Music-induced emotion effects on decision-making

Abstract

This qualitative systematic review aimed to examine music-induced emotion effects on decision-making. Empirical articles published from 2006 to 2016 were searched for in PubMed and PsycInfo. Through the main search, 634 records were identified and 15 articles were included and analyzed according to the following categories: aims of the study, participants, study design, type of music, type of emotion, decision-making tasks, and study results. The included studies aimed to investigate the effects of music on prosociality and aggression, risk-taking, and other decision-making processes. All studies used experimental designs and most of them used pre-selected music and stimulation through music listening. Different decision-making tasks were used, such as a gambling task, helping behavior tasks, and even assessment of aggressive behavior through adding hot chili sauce to food. Results showed that music is a powerful and engaging stimulus that influences decision-making processes and risk-taking, promotes prosociality, and affects customers’ behavioral choices. Different underlying processes in the interaction between music, mood, and decision-making are discussed. This review contributes to promoting research on the use of background music as an arousal and mood inducer in different contexts, as well as music-based interventions to foster prosociality and decrease aggressive behaviors.

Keywords

decision-making emotion mood music systematic review

Decision-making can be defined as “the process of choosing a preferred option or course of action from among a set of alternatives” (Kitajima & Toyota, 2013; Shafir, 1999, p. 220). Throughout history, theories have attempted to explain decision-making in terms of a strict, normative mode of rationality (Loewenstein & Lerner, 2003). Such theories usually implied a decision-maker capable of boundless rationality and ideal sensitivity to information. Tversky and Kahneman (1974), however, have demonstrated effects of variations in information presentation on judgment and decision-making. Since then, researchers have investigated other possible variables interfering in decision-making. These newer models stress the importance of describing how real decision-making actors behave and which factors influence their decisions. Hypotheses involving variables such as emotions (Zheng, Yang, Jin, Qi, & Liu, 2017), social or moral implications (Greene, Nystrom, Engell, Darley, & Cohen, 2004; Zinchenko & Arsalidou, 2018), and pecuniary motivation (Brockmeyer, Simon, Becker, & Friederich, 2017) have been thoroughly explored in the last decades.

Current theories agree that human decisions cannot be explained only by rational imperatives and that decision-making is fundamentally supported by emotions (Frank, Cohen, & Sanfey, 2009; Loewenstein, 1996; Phelps, 2009; Sokol-Hessner, Camerer, & Phelps, 2013). Despite the healthy diversity of theories (Damasio, 1994, 1996; Damasio, Tranel, & Damasio, 1991; Hansen, 2005; Hansen & Christensen, 2007; Loewenstein, Weber, Hsee, & Welch, 2001), all focus on understanding how, when, where, and to what extent emotion is related to decision-making. Hence, current experimental studies control the emotion of participants when investigating a decision-making process.

According to Russell’s (1980) circumplex model, emotions vary in a two-dimensional space, with one dimension corresponding to arousal (or activation) and the other to mood (or valence). Arousal refers to the physical and psychological activation of an emotion, while valence is related to a pleasure–displeasure continuum (Posner, Russell, & Peterson, 2005). According to this model, specific emotions “arise out of patterns of activation within these two neurophysiological systems, together with cognitive interpretations and labeling of these core physiological experiences” (Posner et al., 2005, p. 716). However, it is important not to reduce all affective states to these two dimensions, since “individuals do not experience, or recognize, emotions as isolated, discrete entities, but they rather recognize emotions as ambiguous and overlapping experiences” (Posner et al., 2005, p. 719).

From the many different variables that may influence emotional states, music has shown robust effects (Coan & Allen, 2007), because it can influence the research setting even without the conscious attention of participants to the stimulus, thus representing real-life situations (Halko, Mäkelä, Nummenmaa, Hlushchuk, & Schürmann, 2015). Moreover, music is ubiquitous, since musicality is a universal and cross-cultural characteristic of humankind, deeply connected to its social and cultural context (Trehub, Becker, & Morley, 2015). Musical abilities develop very early in infants’ development and are very sophisticated even in non-musicians. Music listening and music making involve complex cognitive processes that are shared with the processing of language, suggesting that musicality is a natural skill of the human brain (Koelsch, 2011).

Neuroimaging evidence has shown that music activates a network comprising amygdala, hippocampus, parahippocampal gyrus, hypothalamus, insula, cingulate cortex, orbitofrontal cortex, and temporal poles, which is involved in the modulation of emotions (Koelsch, 2009, 2014). Moreover, listening to pleasurable music activates the nucleus accumbens and the dopaminergic reward system (Blood & Zatorre, 2001; Koelsch, 2009, 2014). It is impossible to find an inter-cultural definition of music (Trehub et al., 2015). Thus, for this review we opted for a broad definition of music as temporally structured human activities, social and individual, in the production and perception of sound organized in patterns that convey non-linguistic meaning (Merriam-Webster, n.d.).

Emotional responses to music have been studied through subjective, behavioral, and physiological measures, corresponding to different brain networks and different levels of cognitive processes (Koelsch, 2015). At the level of the brain stem, music can modulate the arousal component of emotion, through changes in physiological responses, and elicit motor reactions, such as a smile. Such responses are related to automatic and noncognitive processes. At the level of the neocortex, music is responsible for more cognitive and conscious processes, involving subjective feelings, as well as voluntary actions and behaviors, such as singing or foot tapping (Koelsch, 2014, 2015).

Besides the properties of music, factors such as musical preferences and competences, personality traits, the stimulus selection, participants’ involvement in the music, the social or solitary setting, as well as the quality of the performance have been shown to make important contributions to emotional reactions to music (Dunbar, Kaskatis, MacDonald, & Barra, 2012; Liljeström, Juslin, & Västfjäll, 2012; Koelsch, 2015). Music-based interventions and musical training have also been shown to improve cognition, language, attention, memory, and executive functions (Fusar-Poli, Bieleninik, Brondino, Chen, & Gold, 2017; Koelsch, 2009, 2011; Sachs, Kaplan, Der Sarkissian, & Habibi, 2017; Zumbansen, Peretz, & Hébert, 2014). Nevertheless, the effects of music in the decision-making process have been less investigated. Additionally, we did not find previous reviews about the music effects on decision-making.

Therefore, this qualitative systematic review aimed to examine music-induced emotion effects on decision-making. In particular, we wanted to investigate how music elicits the arousal and mood dimensions of emotion, thereby influencing the decision-making process. Qualitative systematic reviews allow a comprehensive understanding of the process of interventions and studies, rather than focus on their effectiveness (Stern, Jordan, & McArthur, 2014), and they generally organize data into categories and themes (Butler, Hall, & Copnell, 2016). We analyzed the included articles according to the following categories: aims of the study, participants, study design, type of music, type of emotion, decision-making tasks, and study results.

Method

In undertaking this review, we followed the criteria of the PRISMA checklist (Moher, Liberati, Tetzlaff, Altman, & The PRISMA Group, 2009; see Figure 1 for a summary). Initially, a preliminary search was performed in the electronic databases PubMed and PsycInfo in order to identify the more appropriate keywords. The first two authors independently screened PubMed and PsycInfo in September 2016 using the following terms: “music” AND “decision making”; “music” AND “decision making” AND “mood”; “music” AND “decision making” AND “emotion”; “sound” AND “decision making” AND “emotion”; and, “sound” AND “decision making” AND “mood.” Searching was restricted to articles published in the last ten years (2006–2016). All relevant records were extracted to Zotero 4.0 reference management software (Roy Rosenzweig Center for History and New Media). Studies were included if they met the following criteria: empirical articles published in English and involving human participants; music considered as a temporally structured human activity, involving and integrating rhythm, melody, harmony, and lyrics (Merriam-Webster, n.d.); emotion considered both as the arousal and the mood dimension (Russell, 1980); decision-making considered as “the process of choosing a preferred option or course of action from among a set of alternatives” (Shafir, 1999). Every included article had its references screened. Data searching, screening and selection were undertaken by two independent reviewers. Disagreements were resolved through discussion with the third author. All included articles were entirely read and examined focusing on their main characteristics and the similarities and differences between them.

Figure 1.

Flow of studies through the review process.

Results

From the database searching, 634 records were identified. After removing duplicates, we obtained 508 articles and screened their titles and abstracts. Theses, dissertations, book chapters, articles in languages other than English, and irrelevant records were excluded. According to the eligibility criteria abovementioned, eight studies were included. Subsequently, all the references of these eight articles were screened and eight other articles were included, identifying a total of 16 articles. After the full-text analysis, one article was excluded for being a duplicate study of another article. In the end, 15 articles were included in this review and analyzed according to the following categories: aims of the study, participants, study design, type of music, type of emotion, decision-making tasks, and study results. Table 1 presents these categories.

