Abstract
This study examines whether background music gives young Japanese adults a more favorable impression of conversation partners of the opposite sex for the first meeting during conversation for konkatsu. A total of 32 students in their teens and twenties (16 males and 16 females) participated in the experiment. The two conditions were as follows: the ‘with-music’ condition, where BGM was played during a 20-minute conversation, and the ‘no-music’ condition, where BGM was absent. The participants were divided into eight small groups, consisting of two males and two females each. Four other students (two males and two females) participated in pairs (two couples) in each group as the target conversation partners (called ‘guests’ here). The participants rated their impressions of the guests of the opposite sex before and after the conversation. The guests were instructed to keep the conversation smooth and amicable. The results showed that irrespective of the presence of background music, face-to-face conversation between young males and females led to a change in their impression of their partner and created intimacy between them. The biggest difference between the no-music and with-music conditions is the point that music facilitated the feelings of love for the partner of the opposite sex.
Music and conversation are a part of various facets of our daily life, for example, at konkatsu parties (‘the activity of finding a better marriage partner’) in Japan. As part of this activity, parties for matchmaking, blind-dates and meeting potential dates have undergone a recent boom in Japan (Yamada & Shirakawa, 2008). These parties, called gokon, are held very frequently around the country. Matchmaking companies have also boomed. At such parties, young people search for a potential spouse and simultaneously try to showcase their own personal appeal and attractiveness. The party organizers put on various programs for participants to attract prospective candidates, and music is one of them. For example, party participants may enjoy singing karaoke or talking as background music is played. Music is thought to evoke romantic feelings and create intimacy between young adults. Recently, Guéguen, Jacob, and Lamy (2010) reported the effects of music on courtship and love. However, studies have not yet investigated the effect of background music on the impressions formed by people when they first meet. Therefore, this research first examines whether background music influences impression formation in first meetings.
Background music is played in a variety of public settings, such as hospitals or restaurants. For example, it is used to reduce anxiety during hospital stays. Music is one of the most effective means of mood induction, as many researchers have showed that it can alter peoples’ moods and emotions (Västfjäll, 2002). For example, using four conditions, Fried and Berkowitz (1979) investigated the moods of three groups of participants who listened to a 7-minute musical selection that was either soothing, stimulating or aversive in nature, while the remaining group of participants was not exposed to any music. The mood rating results indicated that soothing and stimulating music created somewhat different positive moods (e.g. joyful and soothed) while aversive music tended to arouse negative feelings (e.g. bored and distressed). Those who heard soothing music were most likely to be helpful immediately afterwards, significantly more so than those who heard aversive music or no music at all, perhaps because of the ideas evoked by the particular musical selection.
Participants may not be fully aware of the effects of background music, although in most cases, the effect may be minimal. For the purpose of investigating the effects of background music on emotional behavior, many researchers ask participants to listen carefully to preselected music or to choose their favorite music (Västfjäll, 2002) so that they focus intentionally on the selection. When listeners focus intentionally on background music that is either preferred or disliked, there are various positive and negative effects (see, for example, Juslin & Sloboda, 2001; Levitin, 2006). Then, we may ask, what effects does music have in cases where the individual is not consciously paying attention to it? Our understanding of the effects of background music is thus somewhat limited.
Furthermore, only a few studies have investigated the effect of background music in field settings (e.g. stores). For example, North, Hargreaves, and McKendrick (1999) reported that in the alcoholic beverages section of a supermarket, French background music led to French wines outselling German ones, whereas German background music had the opposite effect. However, it is difficult to control all the factors in field experiments. One might argue that North et al.’s (1999) results are influenced by the number of consumers or the types of goods displayed on the sales floor. It is possible that the effects attributed to background music may be dependent on other environmental factors.
