of 33 heard were in minor, all on strong beats.Minor (sad) four beats in the major mode. Only one out of four major

chords occurred on the strong beat, three occurred onweak beats. (88% in minor)

Tempo Selection 1 J • = 108 beats per minute as performed by Glen Gould.Fast (happy) Subjects heard 70 beats of 104 total beats in selection.Selection 2 J = 46 beats per minute as performed by Glen Gould.Very slow (sad) Subjects heard 33 beats of 188 total beats in selection.

Dynamics Selection 1 Lively (vivace).Loud (happy)Selection 2 Expressive (espressivo) throughout.Ouiet (sad)

Rhythm Selection 1 Rapid rhythmic activity alternating between bass and treble.Quick (happy) energetic, dynamic movement in 3/8 meter.Selection 2 Slow movement of repeated rhythmic motives in treble;Slow (sad) pedal in bass (same note repeated for first 16 beats), slows

down harmonic rhythm and overall movement in 4/4meter.

Music: -Prelude III in C# major" from Volume I of The Well Tempered Clavier by J. S, Bach (scleclion1).••Prelude XXII m Bb minor" from Volume I of The Well Tempered Clavierby). S. Bach (scleclion2).

music. It remained for the main study to determine, with a more sensitiveinstrument, whether using these selections as background in simulatedadvertisements would replicate hypothesized variations on subject's moodsin an advertising context.

Design and Procedure

The experimental subjects were students in three "Principles of Mar-keting" classes (the groups). Although the use of convenience samples ofstudents limits generalizability, their use in this study seems appropriatefor theory application and testing to investigate the basic process of musicalinfluence on emotions and attitudes (Calder, Phillips, and Tybout, 1981).Extensions using other samples would be justified if initial results werestimulating.*

The design was a mixed factorial, with repeated measures on 2 factors(card and music) and between subjects for the third factor (group). The

Specific musical passages might vary in evoked responsed between student and nonstudentsamples, particularly for contemporary or "pop" music. However, responses to Bach prel-udes used here might not vary substantially.

120 SUMMER 1990

MUSIC INFLUENCES

TABLE 2Experimental Design

Card/Music Card I Sad Card 2 Happy Card 3 Neutral

Group 1 Happy Music Sad Music No MusicGroup 2 Sad Music No Music Happy MusicGroup 3 No Music Happy Music Sad Music

key treatment manipulation was the sad, versus happy, versus no music,paired with different cards, since each group heard the same musical se-lections randomly matched with identical greeting cards (see Table 2).Across the three groups, each music type appeared first, second, and thirdan equal number of times to balance for order of musical exposure.

Each treatment session was given once during regularly scheduledclass periods. Each group heard the experimenter's introduction explainingthat the researchers were interested in consumers' preferences for greetingcards and their feelings about their advertisements. They were asked toview the simulated greeting advertisements, some of which would be ac-companied by music such as might be found in a commercial. They wouldbe given practice sessions with the mood monitor and rating scales for thegreeting cards, followed by the specific cards for which responses andratings were sought.

On the cover sheet of the questionnaire packet, respondents indicatedthe frequency of greeting card purchases. To restrict the sample to thosefor whom the product had relevance, persons who did not purchase greetingcards were later discarded from the analysis, leaving 48 usable subjects.Next, the measures were explained and two trial runs of the mood monitorwere administered along with sample cards and music. Each trial wasfollowed by practice ratings of the subject's mood, perceived attributes ofthe greeting card, including overall impression and purchase intention.Then came the first treatment card and music combination, during whichsubjects used the mood monitor. This was followed by a gap for completingthe scales for card attributes, evaluations, and purchase intention. Ad-ministering these measurements between treatments was done to lessenmemory problems as well as mitigate moods of prior treatment levels. Thesecond and third cards were seen next with the appropriate musical back-ground, accompanied by the same measurements.*

Order was balanced within each treatment so that each of the three music conditions waspresented equally often in the first, second, and third positions. The effect of the three-order presentations was tested, and for all cards there was no main effect of order nor anyinteractions with the other treatments on any of Ihe dependent variables. The responsesto ads accompanied by varying music conditions did not vary as a function of presentationorder, and there were no significant carry-over effects of moods, following the interval ofsubjects' providing responses lo the ads.