Table 1.

Included studies characterized according to the analyzed categories.

References	Aims of the study	Participants	Study design	Type of music	Type of emotion	Decision-making tasks	Study results
Brooks & Schweitzer (2011, Study1)	Explore the influence of anxiety on negotiation processes and outcomes.	Students (N = 136) from a Northeastern university (US), (81 females, 55 males)	Experimental design, randomization, two between-subject conditions	Anxiety condition: theme from the movie Psycho. Neutral condition: Handel’s “Air” from Water Music. Both of these audio clips are orchestral compositions with no vocal parts. Length of 3 minutes, played on a continuous loop.	Anxiety vs. Neutral	A negotiation of price, warranty period, and service period of a cell-phone shipment, where buyers and sellers were participants.	Anxious participants expected to earn less than non-anxious participants. Anxious participants earned less profit than non-anxious participants, as well as participants who were paired with an anxious counterpart earned significantly more profit than participants who were paired with a non-anxious counterpart.
Chung, Lee, Jung, & Kim (2016)	Investigate how mood influences socioeconomic decision-making.	Forty healthy volunteers (22 females, M = 21.55, SD = 1.84 years; 18 males, M = 21.55, SD = 3.89 years)	Experimental design, randomization, between-subject	Musical excerpts identified from a previous study (Roy et al., 2008). Pleasant excerpts from: “Love and Happiness” by Ernst Ranglin, “William Tell Overture” by Rossini, and “French Cancan” by Canissimo. Unpleasant excerpts from: “Pendulum Music” by Steve Reich, “The Threshold of Deafening Silence” by Paul Dolden, and “Fascicles” by the Thirteen Ghosts with Derek Bailey and Thurston Moore.	Pleasant vs. Unpleasant	Ultimatum game	Pleasant group preferred the music more than the unpleasant group. Unpleasant group reported increased negative affect after mood induction and pleasant group reported marginally decreased negative affect after mood induction. Significant interaction between time and group for positive affect scores. The less favorable the offers were, the more participants rejected them. Significant interaction between offer and group: unpleasant group rejected unfair offers more often than the pleasant group. Rejection rates of unfair offers were negatively correlated with: felt fairness and felt happiness in unpleasant group; and felt happiness and rejection in the pleasant group.
Day, Lin, Huang, & Chuang (2009)	Examine the separate and joint effects of tempo of background music and task difficulty on the performance of multi-attribute decisions.	Forty college students at the Southern Taiwan University of Technology (Taiwan), 11 males and 29 females (19–26 years)	Experimental design, randomization, between-subject	“Through the Arbor”, by Kevin Kern, 1996. New Age style. Instrumental, using piano. Two manipulations of background music: faster background music was made by speeding up the original piece by 25% and slower background music was made by slowing down the original piece by 25%.	Arousal inducer vs. distractor	Alternative-attribute matrix with weighted additive rule (WADD) and elimination-by-aspects method (EBA) as decision strategies	Significant differences for accuracy and search pattern were found. Faster tempo condition presented higher accuracy and lower search pattern. Interaction between music tempo and task difficulty for search pattern, faster tempo presented lower search pattern. At EBA reacquisition rate was significantly lower with the presence of the faster tempo than the lower tempo.
Fischer & Greitemeyer (2006, Studies 1 and 3)	Impact of misogynous and man-hating song lyrics on aggression related thoughts, emotions, and behavior toward the same and the opposite sex. Study 1: misogynous song lyrics increase aggressive responses of men toward women. Study 3: Aggressive cognition, attitudes toward women and men, positive and negative emotions and aggressive behavior toward female and male after misogynous, man-hating, or neutral songs.	Study 1: 161 students (88 male, 73 female) from the Ludwig- Maximilians-University, Munich (Germany). Ages 19–60 years (M = 27.13, SD = 7.97). Study 3: 107 (56 female, 51 male) ages 18–49 years (M = 24.11, SD = 5.17)	Study 1: experimental design, 2 (music genre: misogynous vs. neutral) × 2 (sex of participants) × 2 (sex of confederate) factorial design. Between-subject design. Study 3: experimental design, 3 (music genre: misogynous vs. man-hating vs. neutral) × 2 (sex of participant) between-subject design	Study 1: four songs with misogynous content and four songs without misogynous content. Comparable music genres and same length. A song with English lyrics and also to one with German lyrics.	Aggressive behavior. Controlled for arousal and positive feelings elicited by misogynous, man-hating and neutral songs.	Study 1: add hot chili sauce to a sandwich prepared for either a female or male confederate. Study 3: aggression behavior task: assign the time for two subsequent participants to keep their hand in ice water.	Study 3 Aggressive behavior: significant three-way interaction. Men who listened to misogynous song lyrics assigned longer times of ice water treatment to the female target person than (a) women who listened to misogynous music, and (b) men who listened to neutral and man-hating song lyrics. Women who listened to man-hating music assigned longer times of ice water treatment to men than did women who listened to neutral and misogynous music, but did not differ from men who listened to man-hating music. Men who listened to man-hating music responded more aggressively toward the male target person than did women who listened to neutral music.
Greitemeyer (2009a, Study 3)	Examine whether listening to prosocial songs affects prosocial action.	Ninety students (55 female, 35 male) of the Ludwig-Maximilians University, Munich (Germany)	Experimental	One song with English lyrics and one song with German lyrics. Prosocial condition songs: “Love Generation” (Bob Sinclair) and “Kommt zusammen” (2raumwohnung). Neutral condition songs: “Rock this Party” (Bob Sinclair) and “Lachen und Weinen” (2raumwohnung). Artists and genres were matched.	Prosocial behavior	Donation of the €2 earned from the researcher for a non-profit organization.	Participants who had listened to prosocial songs were more likely to donate money than participants who had listened to the neutral songs. No significant effect of participant sex on donating.
Greitemeyer (2009b, Study 1, 2, 3 and 4)	Study 1: investigate if exposure to prosocial songs, compared to neutral songs, increases prosocial behavior. Studies 2–4: clarify the underlying mechanism. Studies 2 and 4: test the possibility that listening to prosocial, relative to neutral, songs increases the accessibility of prosocial thoughts, which then promotes prosocial behavior. Studies 3 and 4: test the possible mediating effects of empathy.	Study 1: 33 students (16 female, 17 male) of the Ludwig-Maximilians University, Munich (Germany). Study 2: 50 students at the University of Sussex, Brighton (UK). Study 3: 59 students (48 female, 11 male) at the University of Sussex, Brighton (UK). Study 4: 50 students (31 female, 19 male) at the University of Sussex, Brighton (UK).	Experimental design, randomization	Study 1: 4 songs. Prosocial condition songs: “Heal the World” (Michael Jackson), “Ein bißchen Frieden” (Nicole), “We Are the World” (Liveaid), and “Help” (Beatles). Neutral condition songs: “On the Line” (Michael Jackson), “Spiel um deine Seele” (Peter Maffay), “An Englishman in New York” (Sting), and “Octopus’s Garden” (Beatles). Studies 2, 3 and 4: Prosocial condition: “Love Generation” (Bob Sinclair) and “Feed the World” (U2 with Band Aid). Neutral condition: “Rock This Party” (Bob Sinclair) and “Vertigo” (U2).	Study 1: control for mood with the Positive and Negative Affect Schedule (Watson, Clark, & Tellegen, 1988)	Study 1: Prosocial behavior: help the experimenter to pick up 20 pencils that he had “accidentally” knocked off the table and on the floor. Study 2: Participants’ willingness to take part (yes or no) in further experiments and how much time (in minutes) they would give. Study 3: Imagine that a person died in an accident leaving his or her two children without a home. Participants’ willingness of help and the amount of help. Study 4: economic decision-making task (the dictator game decide how much money to give to a receiver.	Study 1: Participants who had listened to the prosocial songs picked up more pencils than did those who had listened to the neutral songs. Even when controlling for mood, there was a significant effect of type of song on prosocial behavior. Study 2: Listening to prosocial, relative to neutral, songs promoted prosocial behavior: participants who had listened to the prosocial songs showed more willingness to assist in further studies and for more time than participants who had listened to the neutral songs. Study 3: Participants who had listened to the prosocial songs reported a greater willingness to help than did participants who had listened to the neutral songs. Helping was significantly associated with empathy. Study 4: Exposure to prosocial, relative to neutral, songs increased cooperative behavior in the dictator game: participants who listened to prosocial songs left more 10-pence coins than did participants who had listened to the neutral songs. Significant association of cooperative behavior in the dictator game with empathy, and marginally significant association with prosocial thought.