Although many studies have demonstrated the effects of music on listeners’ emotions or moods, it has not yet been examined whether music influences individuals’ impressions specifically of opposite-sex conversation partners. Rentfrow and Gosling (2006) showed that, using an on-line bulletin-board system, music is a common topic of conversation among young strangers (same-sex and opposite-sex pairings, analysed together) engaged in the task of getting to know one another, and that music preferences are related to targets. Conversation is also one of the essential factors that facilitate the relationship between two people meeting for the first time, but its function has not been investigated in detail. Ogawa (2000) examined the balance of the amount of utterances exchanged in conversation in dyads and found a positive, high correlation between one’s impression of a partner and impression of the conversation; a good impression would lead to an expectation of a good future relationship between the two people. Thus, the conversation that occurs at a first meeting is very important. Panksepp and Bernatzky (2002) described how music serves as a magnet for social activity and that we may expect music has clear effects on neurochemical systems in the brain that help mediate social processes. It therefore seems important to investigate the effect of music on conversation. In this study, we test the effects of background music on young adults’ impressions of their conversation partners in the context of meeting them for the first time at a gokon party for konkatsu. This research thus examines the effects of background music on interpersonal impression formation in conversations that occur during first meetings.
Method
Participants
The participants, 32 young students (average age = 20.3 years, SD = 2.1), were divided into two groups, each comprising 16 students (8 males, 8 females). As they talked, background music was played for one group but not for the other.
Controlling the various conditions in actual settings where background music is played can be difficult. This research replicates the conditions of a gokon party for konkatsu in a laboratory setting. For this purpose, the participants in this research were not instructed to listen to background music. They were only told to enjoy the conversation and rate their impressions of conversation partners of the opposite sex.
Stimuli
The background music selections were chosen beforehand by other undergraduate students at the same university. A preliminary test was conducted to choose five pieces of music that were popular or favorites among young Japanese adults. Eleven different students were given 11 pieces of music of various genres and asked to rate liking for each piece on a 5-point scale from 1 = ‘dislike’ to 5 = ‘like.’ Based on their responses, five pieces of music from three genres were selected as the background music for the study (see Table 1).
Five musical selections used in the experiment as background music.
The ratings for each piece of music (out of 5) were as follows: ‘TSUNAMI’ (4.1, SD = 1.2), ‘Violin Sonata’ (3.5, SD = 1.2), ‘Serenade’ (3.6, SD = 1.3), ‘If we ever meet again’ (3.6, SD = 1.2) and ‘Heavy Rotation’ (3.4, SD = 1.4).
‘TSUNAMI’ is a song released in 2000 by the Southern All Stars, a highly influential Japanese rock band. The first movement of violin sonata No. 5 ‘Spring’ by Beethoven and Serenade ‘Eine kleine Nachtmusik’ by Mozart are both extremely popular pieces of classical music in Japan. ‘If we ever meet again’ was released in 2009 by American popular rapper Timbaland. ‘Heavy Rotation’ was released in 2010 by the Japanese idol group AKB48, which is extremely popular especially among young people in Japan.
Procedure
The participants were divided into eight small groups each consisting of two males and two females. Four other college seniors (two males and two females) participated in pairs (two couples) in each group as the target conversation partners (referred to here as ‘guests’). Therefore, each group consisted of six people (two guests and four participants). The guests were instructed to lead the conversation and make it smooth and amicable. The participants and guests were allowed to talk freely on various topics. The participants rated their impressions of the opposite-sex guest twice: before and after the conversation. Two tables were prepared and two groups in the same condition (‘no-music’ condition or ‘with-music’ condition) were simultaneously tested at separate tables. After the first half of the experiment, the guests switched tables (i.e., groups), and the same process was repeated. Thus the participants were given the same session twice, and they rated the impression of a different guest in each session so that their rating performance would not depend on the unique character of a particular guest.
Figure 1 shows the placement of guests and participants in the experiment. The six successive sessions were run in the first half and repeated in the second half.
The flow of experiment
First half
Session 1. The participants (two males and two females) entered the room and sat down at the table where the guests (a couple) were seated. Each participant sat down beside a guest of the opposite sex. All the members gave a simple self-introduction.
Session 2. The guests left the room, and the participants rated their impressions of the guest of the opposite sex. The rating items consisted of 10 adjective pairs, and the participants rated their impressions of the guest using a seven-level semantic differential method (Table 2).