PSYCHOLOGY & MARKETING/VOL 7, NO 2 ^2^'

MUSIC INFLUENCES

Measurements

Measurements of each subject's feelings/moods were taken while theywere exposed to the treatments. Prior to the treatment, subjects were toldthat the researchers were interested in how ihey feel while viewing the ads.The mood monitor used here is an adaptation of the warmth monitordeveloped and used by Aaker. Stayman. and Hagerty (1986). The warmthmonitor was developed for use in measuring the perceived "warmth" oftelevision commercials, which is analogous to measuring in the presentstudy the perceived "happy/sad" feeling evoked by simulated greeting cardcommercials. The monitor provides a continuous sensitive measure of re-spondents' feelings during a commercial and is portable and inexpensive.*Their findings show good levels of reliability (test/retest correlations av-eraged 0.81) and validity (average correlation with galvanic skin responsefor warm commercials was 0.67).

In the present study, the mood monitor had five scale labels to reflectfeelings of "sad, moderately sad. neutral, moderately happy and happy."Subjects moved a pencil down the paper, to the left (sad), or to the right(happy), while viewing the card and hearing the corresponding music,indicating how sad or happy their feelings were at any given time. Themonitor was scaled from 0 (sad) to 100 (happy), depending on the pencilline's height from the left anchor, and readings were taken at five evenlyspaced percentiles of the respondent's drawn line. Since the illustrationwas constant, as was the prevailing musical mood throughout the 30-secondexcerpt, we defined subject mood as the average of five scores.

Following the exposure to stimulus slides and music and the simul-taneous measurement of mood, respondents turned the page and evaluatedthe greeting card on a series of ten semantic differentials. Imbedded inthis instrument of mostly "placebo" characteristics of cards (e.g., "original/unoriginal") was a scale designed to measure the perceived mood of thecard, scaled as happy sad. The next two measures wereoverall impression: favorable unfavorable and pur-chase intention. "If you were going to send a card to a friend, how likelyis it that you would buy this card?" would buy itwould not buy it.

For review and evaluation of alternative methods of measuring emotional responses toadvertising, see Aaker. Stayman, and Hagerty (1986).

I22 " SUMMER 1990

MUSIC INTLUENCES

MOOD MONITOR

We want to know how happy or sad you feel as you watch the simulatedgreeting card commercials. Please record how happy or sad you feel, between"sad" on the left to "happy" on the right.

Please move your pencil down the paper as you watch the commercials,moving to the left and right to indicate the sadness or happiness you feel at anygiven moment.*

There are no right or wrong answers. Just chart how you feel as hestyou can.

Subjects are also shown a sample mood monitor being drawn. They are told to try to keepa constant rate of speed down the page and no look down at Ihe paper too often whUewatching the commercials.

10

15

20

25

30

35

40

Sad

Moderately

Sad Neutral

Moderately

Happy Happy

FIGURE 1Mood Monitor and Sample.

PSYCHOLOGY & MARKETING/VOL 7, NO 2 123

MUSIC INFLUENCES

RESULTS

First, an SPSS partial factorial ANOVA, with repetitions within sub-jects, was run to test for differences across groups (the rows in Table 2),as well as cards, and music types, for each dependent variable (Nie andHull, 1981). None of the variables had different mean responses across thethree different groups, indicating that all groups responded similarly to thesame cards and music conditions. This is useful for two reasons: (1) itenables the analysis of main effects for music and cards in MANOVA andsingle ANOVAs, which would have been confounded by between-groupdifferences, if present, and (2) it enables a test for interactions betweenmusic and card effects (Edwards 1972; Hayes 1985). Normally partial fac-torial designs assume no interactions. However, the presence of repeatedmeasures per subject (three cards with three varied musical conditions),and absence of differences due to the "groups" blocking variable makespossible the use of two-way ANOVA to test for interaction between themusic and card treatments.