Greitemeyer (2011, Study 5)	Investigate whether exposure to prosocial (relative to neutral) music would decrease aggressive cognition, affect, and behavior. Clarify the causal mechanisms by which exposure to prosocial music decreases aggressive behavior.	Fifty adults (24 women and 26 men) of a community sample in Brighton (UK). Age M = 29.9 years, SD = 11.0	Experimental design	Prosocial condition songs: “Love generation” (Bob Sinclair) and “Feed the world” (U2 with Band Aid). Neutral condition songs: “Rock this Party” (Bob Sinclair) and “Vertigo” (U2). Songs matched on mood and arousal dimensions. Liking and perceived aggressive content of the songs were relatively similar.	State Hostility Scale (Anderson, Deuser, & DeNeve, 1995), Revised Attitudes Toward Violence Scale (RATVS; Anderson, Benjamin, Wood, & Bonacci, 2006), Perceived Arousal Scale (Anderson et al., 1995)	Physical aggression toward a target person (administering hot chili sauce) who had insulted the participant.	Participants in the prosocial music condition administered less hot chili sauce than did participants in the neutral music condition. Significant correlation just between administered chili sauce and state hostility. Significant direct effect of music condition on administered chili sauce. When controlling for state hostility, the effect reduced to non-significance.
Halko & Kaustia (2012)	Test for preference interactions in risk-taking.	Twenty-five teenagers, aged 12–17 years, Helsinki (Finland)	Experimental design, within-subject	Eight self-selected pieces for every participant. Four of what is considered favorite music and four disliked musical pieces.	Mood elevation	Gambling task	Favorite music alleviated and disliked music exacerbated loss aversion. Liked music has slightly increased acceptance rates among the most favorable gambles. Disliked music has lowered the acceptance rates in the least favorable gambles.
Halko et al. (2015)	Show that risky choices and neural coding of risk preferences are modulated by short-term changes in context of the environment.	Twenty-two healthy participants, M = 32, 8 females, range 22–50 years, 17 of 22 right-handed, no professional musicians	Experimental design, within-subject	Eight self-selected pieces for every participant, four of what is considered favorite music and four disliked musical pieces.	Mood valence and arousal	Gambling task	Mean valence was significantly higher and arousal significantly lower in liked music compared to disliked music. The frequency for accepting a gamble grows in this order: disliked music, no music, and liked music. The disliked music had no significant effect on gamble acceptance. Losses were weighted approximately twice as much as gains. Behavioral loss aversion alone was significantly reduced in the liked music. Music had a large effect on the behavior of participants with high emotional contagion scores. Functional MRI images showed that gambles in the disliked music condition with high expected reward induced higher activation than gambles with low expected reward. In the liked music condition the activation pattern was reversed.
Kirschner & Tomasello (2010)	Analyze if prior engagement of 4-year-old children in joint music making increases prosocial behavior.	Ninety-six 4-year-old children (48 males and 48 females, M = 4 years and 6 months, range = 4.0 to 5.0 years), from 16 different German urban day-care centers, mixed socioeconomic backgrounds	Between-participants	Synchronization and singing together with background music. Pre-recorded music with guitar chords, using the frogs as instruments in synchrony with the song’s lyrics. Quack stick to be scraped over the frogs’ back.		Helping test: one child had a sudden accident which presented the partner with the choice to help or to continue pursuing his or her own play activity.	Children of both genders helped one another more after joint music making. Even when children decided not to help or stay until the peer’s problem was solved, the children in the Musical condition more frequently offered verbal excuses than children in the Non-musical condition.
Kniffin, Yan, Wansink & Schulze (2016, Study 1 and 2)	Investigate the influence of music on cooperative behavior within decision-making groups	Study 1: 78 participants (27 female) Study 2: 188 participants (75 female)	Study 1 and 2: experimental design, between-subjects	Study 1: Happy Music Group—set of four “Happy” songs: “Yellow Submarine” by the Beatles; “Walking on Sunshine” by Katrina and the Waves; “Brown Eyed Girl” by Van Morrison; and, the theme song from “Happy Days”; Unhappy Music Group—set of “Unhappy” songs by less familiar “heavy metal” bands: “Smokahontas” by Attack Attack! and “ You Ain’ t No Family” by Iwrestledabearonce. Study 2: same plus Control Group—no music.	Higher mood vs. lower mood from the scale: good mood, cheerful, not very comfortable and edgy or irritable	Study 1 and 2: Voluntary Contribution Mechanism (VCM): participants receive a total of 10 symbolic tokens in each round and decide to allocate 0 to 10 tokens to either the group or private accounts.	Study 1: Participants listening to Happy music contributed more. Study 2: Participants who listened to Happy music showed more cooperative behavior than the Unhappy Music Group and the Control Group.
Moritz et al. (2009)	Explore motivational factors for jumping to conclusions (JTC).	Twenty-seven schizophrenia patients (14 females) from the University Medical Centre Hamburg Eppendorf and 32 healthy controls (21 females) recruited mostly by advertisement.	Experimental design, randomization into three conditions, differing only in the sequence of music. Between-subject factor: schizophrenia group and healthy control group. Within-subject factors: Mood (anxiety, positive, no music), Content of the Alternatives (neutral, positive, and negative content), and Validity (correct, incorrect).	Positive/happy music condition: instrumental version of the song “Take Five” (composed by Paul Desmond for the Dave Brubeck Quartet); Negative/anxious music condition: Michael Myers’ theme from the horror movie “Halloween” (composed by the director of the movie, John Carpenter).	Happy vs. anxious vs. none	Choose to reject the trial or to decide which title corresponds to the painting exposed.	Schizophrenia patients made more decisions than healthy subjects during the negative/anxious music condition. However, there was a significant effect for delusion, with a higher number of decisions made by schizophrenia patients with current delusions under the anxiety-induction condition compared to non-deluded patients and healthy control group.
Schulreich et al. (2014)	Investigate if music-evoked emotions modulate risk attitudes in the gain domain and in particular probability weighting.	Forty-one participants (28 females, 13 males), M = 27.37 years, SD = 7.832 years	Experimental, within-subject design consisting of four conditions (happy music, random tones, no music and sad music)	Happy music: instrumental excerpts from various epochs and styles (classical music, Irish jigs, jazz, reggae, South American, and Balkan music). Sad music: classical and indie-pop pieces, selected on the basis of features that have been shown to evoke sad feelings. Random tone sequences featured acoustically changing stimuli.	Happiness vs. sadness vs. neither happiness nor sadness	Lottery choice task	Music altered the emotional state of happiness in a different manner and this effect was diminished over time. While happy music was associated with greater happiness, sad music and random tone sequences were associated with lower happiness. The changes in participants’ choices were explained through the changes in how they convert objective probabilities into subjective decision weights.
Skulmowski, Bunge, Kaspar, & Pipa (2014)	Examine the interplay between intuitive and controlled cognitive processes related to moral and social judgments through a virtual reality (VR) setting.	Sixty-six students (36 female)	Experimental design, four conditions within-subject, and the fourth condition had a between-subject extra condition.	“Inner Mind Mystique I” by Yamazaki Maso	Negatively valenced emotions and emotional arousal	Modified version of the trolley dilemma using virtual reality: choose who to sacrifice by steering the trolley to the left or to the right.	Behavioral results: in the group condition both male and female participants decided to sacrifice the single person in order to save the group. Participants in the music group showed stronger negatively valenced mood and were more delighted than those participants in the control condition. Eye tracking results: irritating music affected participants’ arousal, while mixed self-reports indicated no specific valence.
Strick, de Bruin, de Ruiter, & Jonkers (2015, Study 1)	Study 1: test the prediction that moving music does not enhance cognitive persuasion variables. Investigate adverse responses and attitudes toward the ad, and expect these to converge with the positive affective response evoked by moving music.	Ninety-two students from the University of Utrecht (Netherlands;48 male, 44 female, age M = 20.92 years).	Experimental design, 2 (music: moving vs. nonmoving) × 2 (ad: positive vs. negative) between-participants design	Two moving excerpts: “Down by the Riverside” (positive music) and “The Funeral” (negative music); two non-moving excerpts: “Always Like This” (positive music) and “Attack el Robot” (negative music)	Music-induced transportation. Positive and negative affective response	Choose how much of the €4 received for participating they would be willing to donate to the respective campaign. Participants had to choose to forward the film clip to five friends in order to raise awareness for the issue.	Participants rated the ad as more moving when it featured moving than non-moving music, and rated the positive ad as more positive and less negative than the negative ad. Moving music increased willingness to donate money and willingness to forward the film clip to five friends. Moving music enhanced transportation and behavioral intentions, but did not increase cognitive reasoning variables.