Eight rating items were selected from the Characteristic Adjective Scales (Hayashi, 1976), and we added two further items. The initial eight items were intended to check the participant’s impression of the partner, while the additional two items were intended to check feelings of attraction for the partner of the opposite sex.
Session 3. In the with-music condition, after the ratings were completed, ‘TSUNAMI’ was played for about 5 minutes through a small loudspeaker placed on the table. Close to the end of the piece, the guests entered the room and sat down in their original positions. Similarly, in the no-music condition, after the ratings, the guests entered and sat down in their original positions.
Session 4. The participants talked freely within their group. The guests were instructed to try to talk with all the group members and to make the conversation smooth and amicable. They were not given any specific instructions on what to talk about. Instead, they were encouraged to talk about anything so that they would enjoy the conversation. In the with-music condition, they were given four pieces of background music in the following order during an approximately 20-minute conversation: ‘Violin Sonata No. 5,’ ‘Eine kleine Nachtmusik,’ ‘If we ever meet again’ and ‘Heavy Rotation.’ In the no-music condition, they engaged in an approximately 20-minute conversation in the absence of background music.
Session 5. In the with-music condition, several minutes after the beginning of ‘Heavy Rotation,’ the guests left the room and then the music faded out. In the no-music condition, the guests left the room after the conversation.
Session 6. After the guests left the room, the participants in both conditions filled out a response sheet in which they rated their impressions of the guests for the second time, answered questions about themselves, rated the environment (with or without music) and responded to various questionnaire items. The participants in the with-music condition were also asked about the background music. For example, they were asked if any of the pieces were preferred or disliked.
Ten adjective pairs that were used for rating impressions of the guests in the experiment.
I feel attracted to you.
Second half
The guests moved to another table, and the participants repeated sessions 1–6. The participants received coupons for books for their participation and then left the room.
The entire experiment lasted for approximately 75 minutes. The procedure was designed to test the extent to which participants’ evaluation of a guest changed after conversing with him or her in a gokon-like setting. The experiment was conducted in a laboratory attached to the Department of Psychology in Aoyama Gakuin University. The music was played through a small loudspeaker (SONY, SRS-BTV5) placed on each table. The presentation level of the music ranged from 50 dB (A) to 76.9 dB (A), with an average level of about 70 dB (LAeq). The time allocated for the evaluation of a guest was about 50s.
Results
The data of one female participant were excluded from the analysis because she failed to follow the instructions. The rating responses of two guests (one from the first half of the experiment and one from the second half) were merged before conducting the analysis of variance (ANOVA). First, we statistically confirmed that participants in the no-music condition gave equivalent ratings of the same guests as participants in the with-music condition. A one-way ANOVA between the two conditions (no-music vs. with-music) was conducted on the averaged rating values of the adjective pairs. The main effect of participants was not significant, F(1, 29) = .822, p = .372. This means that the two groups could be considered equivalent; therefore, the following analyses of the two conditions were conducted separately.
Figure 2 shows the direction and quantity of the change in impression caused by engaging in conversation. Positive values mean that the impression of guests improved, while negative values mean it worsened.
The direction and amount of the change of impression of a partner during the conversation. The ordinate is the difference obtained when the score before the conversation is subtracted from the score after the conversation.
Table 3 shows each averaged value of the 10 rating terms for ‘before conversation’ and ‘after conversation.’
Average score and standard deviation in each condition before and after conversation.
p < .001; ** p < .01; * p < .05; † .05 ≤ p < .10.
From Figure 2 and Table 3, we compared the direction and quantity of the change in rating of guests in the no-music and with-music conditions.
No-music condition
A two-way ANOVA with factors of the timing of impression rating (before vs. after conversation) and rating item (10) yielded a significant interaction, F(9, 135) = 5.57, p < .001. The dependent variable was the value of the rating items. This means that after conversation, impressions of guests either improved or worsened. The main effects of the timing of impression rating (before vs. after conversation) were calculated. As shown in Table 3(a), impressions of guests significantly improved for ‘passive–active,’ F(1, 15) = 30.32, p < .001, ‘inaccessible–accessible,’ F(1, 15) = 11.50, p = .004, ‘unsociable–sociable,’ F(1, 15) = 16.35, p = .001, ‘unfriendly–friendly,’ F(1, 15) = 20.21, p < .001, and ‘hard to like–likeable,’ F(1, 15) = 8.57, p = .010. Impressions of guests worsened for ‘careless–careful,’ although not significantly, F(1, 15) = 0.53, p = .478. For the other rating terms (‘unpleasant–pleasant,’ ‘unassertive–confident,’ ‘short-tempered–patient,’ and ‘would not like to date–would like to date’), the main effects of the timing of impression rating were not significant.