The repeated measures MANOVAs showed that musical backgroundand greeting card variations had approximately equal influence on theoverall profiles of responses. Wilks's Lambda for music was 0.847 {p =0.031) and 0.850 {p = 0.033) for cards. The music x card interaction wasinsignificant.

Figures 2 and 3 show patterns of means for the music and card effectson the two dependent variables significantly affected by music. Music hada significant effect on the subjects' moods (F = 6.07, p < 0.01), as didthe greeting cards (F = 3.13, p < 0.05). Happy music generally producedhigher mood monitor "happiness" scores, followed by no music, then sadmusic, and did so for two of the three cards. This pattern confirms thepredictions of the analyses of the pieces' musical structure. For the thirdcard, happy music evoked happier feelings than did sad music, althoughno music was the lowest in evoked subject mood. For the multiple com-parison tests, happy music produced an average subject mood that washigher than either of the other two music conditions. Thus, Hj was con-firmed.

This pattern was different for perceptions of the card's "mood" perse. The different cards were seen as differently "happy" across the entiresample, controlling for music {F = 22.34, p < 0.01). This may be takenas a manipulation check, as the card pretested as "happiest" was highestin the experiment as well, followed by the cards pretested as "neutral,"and "sad," in the same order as the pretest results. On the other hand,controlling for cards, music did not produce significant variations in per-ceptions of card mood (H^). Even though the means for perceived card"moods" induced by happy music were higher than those for sad musicfor all three cards, the significance level was approximately p = 0.25.

124 SUMMER 1990

MUSIC INFLUENCES

EmotionIndex

65

60

55

50

A5

40

Sad

FIGURE 2Subject Mood.

Happy

CARDS

Neutral

Happy Music

No Music

S.id Music

Overall impressions (liking) of the cards did not vary signifieantlyafter the treatments. The musical background also had no significant impactof this measure of card attitude (H,). Again, means were in the rightdirection, albeit not at conventional significance levels. Card impressionwith sad music was not significantly higher than with happy music {p =0.34) and no music (p = 0.11). This was the one significant interactioneffect found for music x cards {F = 3.08, p = 0.05), which inhibits theability to interpret the significance of main effects for this measure. Sadmusic was more effective than happy music for the neutral card, and slightlybetter with the sad card, but did not show any difference in card liking forthe happy card. This interaction should be interpreted with caution, giventhat only one of four interactions were significant, and the MANOVAresult for music x cards was insignificant (hence this interaction could bespecious).

A clearer pattern emerged for purchase intention. Cards did not differoverall in purchase intent {F = 1.26, ns). However, the music backgrounddid make a difference (F = 3.55, p < 0.05). Further, the multiple com-parison results showed the cards appearing with sad music were significantlymore likely to be selected than those with happy music, while happy and

PSYCHOLOGY & MARKETING VOL 7 NO 2 iii"

MUSIC INFLUENCES

3.6

Purchase 3.2Intention

2.8

2.4

2.0

SM

. . K M

HM

Sad Happy Neutral

CARDS

Happy Music

— . — . — .— No Music

Sad Music

FIGURE 3Purchase Intention.

no music grouped together, and the superiority of sad versus no music wasmarginally significant {p = 0.06). Thus, H4 was confirmed.

DISCUSSION

The results of this study are congenial to the hypothesis that variationsin the formal music structure of background music in commercials mayhave significant influence over the emotional responses of an audience.Prior research in consumer behavior had shown that varying specific back-ground music selections along dimensions of familiarity and liking couldaffect responses to "advertised" products. The present paper extends thediscussion to begin to examine what it is about the musical content thatmay lead to emotional and affective responses among consumers.