Participants

Regarding participants, almost all of the included studies involved healthy people and just one of them investigated schizophrenia patients (Moritz et al., 2009). Most of the studies involved students (Brooks & Schweitzer, 2011; Day et al., 2009; Fischer & Greitemeyer, 2006; Greitemeyer, 2009a, 2009b; Skulmowski et al., 2014; Strick et al., 2015), while one of them investigated adolescents and teenagers (Halko & Kaustia, 2012) and just one of them included children (Kirschner & Tomasello, 2010). In general, the number of participants was quite small, with five studies having less than 40 participants (Chung et al., 2016; Day et al., 2009; Greitemeyer, 2009b, Study 1; Halko & Kaustia, 2012; Halko et al., 2015), and seven studies with 40 to 60 participants (Greitemeyer, 2009b, Study 2, 3 and 4; Greitemeyer, 2011, Study 5; Moritz et al., 2009; Schulreich et al., 2014). Five studies included 60 to 100 people (Greitemeyer, 2009a; Kirschner & Tomasello, 2010; Kniffin et al., 2016, Study 1;¹ Skulmowski et al., 2014; Strick et al., 2015, Study 1), while four of them involved more than 100 participants (Brooks & Schweitzer, 2011; Fischer & Greitemeyer, 2006, Study 1 and 3; Kniffin et al., 2016, Study 2).

Study design

All of the analyzed studies used an experimental design. In particular, some of them used between-subjects designs (Brooks & Schweitzer, 2011, Study 1; Chung et al., 2016; Fischer & Greitemeyer, 2006, Study 1 and 3; Kirschner & Tomasello, 2010; Kniffin et al., 2016, Study 1 and 2; Strick et al., 2015, Study 1). Four studies used within-subjects designs (Day et al., 2009; Halko & Kaustia, 2012; Halko et al., 2015; Schulreich et al., 2014), while two studies combined both between and within-subject factors (Moritz et al., 2009; Skulmowski et al., 2014).

Type of music

In terms of the type of music, we highlight that the vast majority of the studies used standardized songs for all the participants, pre-selected by experts or experimenters. Only the research of Halko and colleagues used liked and disliked music, previously selected by participants (Halko & Kaustia, 2012; Halko et al., 2015).

Nevertheless, almost all the studies that used pre-selected music justified the musical selection and assessed the music-mood congruence through different strategies. Some research carried out pilot studies to pretest the music (Brooks & Schweitzer, 2011, Study 1; Fischer & Greitemeyer, 2006; Greitemeyer, 2009a, Study 3, 2009b, Study 1, 2, 3 and 4, 2011, Study 5; Kniffin et al., 2016; Strick et al., 2015, Study 1), others used musical excerpts identified from previous studies (Chung et al., 2016; Schulreich et al., 2014; Skulmowski et al., 2014), while in Moritz and colleagues’ study (2009) the music–mood congruence was tested by experts (psychologists). Moreover, Day et al. (2009) controlled the ecological validity of the music, selecting a popular style, more accepted by office workers, than classical music which is more commonly used in research.

Almost all of the studies used music listening or background music, except for Kirschner and Tomasello (2010), who used live music making with background music. Regarding the characteristics of the musical stimuli, some studies specifically used vocal music (Fischer & Greitemeyer, 2006; Greitemeyer, 2009a, Study 3, 2009b, Study 1, 2, 3 and 4, 2011, Study 5; Kniffin et al., 2016, Study 1 and 2), while others used instrumental music (Brooks & Schweitzer, 2011, Study 1; Day et al., 2009; Moritz et al., 2009; Schulreich et al., 2014).

Regarding music styles, among the studies that used pre-selected music, Brooks and Schweitzer (2011, Study 1) and Chung et al. (2016) used both classical and popular music. Some studies used popular music, such as New Age Style, German songs, house music, rock and heavy metal (Day et al., 2009; Fischer & Greitemeyer, 2006, Study 1 and 3; Greitemeyer, 2009a, Study 3, 2009b, Study 1, 2, 3 and 4, 2011, Study 5; Kniffin et al., 2016; Moritz et al., 2009; Strick et al., 2015, Study 1). Moreover, Schulreich et al. (2014) selected music from various epochs and styles (classical music, Irish jigs, jazz, reggae, South American and Balkan music), while Kirschner and Tomasello (2010) used children’s music and Brooks and Schweitzer (2011, Study 1) and Skulmowski et al. (2014) selected music from soundtracks.

Some studies specified that music was used to elicit a particular mood or emotion, while others were interested in the effects of some particular musical or non-musical elements of the songs, such as music tempo (faster versus slower background music; Day et al., 2009) or the lyrics. In particular, the research group of Greitemeyer and colleagues investigated the effects of music with misogynous and man-hating content (Fischer & Greitemeyer, 2006, Study 1 and 3) and with prosocial versus neutral lyrics (Greitemeyer, 2009a, Study 3, 2009b, Study 1, 2, 3 and 4, 2011, Study 5). In particular, Fischer and Greitemeyer (2006, Study 1 and 3) matched all the songs according to music style, artist, level of aggression, negativity, and arousal level, and they used songs from different artists in their experiments. This allowed them to control that the effects of the music in the aggressive behavior were primarily attributed to the misogynous, man-hating, or neutral lyrics. Similarly, Greitemeyer (2009a, Study 3) controlled for the mood and the arousal dimension of songs, Greitemeyer (2009b, Study 1, 2, 3 and 4) controlled for mood and Greitemeyer (2011, Study 5) controlled for mood, arousal, as well as for the perceived aggressive content.

Type of emotion

Not all reviewed studies directly referred to an underlying emotion elicited by music and related to a decision-making process. Nevertheless, each of them investigated emotion either as an independent variable, elicited by music, or as a covariable, having a mediator effect between music and decision-making process.

For example, some studies directly investigated emotion or mood as an independent variable, assuming that music affects decision-making and behavior through mood modulation. Using Russell’s (1980) circumplex model of emotions, we can distinguish studies that were interested in the music-induced mood (or valence) effects on decision-making by investigating anxiety (Brooks & Schweitzer, 2011, Study 1), anxiety and happiness (Moritz et al., 2009), happiness and unhappiness (Schulreich et al., 2014), pleasant and unpleasant mood (Chung et al., 2016), or negatively valenced emotions (Skulmowski et al., 2014). Only Day et al. (2009) were interested just in the arousal or activation dimension (Russell, 1980), investigating the effects of tempo on decision-making, while Skulmowski et al. (2014) investigated both dimensions of emotion, the music-induced arousal and a negative valence.

Additionally, Strick et al. (2015) were interested in the effects of music on psychological transportation, which occurs when people are involved in a story or an experience and they “lose” temporarily the real world (Green & Brock, 2000). Transportation is a psychological concept which can be considered similar to arousal, because it is independent of the valence of emotion. Strick et al. (2015) also investigated the positive and negative affective response, which can be related to the mood or valence dimension of Russell’s (1980) model. The studies of Halko and colleagues, despite not being interested in specific emotions, assumed that mood has a mediator effect between music and risk-taking (Halko & Kaustia, 2012; Halko et al., 2015).

Other studies did not assume emotion as a mediator between music and decision-making and were interested in specific behavioral variables, such as aggressive behavior (Fischer & Greitemeyer, 2006; Greitemeyer, 2011) and prosocial behavior, defined by the authors as helping and cooperative behavior derived from social commitment and bonding (Greitemeyer, 2009a, 2009b, Study 1, 2, 3 and 4; Kirschner & Tomasello, 2010). Greitemeyer (2009b) controlled for mood in Study 1, while Kirschner and Tomasello (2010) discussed that the effects of music on prosocial behavior might be attributed to the positive mood induced by music. Kniffin et al. (2016) investigated the effects of music on happiness and unhappiness (Study 1), and on pleasant and unpleasant mood (Study 2). They controlled for mood and found that music effects might be both direct and indirect (mediated by mood).