The results suggest that participants were able to evaluate guests more precisely after the conversation because they had a better understanding of the guest’s personality than they had before the conversation.
With-music condition
A two-way ANOVA with factors of the timing of impression rating (before vs. after conversation) and rating item (10) was conducted. The dependent variable was the values of the rating items. The interaction between the timing of impression rating (before vs. after conversation) and rating item (10) was not significant, F(9, 126) = .57, p = .821. The main effect of the timing of impression rating (before vs. after conversation) was significant, F(1, 14) = 7.03, p = .019. This means that impressions of guests improved after the conversation. In contrast to the no-music condition, in the with-music condition, guests received higher scores after the conversation on every rating item. This means that impressions of guests after the conversation improved on the whole. Statistical tests were then conducted for every rating item. As shown in Table 3(b), the amount of improvement was significant for ‘unpleasant–pleasant,’ F(1, 14) = 4.67, p = .049, and ‘would not like to date–would like to date,’ (F(1, 14) = 17.80, p < .001. In addition, the degree of improvement was marginally significant for ‘inaccessible–accessible,’ F(1, 14) = 3.56, p = .080, ‘careless–careful’ F(1, 14) = 3.89, p = .069, ‘unfriendly–friendly,’ F(1, 14) = 3.92, p = .068, and ‘hard to like–likeable,’ F(1, 14) = 3.65, p = .077. For the other rating terms (‘passive–active,’ ‘unsociable–sociable,’ ‘unassertive–confident,’ and ‘short-tempered–patient’), the effects of the timing of impression rating were not significant.
The results indicate that in the with-music condition, impressions of guests tended to change overall in a positive direction during the conversation.
Comparison between no-music and with-music conditions
A two-way ANOVA between the two conditions (no-music vs. with-music) with the factor of the timing of impression ratings (before vs. after conversation) was conducted for each rating item (see Table 4).
Comparison between the no-music condition and with-music condition. The results of a two-way ANOVA of every rating item.
Timing: before vs. after.
Participant: no-music vs. with-music.
p < .001; ** p < .01; * p < .05; † .05 ≤ p < 0.10.
A significant interaction was found for ‘passive–active,’ F(1, 29) = 5.69, p = .024, and for ‘would not like to date–would like to date,’ F(1, 29) = 6.32, p = .018. The main effects of participant (no-music vs. with-music) were not significant for the former but were significant for the latter, where participants in the with-music condition felt they might ‘like to date’ with the conversation partner more than participants in the no-music condition: before conversation, not significant; after conversation, F(1, 29) = 15.15, p = .001. Furthermore, for ‘hard to like–likeable,’ no significant interaction was found, and two main effects were significant: the timing of impression ratings (before vs. after conversation) was significant, F(1, 29) = 11.45, p = .002, and the effect of participant was significant F(1, 29) = 5.99, p = .021. The results indicate that the impression of guests improved after the conversation irrespective of the existence of music (for the first eight items, there were no significant differences between no-music and with-music conditions). However, participants with music rated their partners as more ‘likeable’ in terms of feeling attracted to them, and ‘would like to date’ them, than participants without music.
Discussion
Face-to-face conversation, in both the no-music condition and the with-music condition, had the effect of allowing participants to learn more about each other. As a result, impressions of their partner changed. The largest difference between no-music and with-music conditions was that music facilitated a feeling of attraction (‘likeable’ and ‘would like to date’) toward a partner of the opposite sex. It is widely known that music directly influences our emotions and appeals to feelings (e.g. Levitin, 2006), and it could be said that the effects of music extend beyond cognition to the senses. Furthermore, listening to music may make us feel not only relaxed but also more positive, boosting positive impressions of another person and evoking the feeling that we are attracted to him or her. As there was no unpleasant or neutral music condition in the current experiment, it is not possible to distinguish between the effect of music in general or that of pleasing music in particular. Further research is necessary to investigate the effect of favorite/neutral music on impressions of conversation partners.