Different profiles of musical structural elements of modality, tempo,dynamics and rhythm, may, all things being equal, lead to a perception ofhappy or sad musical content. In this study, equally liked musical back-grounds that differed in their profile of these structural elements wereshown to affect audience moods in directions predictable from analysis of

126 SUMMER 1990

MUSIC INFLUENCES

the musical structure, across a set of simulated greeting card advertise-ments. This finding has direct relevance to those interested in the impacton mood from factors such as the structural elements in background music.It was noted earlier that simultaneous variation of the entire profile ofelements precludes inferences from this study regarding their relative in-fluence on moods and other dependent variables of interest. However,other research suggests the dominance of major versus minor melodies,all else equal. More important, having found that musical structure doesmake a difference to moods and behavioral intentions towards products"shaded" with music, it may be appropriate to extend the present workwith carefully controlled manipulations of specific structural elements ofmusic. To this end, the methodologies employed by Holbrook, ei al. (1981;1988) may be productively used.

The effects of varying background factors such as musical structurewere less clearly demonstrated for subjects' perceptions of the greetingcards' moods and their stated liking for the cards. Evidence supportive ofmusic's conditioning effect on card mood was not strong; nor did it dem-onstrate significant changes in liking for cards presented with differentbackground music. Although significant conditioning effects have beenpreviously reported for single exposures (cf. Stuart, Shimp, and Engle,1987; Gorn, 1982), advocates of classical conditioning might criticize theuse of a single exposure to the messages and lack of reinforcement. * How-ever, evidence of mood-induced influence, whether through conditioningor not, is demonstrated by the effect on purchase intent.t That this mayoccur in the absence of significant intervening effects on the perceivedsadness and even stated liking for a card may be supportive of peripheralpath processing in this setting. Given that the advertisements presentedno factual claims, motivation to process information via the central routemay have been diminished. The presence of music that evokes emotionsand other "noninformational" aspects of the ad may also stimulate pe-ripheral processing. Accordingly, one might expect to find influence onbehavior (here proxied by behavioral intent) without the necessity of in-tervening attribute perceptual changes or even significant changes in liking.Although these findings should be considered tentative, given the study'slimitations, they were consistent with the views of those who would classify

Extensions to field settings would enable lesting repetition effects with lower risk of demandeffects than might occur in a laboratory setting. It is also possible that the limited reliabilityof single-variable scales for card mood and card impression may have lowered chances forsignificant music effects on these variables. Note, however, that significance was obtainedfor music's impact on the single-variable measure of purchase intention, and cards differedin their effects on card mood.

t Confirmation of the ANOVA results is indicated by the correlation between subject moodand purchase intention (r = -0.355; p < 0.05). Hence, the sadder the person's mood, themore likely to purehase a card for a friend.

PSYCHOLOGY & MARKETING/VOL 7. NO 2 127~

MUSIC INFLUENCES

reactions to greeting card advertisements as a low cognitive involvementsituation. If so, intervening cognitions might not be affected while behav-ioral intentions and perhaps behavior could be (Batra and Ray, 1983;Bettman, 1986; Krugman, 1965; Robertson, 1976; Zajonc, 1968).

Nevertheless, there is clearly a mixed pattern of results. Whetheradditional research will show significance for what are here merely direc-tional indications of music affecting card mood and card impression, orwhether there will be a continuation of the pattern of stronger impact onbehavioral intent than intervening cognitions is unresolved at this point.Further research involving more measures of cognitive and affective re-sponses to music-backed advertisements, and their interactions (Allen andMadden, 1985), will be helpful in explicating the relative roles of peripheraland central route processing in contexts such as these.

Such inquiry will also be useful in further developing the managerialimplications of these concepts. As a laboratory study, the empirical resultshere presented may not easily be extended to an external consumptionenvironment. Aside from the obvious questions regarding the responsesof other samples to these and other greeting cards in the field, it is notclear whether music's ability to affect moods and shortrun behavioral in-tentions would last until actual purchase situations. The main purpose ofthe present study has been to investigate the potential utility of structuralanalysis and the relationships between hypothetical influences on audiencemood and responses to measures of these responses. Having shown po-tential application of the concepts expressed, it would be useful to extendthis work by testing music's effects on moods and intentions for otherproducts, including those for which decisions might be made in the presenceof music. Milliman's work (1982; 1986) in store and restaurant environ-ments might provide a useful example for such inquiry. One could alsonote that situations similar to the present study's context are found intelevision presentations of products to be ordered by phone. Here theinfluences on mood and intentions might be sufficient to produce buying.In other contexts repetition of similar advertisements might produce pur-chase intentions of longer duration, if not toward a specific card (or prod-uct), prehaps towards a brand such as Hallmark. Seeing the advertisedproducts in a store could evoke the mood originally stimulated by musicin the advertising situation. Although such an outcome is plausible and issupported by the literature (Lutz and Lutz, 1978), further research isneeded to test the efficacy of this linking of feelings.