To end, many studies assessed mood and emotion through Likert scales and questionnaires: the Positive and Negative Affect Schedule (PANAS; Watson et al., 1988; Chung et al., 2016; Greitemeyer, 2009a, Study 3, 2009b, Study 1, 2, 3 and 4); the four-item mood short form (Peterson & Sauber, 1983; Kniffin et al., 2016, Study 1 and 2); the Perceived Arousal Scale (Anderson et al., 1995; Greitemeyer, 2009a, Study 3, 2011, Study 5); and the State Hostility Scale (Anderson et al., 1995; Greitemeyer, 2011).

Decision-making tasks

All studies analyzed decision-making quantitatively. Despite the many different tasks, it was possible to group some of them according to a common characteristic: adding hot chili sauce, gambling, and helping someone.

Fischer and Greitemeyer (2006), in Study 1, and Greitemeyer (2011), in Study 5, used a similar decision-making task that consisted of putting hot chili sauce into a cup for another participant to consume in another experiment. The only difference between them was that Fischer and Greitemeyer (2006) presented the person (who was going to consume the sauce) before the task. In both studies the quantity of sauce grams put into the cup was considered the measure of aggression.

Halko and Kaustia (2012) and Halko et al. (2015) used almost the same gambling task, in which participants should accept or reject gambles that offered a 50% chance of gaining money. At the beginning of the task, the gain and loss amount were shown to the participants. Subsequently, they had to make the decision of accepting or rejecting the offer. The next gamble was presented, without providing the participant with the result of the last one.

Helping someone includes tasks involving: donating, picking up objects, or being willing to help. Greitemeyer (2009a), in Study 3, Strick et al. (2015), in Study 1, and Greitemeyer (2009b), in Study 3, used a donation decision-making task. In Greitemeyer (2009a) the task involved actual donating behavior, while Strick et al. (2015) and Greitemeyer (2009b) measured only donating behavior intentions. The donation question of Greitemeyer (2009b) was not just about money, but included options such as “have the kids come live with you until a permanent home was found” (Greitemeyer, 2009b, p. 1505).

During the experiment session of Kirschner and Tomasello (2010) and Greitemeyer (2009b, Study 1), the participant is faced with the opportunity to help another person: the researcher (Greitemeyer, 2009b) or another participant (Kirschner & Tomasello, 2010). Both studies used the number of dropped objects the participant decided to pick up from the floor as a measure of prosocial behavior. Moreover, Kirschner and Tomasello (2010) also considered the effort made by the participant as a variable. Greitemeyer (2009b) designed a scene where the researcher always asked for help, while in Kirschner and Tomasello (2010) the children involved in the “accident” were not aware of what was happening.

Greitemeyer (2009b), in Study 2, used a decision-making task involving the willingness to help others. After the main task, the participants were told the experiment was over. Subsequently, a different researcher entered the room asking for help in another research not involving any compensation, and participants had to indicate their willingness to take part and how much time they could give.

The other studies used heterogeneous decision-making tasks. Brooks and Schweitzer (2011) used a decision-making task where participants had to negotiate price, warranty period, and service period of a cell-phone shipment. In Day and colleagues’ study (2009) participants had to apply previously learned decision strategies to choose one protection lotion brand and four equivalent versions. In Moritz et al. (2009) participants were presented with paintings and possible titles. They could decide to connect a painting to a title or opt not to answer. In the study of Fischer and Greitemeyer (2006) participants were told the experiment was over and were asked to help in another experiment deciding for how long two subsequent participants should keep their hands in ice water.

Several experiments used famous decision-making tasks. Greitemeyer (2009b) used the dictator game, where participants have to decide how much money they want to give to a receiver. Chung et al. (2016) used the ultimatum game (Güth, Schmittberger, & Schwarze, 1982), where a proposer decides how to split a resource between themselves and a responder, who chooses whether to accept or reject the division. The proposal is implemented only if the responder accepts.

The Voluntary Contribution Mechanism (VCM), used by Kniffin et al. (2016), is a traditional public goods experiment about resource allocation (Isaac & Walker, 1988) that incentivizes and rewards cooperation. The participant has to decide how to allocate resources between an individual exchange (private good) and a group exchange (public good).

Schulreich et al. (2014) used the Random Lottery Pairs, where the participant is presented to an array of binary lottery choices and picks one of the lotteries in each pair, facing multiple pairs in sequence. In Schulreich et al. (2014) the two possible payoffs were strictly positive, limiting the experiment to the gain domain, differing from each other only in riskiness. It was expected that participants would make the riskier or the less risky decision. To end, Skulmowski et al. (2014) used the trolley dilemma, which consists of a scenario where a railway trolley is out of control and will run a number of people down. The participant chooses between sacrificing one person to save these people or letting them die.

Study results

Prosociality and aggression

Our review found several significant effects of music on decision-making tasks involving prosocial, cooperative, and aggressive behavior. In particular, Greitemeyer (2009a, Study 3) showed that music affected prosocial behavior, since participants who had listened to prosocial songs were more likely to donate money than participants in the neutral songs condition.

Similarly, in another article, Greitemeyer (2009b, Study 1, 2, 3 and 4) showed that listening to prosocial songs, compared to neutral music, promoted prosocial and cooperative behavior, since participants helped to pick up more dropped pencils (Study 1), reported a greater willingness to devote more time in further studies (Study 2) and to help in a hypothetical decision-making task (Study 3), and showed more cooperative behavior in the dictator game (Study 4). When controlling for positive and negative mood, the effect of type of song on prosocial behavior remained evident, suggesting that different mood states are unlikely to account for the effect of listening to prosocial songs on helping behavior. Conversely, in Study 4 empathy has been shown to mediate between music and cooperative behavior.

Similarly, Greitemeyer (2011, Study 5) showed that prosocial music not only increased prosocial and cooperative behavior, but also decreased aggression. In this study, participants in the prosocial music condition administered less hot chili sauce than did participants in the neutral music condition. Moreover, the significant effect of the music condition was reduced to non-significance when controlling for state hostility, suggesting that the effect of listening to prosocial music on aggressive behavior is mediated by differences in the reported aggressive affect.

Fischer and Greitemeyer (2006) showed that misogynous and man-hating song lyrics increased aggressive behavior in both men and women. Study 1 used the administration of hot chili sauce as a decision-making task to investigate aggressive behavior, while in Study 3 participants were asked to assign the time for two subsequent participants to keep their hands in ice water. Both studies showed that listening to misogynous and man-hating music increased the aggression toward the opposite sex.

Similarly, Kirschner and Tomasello (2010) showed that, after joint music making, children of both genders helped one another more than children in the no-music condition. Interestingly, children in the music condition that did not help offered verbal excuses more frequently than children in the no-music condition, suggesting that music fosters empathy and social commitment. Kniffin et al. (2016, Studies 1 and 2) selected as stimuli a set of happy and unhappy songs, that had been previously tested. Results showed that listening to happy music, compared both to unhappy music and no-music, increased cooperation, since participants in the happy music condition generated higher economic contributions to the public group. Differently from Greitemeyer (2009b, Studies 1, 2, 3 and 4), Kniffin et al. (2016, Study 2) suggested that happy music has both direct and indirect (mediated by mood) effects.

Instead of investigating happy and unhappy music, Strick et al. (2015, Study 1) were interested in the effects of moving and non-moving music on behavioral intentions in the context of video ads. Results showed that participants listening to moving music reported more willingness to donate money and to forward the film clip to five friends. As the authors hypothesized, moving music elicited a highly emotional transportation into the narrative of the video ad, therefore increasing participants’ cooperation and helping behavior by temporarily suppressing their tendency to consider the manipulative intent of the advertiser.

Differently from all the aforementioned studies, Skulmowski et al. (2014) did not find differences between the anxious music and the silent condition in the proportion of sacrificed single persons in a modified version of the trolley dilemma. Nevertheless, they found a significant difference in the score of negative emotions (“annoyed,” “angry,” “irritated,” and “gross”), indicating a stronger negatively valenced mood in the group that listened to music. Moreover, the eye tracking² results showed that the irritating music increased participants’ arousal. Contrary to the hypothesis of the study, music did not appear to affect moral decisions.

Risk-taking

Articles studying risk-taking (Halko & Kaustia, 2012; Halko et al., 2015; Schulreich et al., 2014) have all found correlations between emotion and risk-taking. Halko and Kaustia (2012) verified the impact of music on the tendency to participate in gambles, concluding that liked music alleviates loss aversion, while disliked music exacerbates it. The mean effects of liked and disliked music, compared to no music, are statistically significant. Liked music increased loss amount, while disliked music reduced it. Interestingly, the reduction caused by disliked music was more intense than the increase of loss amount provoked by liked music. The effect varied, however, for different types of gambles.