Furthermore, in the current experiment, the guests (target persons) in the with-music condition also listened to the background music during the conversation. Therefore, the music may have not only influenced the participants’ perceptions but also the guests’ behavior. Although guests were instructed to behave as consistently as possible across sessions, it is highly likely that the music could have made them more social, happy or charming. In this case, participants’ impressions of the guests would be much improved. Another possibility was that music made the interaction between participants and guests more positive. Again, in this case, participants’ impressions of guests may have been much better than in the no-music condition. More specifically, favorite music may facilitate the formation of closer relationships during conversation (Rentfrow & Gosling, 2006), and as a result, the impression ratings may improve. Therefore, clarification is still required to confirm whether this change was caused by different interactions between guests and participants, by changes in guests’ attitudes, or by changes in participants’ judgments. As shown in Table 4, after conversation, impression ratings improved irrespective of the presence of music. However, at the same time, whether or not music caused the change of guests’ attitudes and the change of interaction between guests and participants, music increased the feeling of attraction participants had toward guests. This confirms the effects of background music on emotional behavior (Västfjäll, 2002) and supports the idea that music helps us take action (e.g. dating).
As found in earlier research, music influences our behavior and actions. For example, Guéguen, Jacob, Lourel, and Guellec (2007) reported that when popular music was played, clients stayed longer at an open-air market stall. Emotions are generated in the limbic system and music is known to activate this system (e.g. Brown, Martinez, & Parsons, 2004). Music causes a sense of comfort and influences our actions. Therefore, when we listen to our favorite music, we may be more predisposed to take action. The results shown in Table 4 support this hypothesis, although the effects of background music may be modest.
Our favorite music encourages us greatly and moves us emotionally. However, it is clear that in group activities such as a gokon party for konkatsu in Japan, background music is not always favored by the entire group, as individual preferences vary. Event organizers try to play music that is expected to meet the preference of the majority. The current research investigated this phenomenon under a controlled environment similar to a gokon party for konkatsu. Stronger effects of background music might have been obtained had we used music that was selected as personal favorites by the participants themselves rather than by another sample from the same university. In the current experiment, the participants did not select their favorite music; rather it was selected by other students. However, after the experiment, they were asked whether they liked the music, and 73% of participants in the with-music condition said that their favorite music was included in the background music pieces, while only 13% indicated that the study included music they disliked.
The results obtained here may be partly culture-dependent. Different results might be obtained if the experiment included young adults from cultural backgrounds that are more passionate in love relationships. Additionally, positive music might be more effective than favorite music.
Panksepp and Bernatzky (2002, p. 145) note that ‘the ability of music to promote social interactions is an evident aspect of music used in all societies.’ The present results go some way to demonstrate that music serves as a magnet for social activities.
Conclusion and limitations
The current results indicated that a person’s feelings of attraction toward a conversation partner may be heightened if he or she hears liked music during the conversation. However, this research has some limitations. First, whether the music effect obtained here is the effect of liked music or of music in general is difficult to determine. Second, this research examined group activities and selected music that met the preference of the majority of a previous sample. However, stronger effects of background music might have been obtained if we had used music that had been selected as a personal favorite by the participants. Third, it may be possible that music influenced the guests and the interactions between guests and participants. This might cause changes of impressions of the conversation partner. Fourth, the results obtained here might be culture-dependent.
Although the experiment was conducted under controlled laboratory conditions, and we can be relatively confident of its internal validity, future research with a broader range of participants is necessary to test the generalizability of the findings.
Footnotes
Author note
Part of the results of this experiment were published at the 77th Annual Convention of the Japanese Psychological Association and introduced on the website of the Sony Corporation.
Funding
This research was supported by VECTOR INC. and Sony Corporation.