It was also interesting to note that sad music was more effective ininfiuencing purchase intent than were happy music and silence. As notedearlier, research cited by Gardner (1985) has generally found positive cor-relations between mood inducers, moods and a number of dependent vari-ables such as evaluations. However, studies such as Caildini and Kenrick(1976) found that older children were more generous when self-generated

128 SUMMER 1990

MUSIC INFLUENCES

thought made them sad. As Gardner (1985), and Park and Young (1986)have stated, a key factor is the congruence between associated feelings andbehaviors consistent with that advocated in a message. In this situation,college students may have responded more positively to sad emotionalevocations (induced by music) in the context of sending greeting cards todistant friends (with messages like "missing you"). Certainly Hallmarkpositions (verbally and nonverbally) many of their greeting cards to appealto this market segment and situation. So did AT»feT with their memorableads supported by the song, "Feelings." What the musical structure mayhave been able to do in this study was evoke a feeling of melancholy, whichmay have affectively linked the audience to responding positively to sendinggreeting cards which were associated with that feeling.

The importance of music as a stimulus to emotional response toadvertising has both theoretical and practical implications. While the abilityto influence mood seems likely, the degree of influence and the desirabledirection may vary in other situations. Visits to an amusement park, forexample, may be more effectively advertised with happy music than withsad.*

Happy and sad may well be multidimensional constructs. Differentgradations within these emotions may require different inducers and mayin turn produce different responses and behavior. For example, there maybe different kinds of sadness (or happiness), influenced by different factors,and may lead to different responses (relaxation after completion of a dif-ficult task, expressions of joy and the like). In addition, music has a hostof elements that may be influential, beyond the musical structure. Theseinclude the words, artistic interpretation, specific memories that may beassociated with the selection, type and period of music and the interactionof all of these with the product and use-situation stressed in the advertise-ment. Additional research may eventually be able to decompose overalleffects into elements of all of these components, taking into account theeffect of moderator variables such as the respondents' demographics, per-sonality and lifestyle, cognitive and affective involvement in the commu-nication setting and familiarity with the music. The tasks in pursuing theseissues are considerable, but it seems worthwhile to decompose factors suchas musical influence into theoretical elements and their combinations. Itis encouraging in this process of inquiry to find that predictions from musi-cal theory may be derived that show correspondence in the emotionalresponses of audiences. To the extent that this phenomenon might bevalidated in future experiments, it may be possible to provide better

Hence future research may be productiveiy directed at the interactions among music type,card type and situation. Happy music may "help" happy cards, when purchased for joyousoccasions; sad music may help sad cards more than happy ones, for sad situations, suchas funerals.

PSYCHOLOGY & MARKETING/VOL 7. NO 2 i29~

MUSIC INFLUENCES

explanation of this source of emotional response to commercials, as wellas screen potential advertisements for predicted influences.

It seems appropriate that there remain some unresolved issues at thisstage of the research into the role of background variables on affectiveresponses and consumer behavior. It has been shown that there may be aplausible link between the profile of music structural elements and audiencemood, as well as significant impact on behavioral intentions towards theproducts advertised with music as background. The precise process bywhich these messages are formed and processed as information is open tofurther debate. There are promising suggestions following from researchand conceptualizing in the literatures reviewed here. We encourage furtherefforts to explore complexities within the musical and nonmusical stlmuhin advertising, and their relationships to the responses of consumers.