Halko et al. (2015) used functional MRI to analyze the brain mechanisms underlying music-induced unstable risk preferences. They gathered evidence that subjectively preferred music increases financial risk-taking and reduces loss aversion. When liked music was played, mean valence was significantly higher and arousal significantly lower compared to disliked music. The frequency for accepting a gamble grew in this order: disliked music, no music, and liked music. The disliked music, however, had no significant effect on gamble acceptance. Losses were weighed approximately twice as much as gains. Behavioral loss aversion, when calculated separately for liked music and disliked music, was significantly reduced in the liked music.

In Schulreich et al. (2014) music altered the emotional state of happiness in a different manner and this effect was diminished over time. While happy music was associated with greater happiness, sad music and random tone sequences were associated with lower happiness. The changes in participants’ choices were explained through the changes in how they convert objective probabilities into subjective decision weights. The affective nature of this link was corroborated by the results, showing that in the happy condition the decision weights associated with the larger payoffs were higher than in the sad and in the random tone conditions. In the happy condition participants chose more the riskier lottery compared to the “sad” condition. Combined, these findings suggest that subjectively liked music and happy music increase risk-taking, while disliked music and sad music decrease it.

Other decision-making processes

By comparing healthy controls with schizophrenia patients, Moritz et al. (2009) found that schizophrenia patients with current delusions made a higher number of decisions under the anxiety-induction condition, compared to non-deluded patients and healthy subjects.

Day et al. (2009) analyzed the differences between two different tasks (in terms of difficulty) and between two different music tempos. The independent variables were different dimensions of decision-making processes: decision accuracy, decision time, search pattern, percentage of information searched, reacquisition rate, and speed of information processing. Significant differences were found for accuracy and search pattern. Faster tempo condition presented higher accuracy, suggesting that participants in this group solved more correctly the decision problem. Moreover, faster tempo condition showed a lower search pattern, which is “used to reflect whether a subject’s information search direction tends to be attribute-based (intra-dimensional) or alternative-based (inter-dimensional)” (Day et al., 2009, p. 141). Participants who listened to faster music used more intra-dimensional search to solve the problem, compared to participants in the slower tempo condition. This suggests that faster music allows attribute-based processing, where “the values of several alternatives on a single attribute are processed before information about a second attribute is processed” (Day et al., 2009, p. 138). An interaction between music tempo and task difficulty was found for search pattern, showing that only for a WADD task there was a significant effect, where faster tempo presented a lower search pattern. A significant interaction effect of music tempo and task difficulty was found for only one of the tasks (EBA), where the reacquisition rate was significantly lower with the presence of the faster tempo.

Chung et al. (2016) induced pleasant and unpleasant moods, using musical excerpts that had been selected in a previous study. Participants were assigned either to a pleasant group, which listened to pleasant music, or an unpleasant group, which listened to unpleasant excerpts. Results showed that the unpleasant group reported increased negative affect after mood induction relative to pre-induction, while the pleasant group reported marginally decreased negative affect. A significant interaction between time and group was found for positive affect scores, suggesting that the less favorable the offers were, the more participants rejected them. A significant interaction was also found between offer and group, revealing that the unpleasant group rejected unfair offers more often than the pleasant group. Rejection rates of unfair offers were negatively correlated with felt fairness and felt happiness in the unpleasant group, indicating that the unpleasant group rejected more often, as they perceived the offer to be less fair and were less happy. The pleasant group also showed a negative correlation between felt happiness and rejection rates.

Brooks and Schweitzer (2011) have compared an independent variables emotion condition (anxiety or neutral) and role (buyer or seller) with dependent variables such as aspirations, expectations, first offers, exit decisions, and individual outcomes. Emotion showed a significant effect on expectations. Participants in the anxious music condition expected to earn less than participants in the non-anxious music group. Considering the first offers, participants in the anxious music condition made significantly lower first offers than participants in the non-anxious music condition. With regard to the individual outcomes, the analyses showed that participants in the anxious music group earned less profit than participants in the non-anxious music condition, as well as that participants who were paired with an anxious counterpart earned significantly more profit than participants who were paired with a non-anxious counterpart.

Discussion

This qualitative systematic review aimed to examine the music-induced emotion effects on decision-making. Our results showed that music affects the decision-making process, influences risk-taking and promotes prosociality. Music is a complex and powerful stimulus in people’s daily life and several factors may contribute to the emotional or behavioral responses to music: the musical properties of the stimulus, participants’ individual characteristics, and contextual factors (Liljeström et al., 2012).The majority of the studies did not take into account participants’ musical preferences. Nevertheless, there is a general consensus that both properties of musical stimulus and individual differences contribute significantly to the induction of musical emotion (Kreutz, Ott, Teichmann, Osawa, & Vaitl, 2008; Liljeström et al., 2012; Trehub et al., 2015). In particular, Liljeström et al. (2012) showed that self-chosen music aroused more intense and more positive emotions than randomly sampled music. Future research must consider participants’ musical preferences in order to increase their ecological validity.

Evidence has shown that making music induces more positive affect than music listening (Dunbar et al., 2012). Nevertheless, just Kirschner and Tomasello (2010) used live music making. An active involvement in music, as in singing, dancing, and drumming, promotes more endorphin release and elicits more positive affect, while simply listening to music does not induce the same effects (Dunbar et al., 2012). Kirschner and Tomasello (2010) found that joint music making has positive effect on prosociality, supporting the evolutionary hypothesis of the origin of music, which suggests that it allows the creation and maintaining of social bonds and prosocial commitment among social groups. Interestingly, almost all the studies that investigated the music effects on prosocial or aggressive behavior used vocal music instead of instrumental. This is in line with research that shows that singing is particularly powerful in fostering caregiver–infant bonding (Trehub et al., 2015), suggesting that vocal music may be more efficient in promoting prosocial and cooperative behaviors. Similarly, a recent study by Weinstein, Launay, Pearce, Dunbar, and Stewart (2016) showed that singing, in both small and large groups, increases positive affect, feelings of inclusion and social connection, and causes elevated pain thresholds, generally associated to a higher endorphin release. These effects are greater in larger groups than in smaller and more familiar groups, suggesting that music making is particularly effective in fostering social connections between large groups of unfamiliar individuals. Future studies might compare the use of vocal and instrumental music in decision-making processes that involve aggressive and prosocial behavior.

We did not find homogeneity in decision-making tasks, probably due to the different theoretical background of the included articles. The Iowa Gambling Task, for example, is a risk-taking decision-making task, commonly used to investigate the role of emotion in decision-making processes (Reimann & Bechara, 2010), but none of the included studies employed it. Some studies have focused on decision-making processes that involve a long deliberation and many steps to be followed (for example, Day et al., 2009), while others have used almost automatic responses as decision-making tasks (for example, Greitemeyer, 2011; Kirschner & Tomasello, 2010).

The studies highlighted explications for the interaction or the mediator factors between music and decision-making. For example, Day et al. (2009) investigated the mediation of attention, by comparing the two different perspectives of music as a distractor or as an arousal inducer. According to Kahneman’s capacity model (Kahneman, 1973), background music might be a distractor, because it demands cognitive capacity and reduces the mental resource available for the cognitive task, thus impairing the decision-making process. On the contrary, music might be an arousal inducer and might increase listeners’ mental resource, thus promoting decision-making. The results of the studies supported the arousal effect of music.

One of the most investigated mediator factors between music and decision-making is mood. While many studies supported the mood induction effect of music, others challenged it. Studies that investigated prosociality showed that neither cognition or mood can explain the interaction between music and decision-making process. On the contrary, specific prosocial feelings like interpersonal empathy were elicited by prosocial songs and promoted prosocial behavior. Results were in line with the General Aggression Model (GAM; Bushman & Anderson, 2002) and the General Learning Model (GLM; Buckley & Anderson, 2006). The GAM suggests that violent media may affect aggressive ideas, by evoking associations to aggressive cognitions, arousal, and affect related to violence. The GLM suggests that exposure to media may affect a person’s cognition, affect and arousal, thus affecting their behavior.

Nevertheless, both models were only partially supported, because the effects of music exposure on social behavior were mediated more by the affective route than the cognitive route, which is more commonly activated by video games. Combining these findings, we suggest that music affects both directly and indirectly (with the mediation of mood, emotion, and affection) the decision-making process, while the cognitive route seems to have a smaller influence. Nevertheless, as a limitation of these studies we highlight that the effects of music on prosociality may be partially affected by the prosocial content of the lyrics. The aforementioned study by Weinstein et al. (2016) also investigated the effects of music on prosociality. Even if the music used in this study had lyrics, it did not have prosocial content and researchers were interested in the effects of music making, rather than the effects of music per se. This research also highlighted the relationship between music and mood, showing that singing increased positive affect, feelings of inclusion, and social connection. Future research might investigate separately the effects of prosocial music without the lyrics, prosocial song lyrics without the music, and the interaction of both.