REFERENCES

Aaker, D. A., Stayman, D. M., & Hagerty, M. R. (1986). Warmth in advertising; Mea-surement, Impact, and Sequence Effect. Journal of Consumer Research, /2(March),365-381.

Aaker. D. A., Stayman, D. M., & Vezina, R. (1988). Identifying feelings elicited by adver-tising. Psychology & Marketing, 5(Spring), 1-16.

Aderitian, D. (1972). Elation, depression, and helping behavior. Journal of Personality andSocial Psychology, 24(1), 91-101.

Allen, C. T., & Madden, T. J. (1985). A Closer Look At Classical Conditioning. Journal ofConsumer Research, /2(December). 301-315.

Batra, R., & Ray, M. L. (1983). Advertising situations; The implications of differentia!involvement and accompanying affect responses. In R. J. Harris (Ed.), InformationProcessing Research in Advertising, (pp. 127-151). Hillsdale, New Jersey: Eribaum.

Baumann. D. J., Cialdini, R. B., & Kenrick, D. T. (1981). Altruism as hedonism; Helpingand self-gratification as equivalent responses. Journal of Personality and Social Psy-chology, 40(6). 1039-1046.

Beltman, J. R. (1986). Consumer Psychology. Annual Review of Psychology, 37, 257-289.Bierley. C . McSweeney, F. K , & Vannieuwkerk, R. (1985). Classical Conditioning of Pref-

erences for Stimuli. Journal of Consumer Research, /2(December). 316-323.Biller, J. D., Olson, P. J.. & Breen, T. (1974). The Effects of "Happy" Versus "Sad" Music

and Participation Anxiety. Journal of Music Therapy, /^(Summer), 69-73.Bower, G. H. (1981). Mood and Memory. American Psychologist, 36, 129-148.Bradley, I. L. (1971). Repetition as a factor in the development of musical preferences.

Journal of Research in Music Education, 19, 295-298.Calder, B. S., Phillips, L., & Tybout, A. (1981). Designing Research for Application. Journal

of Consumer Research, S(September), 197-207.Carson, T. & Adams, H. (1980). Activity valence as a function of mood change. Journal of

Abnormal Psychology, 89(3), 368-377.Cialdini, R., & Kenrick, D. (1976). Altruism as hedonism; A social development of negative

mood state and helping. Journal of Personality and Social Psychology, 34{5), 907-914.Clark, M., & Isen, A. (1982). Toward understanding the relationship between feeling states

and social behavior. In A. Hastorf & A. Isen, (Eds.). Cognitive Social Psychology(pp. 73-108). New York; Elsevier/North-Holland.

Dixon, W. J. (Ed.). (1981). BMDP, biomedical computer programs. Berkeley; University ofCalifornia Press.

Donnerstein, E., Donnerstein, M., & Munger, G. (1975). Helping Behavior as a Function

130 SUMMER 1990

MUSIC INFLUENCES

of Pictorially Induced Moods. The Journal of Social Psychology, 97(December), 221-

Edwards, A. L. (1972), Experimental design in psychological research. New York: Holt,Rinehart and Winston.

Galizio, M., & Hendrick, C. (1972). Effect of musical accompaniment on attitude: The guitar^^aprop ioT persuasion. Journal ofApplied Social Psychology. 2(Ocioher/DecembeT),

Gardner. M. P. (1985). Mood states and consumer behavior: A critical review. Journal ofConsumer Research, /i(December), 281-300.

Gardner, M. P., & Hill, R. P (1988). Consumers' mood states: Antecedents and consequencesof experiential versus informational strategies for brand choice. Psycholoev & Mar-keting, 5(Summer), 169-182,

Gardner. M. P . & Vandersteel, M. (1984). The consumer s mood: An important situationalvariable. In T. Kinnear (Ed.). Advances in Consumer Research (Vol. 11, pp, 525-529). Provo, UT: Association for Consumer Research.

Goldberg. M. E., & Gorn. G. J. (1987). Happy and sad tv programs: How they affect reactionsto commercials. Journal of Consumer Research. /-/(December), 387-403.