With regards to risk-taking, the included studies confirmed the current literature (Reimann & Bechara, 2010; Guillaume et al., 2009) by showing that emotion interferes in decision-making. For the other articles not related to risk-taking or prosocial behavior, music was also an effective arousal or mood inducer. Nevertheless, we highlight that some studies did not find significant effects in all the investigated conditions. For example, Moritz et al. (2009) only found significant effect when they split the patient group for delusion to find an interaction between music and group. Another example is in Brooks and Schweitzer’s study (2011), where no significant effect of emotion (anxious vs. neutral) was found on aspiration, and no interaction effect of the negotiator’s emotion and the counterpart’s emotion on individual outcomes.

Despite these non-significant or inconclusive results, most of the studies showed that music is effective in eliciting arousal or mood variability in different kind of tasks and in affective decision-making. This variability of tasks supports a consistency of the effect, and provides a rationale for the use of music as an arousal and mood inducer in different areas of research. However, our findings showed that music is still not so widespread as one might expect in the research involving decision-making. Interestingly, six of the articles included in this review are part of a line of studies made by two groups of researchers and their colleagues. This shows that, once music is used in the experiments and it shows positive outcomes to the research field, researchers are motivated to continue their investigation with this stimulus, to provide other evidence and improve the theory.

To end, this review allows a reflection about the necessity of environmental control during any kind of experiment, in order to guarantee the internal validity of research. Our results showed that music might be an effective intervenient variable over emotion, interfering in research results. Generalizing our findings, even experiments that do not investigate music must take into account and control the emotional and behavioral effects of musical and acoustic stimuli (e.g. the presence of music or noise in the acoustic background before or during the experimental setting). Finally, this review provides a rationale for the use of music-based interventions to foster prosocial and cooperative behavior and decrease aggression.

Footnotes

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This publication was supported by the Brazilian foundation “Comissão de Aperfeiçoamento de Pessoal de Nível Superior” (CAPES).

Notes

References

Anderson

C. A.

Deuser

W. E.

DeNeve

(1995). Hot temperatures, hostile affect, hostile cognition, and arousal: Tests of a general model of affective aggression. Personality and Social Psychology Bulletin, 21, 434–448.

Anderson

C. A.

Benjamin

A. J.

Wood

P. K.

Bonacci

A. M.

(2006). Development and testing of the Velicer Attitudes Toward Violence Scale: Evidence for a four-factor model. Aggressive Behavior, 32, 122−136.

Blood

Zatorre

R. J.

(2001). Intensely pleasurable responses to music correlate with activity in brain regions implicated in reward and emotion. Proceedings of the National Academy of Sciences USA, 98, 11818–11823.

Brockmeyer

Simon

J. J.

Becker

Friederich

H. C.

(2017). Reward-related decision making and long-term weight loss maintenance. Physiology & Behavior, 181, 69–74. doi:10.1016/j.physbeh.2017.09.008

Brooks

A. W.

Schweitzer

M. E.

(2011). Can Nervous Nelly negotiate? How anxiety causes negotiators to make low first offers, exit early, and earn less profit. Organizational Behavior and Human Decision Processes, 115, 43–54.

Buckley

K. E.

Anderson

C. A.

(2006). A theoretical model of the effects and consequences of playing video games. In Vorderer

Bryant

(Eds.), Playing video games: Motives, responses, and consequences (pp. 363 −378). Mahwah, NJ: Lawrence Erlbaum.

Bushman

B. J.

Anderson

C. A.

(2002). Violent video games and hostile expectations: A test of the general aggression model. Personality and Social Psychology Bulletin, 28, 1679 −1686.

Butler

Hall

Copnell

(2016). Guide to writing a qualitative systematic review protocol to enhance evidence-based practice in nursing and health care. Worldviews on Evidence-Based Nursing, 13(3), 241–249.

Coan

J. A.

Allen

J. J. B.

(2007). The handbook of emotion elicitation and assessment. New York, NY: Oxford University Press.

10.

Chung

Lee

E. J.

Jung

Y. J.

Kim

S. H.

(2016). Music-induced mood biases decision strategies during the ultimatum game. Frontiers in Psychology, 7(453). doi:10.3389/fpsyg.2016.00453

11.

Damasio

A. R.

(1994). Descartes’ error: Emotion reason, and the human brain. New York, NY: Putnam.

12.

Damasio

A. R.

(1996). The somatic marker hypothesis and the possible functions of the prefrontal cortex. Philosophical Transactions of the Royal Society B: Biological Sciences, 351, 1413–1420.

13.

Damasio

A. R.

Tranel

Damasio

(1991). Somatic markers and the guidance of behavior: Theory and preliminary testing. In Levin

H. S.

Eisenberg

H. M.

Benton

A. L.

(Eds.), Frontal lobe function and dysfunction (pp. 217–229). New York, NY: Oxford University Press.

14.

Day

R. F.

Lin

C. H.

Huang

W. H.

Chuang

S. H.

(2009). Effects of music tempo and task difficulty on multi-attribute decision-making: An eye-tracking approach. Computers in Human Behavior, 25(1), 130–143. Retrieved from https://doi.org/10.1016/j.chb.2008.08.001

15.

Dunbar

R. I. M.

Kaskatis

MacDonald

Barra

(2012). Performance of music elevates pain threshold and positive affect: Implications for the evolutionary function of music. Evolutionary Psychology, 10, 688–702.

16.

Fischer

Greitemeyer

(2006). Music and aggression: The impact of sexual-aggressive song lyrics on aggression-related thoughts, emotions, and behavior toward the same and the opposite sex. Personality and Social Psychology Bulletin, (32), 1165–1176. doi:10.1177/0146167206288670

17.

Frank

M. J.

Cohen

M. X.

Sanfey

A. G.

(2009). Multiple systems in decision making: A neurocomputational perspective. Current Directions in Psychological Science, 18(2), 73–77.

18.

Fusar-Poli

Bieleninik

Ł.

Brondino

Chen

X. J.

Gold

(2017). The effect of music therapy on cognitive functions in patients with dementia: A systematic review and meta-analysis. Aging & Mental Health, 10, 1–10. doi:10.1080/13607863.2017.1348474

19.

Green

M. C.

Brock

T. C.

(2000). The role of transportation in the persuasiveness of public narratives. Journal of Personality and Social Psychology, 79, 701–721. Retrieved from https://dx-doi-org.web.bisu.edu.cn/10.1037/0022–3514.79.5.701

20.

Greene

J. D.

Nystrom

L. E.

Engell

A. D.

Darley

J. M.

Cohen

J. D.

(2004). The neural bases of cognitive conflict and control in moral judgment. Neuron, 44(2), 389–400.

21.

Greitemeyer

(2009a). Effects of songs with prosocial lyrics on prosocial thoughts, affect, and behavior. Journal of Experimental Social Psychology, 45, 186–190.

22.

Greitemeyer

(2009b). Effects of songs with prosocial lyrics on prosocial behavior: Further evidence and a mediating mechanism. Personality and Social Psychology Bulletin, 35, 1500−1511.

23.

Greitemeyer

(2011). Exposure to music with prosocial lyrics reduces aggression: First evidence and test of the underlying mechanism. Journal of Experimental Social Psychology, 47, 28–36.

24.

Guillaume

Jollant

Jaussent

Lawrence

Malafosse

Courtet

(2009). Somatic markers and explicit knowledge are both involved in decision-making. Neuropsychologia, 47, 2120–2124.

25.

Güth

Schmittberger

Schwarze

(1982). An experimental analysis of ultimatum bargaining. Journal of Economic Behavior and Organization, 3, 367–388. doi:10.1016/0167–2681(82)90011–7

26.

Halko

M. L.

Kaustia

(2012). Are risk preferences dynamic? Within-subject variation in risk-taking as a function of background music. SSRN. Retrieved from: http://ssrn.com/abstract!1997520 or https://dx-doi-org.web.bisu.edu.cn/10.2139/ssrn.1997520

27.

Halko

M. L.

Mäkelä

Nummenmaa

Hlushchuk

Schürmann

(2015). Hedonic context modulates risky choices and reward responses in amygdala and dorsal striatum. Journal of Neuroscience, Psychology, and Economics, 8(2), 100–115. Retrieved from https://dx-doi-org.web.bisu.edu.cn/10.1037/npe0000036

28.

Hansen

(2005). Distinguishing between feelings and emotions in understanding communication effects. Journal of Business Research, 58, 1426–1436.

29.

Hansen

Christensen

S. R.

(2007). Emotions advertising and consumer choice. Copenhagen, Denmark: Copenhagen Business School Press.

30.

Isaac

Walker

(1988). Group size effects in public goods provision: The voluntary contribution mechanism. Quarterly Journal of Economics, 103, 179–199.

31.

Kahneman

(1973). Attention and effort. New Jersey, NJ: Prentice-Hall.

32.

Kirschner

Tomasello

(2010). Joint music making promotes prosocial behavior in 4-year-old children. Evolution and Human Behavior, 31, 354–364

33.

Kitajima

Toyota

(2013). Decision-making and action selection in Two Minds: An analysis based on Model Human Processor with Realtime Constraints (MHP/RT). Biologically Inspired Cognitive Architectures, 5, 82–93.

34.

Kniffin

K. M.

Yan

Wansink

Schulze

W. D.

(2016). The sound of cooperation: Musical influences on cooperative behavior. Journal of Organizational Behavior, 38(3), 372–390.

35.

Koelsch

(2009). A neuroscience perspective on music therapy. Annals of the New York Academy of Sciences, 1169, 374–384.

36.

Koelsch

(2011). Toward a neural basis of music perception: A review and updated model. Frontiers in Psychology, 2, 110. Retrieved from http://doi.org/10.3389/fpsyg.2011.00110

37.

Koelsch

(2014). Brain correlates of music-evoked emotions. Nature Review Neuroscience, 15(3), 170–180. doi:10.1038/nrn3666

38.

Koelsch

(2015). Music-evoked emotions: Principles, brain correlates, and implications for therapy. Annals of the New York Academy of Sciences, 1337, 193–201. doi:10.1111/nyas.12684

39.

Kreutz

Ott

Teichmann

Osawa

Vaitl

(2008). Using music to induce emotions: Influences of musical preference and absorption. Psychology of Music, 36(1), 101–126.

40.

Liljeström

Juslin

P. N.

Västfjäll

(2012). Experimental evidence of the roles of music choice, social context, and listener personality in emotional reactions to music. Psychology of Music, 41(5), 579–599.

41.

Loewenstein

(1996). Out of control: Visceral influences on behavior. Organizational Behavior and Human Decision Processes, 65(3), 272–292.

42.

Loewenstein

Lerner

(2003). The role of affect in decision making. In Davidson

R. J.

Scherer

K. R.

Goldsmith

H. H.

(eds.), Handbook of affective science (pp. 619–642). Oxford, UK: Oxford University Press.

43.

Loewenstein

G. F.

Weber

E. U.

Hsee

C. K.

Welch

(2001). Risk as feelings. Psychological Bulletin, 127, 267–286.

44.

Malik

A. S.

Amin

H. U.

(2017). Designing EEG experiments for studying the brain: Design code and example. London, UK: Academic Press.

45.

Merriam-Webster. (n.d.). Music. In Merriam-Webster Dictionary. Retrieved from http://www.merriam-webster.com/dictionary/music.

46.

Moher

Liberati

Tetzlaff

Altman

D. G.

, & The PRISMA Group.(2009). Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. PLoS Medicine, 6(7). doi:10.1371/journal.pmed.1000097

47.

Moritz

Veckenstedt

Randjbar

Hottenrott

Woodward

T. S.

Eckstaedt

F. V. V.

Lincoln

T. M.

(2009). Decision making under uncertainty and mood induction: Further evidence for liberal acceptance in schizophrenia. Psychological Medicine, 39(11), 1821–1829. Retrieved from https://doi.org/10.1017/S0033291709005923

48.

Peterson

R. A.

Sauber

(1983). A mood scale for survey research. In Murphy

Laezniak

Anderson

Belk

Ferrell

O. C.

Lusch

Shimp

Weinberg

(Eds.), Educator’s proceedings (pp. 409–414). Chicago, IL: American Marketing Association.

49.

Phelps

E. A.

(2009). The study of emotion in neuroeconomics. In Glimcher

P. W.

Camerer

C. F.

Fehr

Poldrack

R. A.

(Eds.), Neuroeconomics: Decision making and the brain (pp. 233–250). London, UK: Academic Press.

50.

Posner

Russell

J. A.

Peterson

B. S.

(2005). The circumplex model of affect: An integrative approach to affective neuroscience, cognitive development, and psychopathology. Development and Psychopathology, 17, 715–734.

51.

Reimann

Bechara

(2010). The somatic marker framework as a neurological theory of decision-making: Review, conceptual comparisons, and future neuroeconomics research. Journal of Economic Psychology, 31, 767–776.

52.

Russell

J. A.

(1980). A circumplex model of affect. Journal of Personality and Social Psychology, 39(1), Supp. 6, 1161–1178.

53.

Sachs

Kaplan

Der Sarkissian

Habibi

(2017). Increased engagement of the cognitive control network associated with music training in children during an fMRI Stroop task. PLoS One, 12(10): e0187254. doi:10.1371/journal.pone.0187254.

54.

Shafir

(1999). Decision making. In Wilson

R. A.

Keil

F. C.

(Eds.), The MIT encyclopedia of the cognitive sciences (pp. 220–223). Cambridge, MA: MIT Press.

55.

Schulreich

Heussen

Y. G.

Gerhardt

Mohr

P. N. C.

Binkofski

F. C.

Koelsch

Heekeren

H. R.

(2014). Music-evoked incidental happiness modulates probability weighting during risky lottery choices. Frontiers in Psychology, 4, 981. Retrieved from https://dx-doi-org.web.bisu.edu.cn/10.3389/fpsyg.2013.00981

56.

Skulmowski

Bunge

Kaspar

Pipa

(2014). Forced-choice decision-making in modified trolley dilemma situations: A virtual reality and eye tracking study. Frontiers in Behavioral Neuroscience, 8, 426. Retrieved from http://doi.org/10.3389/fnbeh.2014.00426

57.

Sokol-Hessner

Camerer

C. F.

Phelps

E. A.

(2013). Emotion regulation reduces loss aversion and decreases amygdala responses to losses. Social Cognitive and Affective Neuroscience, 8, 341–350.

58.

Stern

Jordan

McArthur

(2014). Developing the review question and inclusion criteria: The first steps in conducting a systematic review. American Journal of Nursing, 114(4), 53–56.

59.

Strick

de Bruin

H. L.

de Ruiter

L. C.

Jonkers

(2015). Striking the right chord: Moving music increases psychological transportation and behavioral intentions. Journal of Experimental Psychology: Applied, 21(1), 57–72. Retrieved from https://dx-doi-org.web.bisu.edu.cn/10.1037/xap0000034

60.

Trehub

S. E.

Becker

Morley

(2015). Cross-cultural perspectives on music and musicality. Philosophical Transactions of the Royal Society B, 370, 20140096. Retrieved from https://dx-doi-org.web.bisu.edu.cn/10.1098/rstb.2014.0096

61.

Tremblay

R. E.

(2000). The development of aggressive behavior during childhood: What we learned in the past century? International Journal of Behavioral Development, 24, 129–141.

62.

Tversky

Kahneman

(1974). Judgment and uncertainty: Heuristics and biases. Science, 185(4157), 1124–1131.

63.

Watson

Clark

L. A.

Tellegen

(1988). Development and validation of brief measures of positive and negative affect: The PANAS scales. Journal of Personality and Social Psychology, 54(6),1063–1070.

64.

Weinstein

Launay

Pearce

Dunbar

R. I. M.

Stewart

(2016). Group music performance causes elevated pain thresholds and social bonding in small and large groups of singers. Evolution and Human Behavior, 37(2), 152–158. Retrieved from http://doi.org/10.1016/j.evolhumbehav.2015.10.002

65.

Zheng

Yang

Jin

Liu

(2017). The influence of emotion on fairness-related decision making: A critical review of theories and evidence. Frontiers in Psychology, 19. doi:10.3389/fpsyg.2017.01592

66.

Zinchenko

Arsalidou

(2018). Brain responses to social norms: Meta-analyses of fMRI studies. Human Brain Mapping, 39(2), 955–970. doi:10.1002/hbm.23895.

67.

Zumbansen

Peretz

Hébert

(2014). Melodic intonation therapy: Back to basics for future research. Frontiers in Neurology, 5(7), 1–11.