Gorn, G. i. (1982). The effects of music in advertising on choice behavior: A classicalconditioning approach. Journal of Marketing, •*6(Winter). 94-101.

Griffitt, W. (1970), Environmental effects on inlerpersona! affective behavior: Ambient tem-perature and attraction. Journal of Personality and Social Psychology /5(3) 240-244

Griffitt, W.. & Guay. P (1969). "Object" Evaluation and Conditioned Affect. Journal ofExperimental Research in Personality. 4(iu\y). 1-8.

Hayes, W. B. (1985). Personal communication, (October).Hevncr. K. (1935). The affective character of the major and minor modes in music American

Journal of Psychology, 47, 103-118.Hevner, K. (1936). Experimental studies in the elements of expression in music American

Journal of Psychology, 48, 246-268.Holbrook. M. B. (1981). Integrating compositional and decompositional analyses to represent

the intervening role of perceptions in evaluative judgements. Journal of MarketineResearch, 7S(February), 13-28.

Holbrook, M. B.. & Anand, P. (1988). Aims, concepts, and methods in marketing researchon consumer esthetics: The effects of tempo on perceptual and affective responses tomusic. Working paper (March).

Holbrook. M. B . & Corfman. K. P (1985). Ouality and value in the consumption experience-Phaedrus rides again. In J. Jacoby and J. C Olson (Eds) . Perceived Quality~HowConsumers View Stores and Merchandise. Lexington. Massachusetts: Lexington Books

Infante. D, A., & Berg, C. M. (1979). The impact of music modahty on the perception ofcommunication situations in video sequences. Communication Monographs, 46{!i\iTit),

Isen, A. M. (1984). The influence of positive affect on decision making and cognitive or-ganization. In T. C. Kinnear ( E d ) , Advances in Consumer Research (Vol. 11. pp.534-537). Provo. UT: Association for Consumer Research

Isen. A. M., Shalker, T E., Clark. M. S.. & Karp, L. (1978). Affect accessibility of materialin memory, and behavior: A cognitive loop? Journal of Personality and Social Psy-chology, ,?6(January). 1-12.

Johnson, E, J.. & Tversky, A. (1983). Affect, generalization, and the perception of riskJournal of Personality and Social Psychology. 45(July). 20-31

Kellaris. J. J.. & Cox. A. D. (1987), The effects of background music in advertising' Arephcation and extension. In S. P Douglas, M. I. Alpert, J. C. Anderson P DoyleG. T Ford. G. L. Frazier, V. Mahajan. W. M, Pride & M. R. Solomon (Eds ) AMAEducators' Proceedings (Vol. 53, p. 283). Chicago: American Marketing Association.

Kotler. P (1974). Atmospherics as a marketing tool. The Journal of Reiailmg, -/Vl Winter).48-64.

Kroeber-Riel, W. (1984). Emotional product differentiation by classical conditioning (withconsequences for the low-involvement hierarchy ). In T. C. Kinnear (Ed.), Advances

PSYCHOLOGY & MARKETING/VOL 7, NO. 2 131

MUSIC INFLUENCES

in Consumer Research (Vol. 11, pp. 538-543). Provo, UT: Association for ConsumerResearch.

Krugman, H. E. (1965). The impact of television advertising: Learning without involvement.Public Opinion Quarterly. 29, 349-356.

Lazarus, R. S. (1982). Thoughts on the relations between emotion and cognition. AmericanPsychologist. 37, 1019-1024.

Lazarus, R. S. (1984). On the primacy of cognition. American Psychologist, J9(February),124-129.

Lutz, K., & Lutz, R. J. (1978). Imagery-eliciting strategies: Review and implications forresearch. In H. K. Hunt (Ed) , Advances in Consumer Research (Vol. 5, pp. 611-620). Chicago: Association for Consumer Research.

McSweeney, F. K., & Bierley, C. (1984). Recent developments in classical conditioning.Journal of Consumer Research, //(September), 619-631.

Meyer, L. B. (1956). Emotion and meaning in music. Chicago: The University of ChicagoPress.

Milliman, R. E. (1982). Using background music to affect the behavior of supermarketshoppers. Journal of Marketing, 46(Summer), 86-91.

Miiliman, R. E. (1986). The influence of background music on the behavior of restaurantpatrons. Journal of Consumer Research, /^(September), 286-289.

Nie, N. H., & Hull, C. H. (1981). SPSS Update 7-9. New York: McGraw-Hill.Paivio, A. (1971). Imagery and verbal processes. New York: Holt, Rinehart and Winston.Park, C. W., & Young, S. M. (1986). Consumer response to television commercials: The

impact of involvement and background music on brand attitude formation. Journal ofMarketing Research, 2i(February), 11-24.

Petty, R. E., Cacioppo, J. T., & Schumann, D. (1983). Central and peripheral routes toadvertising effectiveness: The moderating role of involvement. Journal of ConsumerResearch, /O(September), 135-146.

Robertson, T. S. (1976). Low-commitment consumer behavior. Journal of Advertising Re-search, 76(April), 19-24.

Rohner, S. J., & Miller, R. (1980). Degrees of familiar and affective music and their effectsof state anxiety. Journal of Music Therapy, /7(Spring), 2-15.

Shimp, T. A. (1981). Attitude toward the ad as a mediator of consumer brand choice. Journalof Advertising, /0(2), 9-15.

Smith. P. C & Curnow. R. (1966). Arousal hypotheses and the effects of music on purchasingbehavior. Journal of Applied Psychology, 50(3), 255-256.

Smith, R. E., & Hunt, S. D. (1978), Attributional processes and effects in promotionalsituations. Journal of Consumer Research, ^(December), 149-158.

Staats, C. K , & Staats, A. W. (1957). Meaning established by classical conditioning. Journalof Experimental Psychology. 54(July). 37-40.

Stayman, D. M., & Hagerty, M. R. (1985). Methodological issues in simulated shoppingexperiments. In E. Hirschman & M. Holbrook (Eds.). Advances in Consumer Research(Vol. 12, pp. 173-176). Ann Arbor: Association for Consumer Research.

Stuart, E. W., Shimp, T. A., & Engle, R. W. (1987). Classical conditioning of consumerattitudes: Four experiments in an advertising context. Journal of Consumer Research./^(December), 334-349.

Tsal, Y. (1985). On the relationship between cognitive and affective processes: A critique ofZajonc and Markus. Journal of Consumer Research. /2(December), 358-362.

Watson, J. B., & Rayner, R. (1920). Conditioned emotional reactions. Journal of Experi-mental Psychology, i(Febniary), 1-14.

Weiner, R. B. (1979). The cognition-emotion process in an achievement-related context.Journal of Personality and Social Psychology, 37. 1211-1220.

Zajonc, R. B. (1968). Attitudinal effects of mere exposure. Journal of Personality and SocialPsychology, 25(June), 1-27.

Zajonc. R. B. (1980). Feeling and thinking: Preferences need no inferences. American Psy-chologist, 35. 151-175.

Zajonc, R. B., & Markus. H. (1982). Affective and Cognitive Factors in Preferences. Journalof Consumer Research, 9(2), 123-131.

132 SUMMER 1990

MUSIC INFLUENCES

Zajonc. R. B.. Markus. H., & Wilson, W. R. (1974). Exposure effects and associativelearning. Journal of Experimental Social Psychology, 7f>(May), 248-263.

Requests for reprints should be addressed to Mark I. Alpert. Department of Mar-keting. The University of Texas at Austin, Austin. TX 78712-1176.Judy I. Alpert is Adjunct Professor ot Marketing at the Division of BusinessAdministration. St. Edwards University, Austin. Texas.Mark I. Alpert is Foley's-Federated. Inc.. Centennial Professor in Retailing, De-partment of Marketing. The University of Texas at Austin.The authors gratefully acknowledge the computational assistance of V. Kumar, thehelpful suggestions of Pam Henderson and Doug Stayman, and the cooperation ofthe students and faculty at St. Edwards University.

PSYCHOLOGY & MARKETING/VOL 7. NO 2 133