Quantitative results
To compare the importance of intimacy, passion, and commitment across the three relationship phases (RQ1), we performed a repeated-measures ANOVA with aspect and phase as within-subject factors predicting the importance of music. This analysis revealed significant main effects for both aspect, F(1.39, 240) = 195.75, p < .001, ηp² = 0.53, and phase, F(1.62, 280) = 13.44, p < .001, ηp² = 0.07, and no interaction effect, F(3.56, 616) = 0.87, p = .475, ηp² = 0.01. Pairwise comparisons using Bonferroni correction indicated significant differences between all relationship aspects (p < .001). Regarding relationship phases, there were no differences between the attraction and building phases (p = 1.00), but there were differences between the maintenance phase and both the attraction and building phases (ps < .001). The perceived general role of music did not differ across relationship phases, F(1.85, 319.2) = 0.77, p = .462, ηp² < 0.01. As shown in Figure 1, participants reported a high general role of music in all three phases (M = 4.00, SD = 0.89). The importance of music for the three components was higher in the attraction (M = 3.61, SD = 1.08) and building (M = 3.59, SD = 1.02) phases than in the maintenance phase (M = 3.31, SD = 1.05), while it appeared to be most important for intimacy (M = 3.96, SD = 0.97), followed by passion (M = 3.76, SD = 0.99), and commitment (M = 2.79, SD = 1.18).
Figure 1. The importance of music across the three phases of romantic relationships.
Note. A combination of boxplots and violin plots is used to visualise the data distributions. Mean values are indicated by black circles. The importance of music was measured on a scale from 1 = not at all to 5 = totally.
To determine whether there are personal characteristics associated with the role of music in romantic relationships (RQ2), we calculated correlations between the role of music in general and for the three aspects aggregated across phases with the assessed personal characteristics: demographics (gender, age, education), musical expertise, and music reward experiences. As detailed in Table 1, of the characteristics assessed, music reward experiences were associated with all outcomes, whereas musical expertise was correlated with the general role of music.
Table 1. Correlations between the importance of music and assessed characteristics including means and standard deviations.
| Variable | M (SD) | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 |
|---|
| 1 Role of music overall | 4.00 (0.89) | | | | | | | | |
| 2 Importance of music for passion | 3.76 (0.99) | .73*** | | | | | | | |
| 3 Importance of music for intimacy | 3.96 (0.97) | .71*** | .88*** | | | | | | |
| 4 Importance of music for commitment | 2.79 (1.18) | .58*** | .63*** | .61*** | | | | | |
| 5 Musical expertise | 0.00 (0.83) | .24** | .11 | .08 | .14 | | | | |
| 6 BMRQ Total Score | 4.04 (0.46) | .50*** | .44*** | .40*** | .24** | .28*** | | | |
| 7 Gender::female | 0.83 (0.37) | .05 | .00 | .05 | -.02 | -.01 | .12 | | |
| 8 Age | 27.18 (9.27) | .07 | -.02 | .00 | .13 | .14 | -.14 | -.07 | |
| 9 Education | 4.77 (1.17) | -.04 | -.07 | -.06 | .04 | .14 | -.13 | .08 | .50*** |
In more detail, Table S1 in the Extended Data (see OSF repository) outlines the correlations between musical expertise and each BMRQ subscale across different relationship phases and their respective outcomes. This analysis suggests that musical expertise is specifically associated with higher importance of music in the attraction and building phases, but not in the maintenance phase. Moderate correlations were observed between all BMRQ scales and outcomes, with the BMRQ Social Reward subscale showing the strongest associations, while associations with the BMRQ Sensory Motor subscale appeared to be somewhat weaker.
In response to additional 5-point scale questions, participants indicated that they typically discussed their musical preferences with their (potential) partners (M = 4.28, SD = 0.86) and placed moderate importance on their partner's musical taste (M = 3.52, SD = 1.06), but less importance on their partner's musical ability (M = 2.19, SD = 1.14). On average, participants tended to have songs that they considered defining their relationships with (M = 3.70, SD = 1.37). Overall, the participants did not perceive musical compatibility as more important than compatibility in other leisure activities (M = 2.54, SD = 1.05). Consistent with findings regarding the role of music in romantic relationships, higher scores on these additional questions were associated with higher levels of musical expertise and musical reward (see Table S2 in the OSF repository for more details).
Qualitative results
Out of 174 participants, 129 provided responses on musical experiences in the attraction phase, 119 in the building phase and 103 in the maintenance phase, resulting in 351 coded responses. Qualitative template analysis, as described in the Methods section, yielded 55 distinct codes grouped into nine higher order themes (memories, listening to music, sharing and discussing music, active musical activities, attraction, compatibility, self-disclosure and communication, bonding and (re)connecting, emotion/affect, and regulation) and two aggregate dimensions (action and outcome), as shown in Figure 2.
Figure 2. Full list of codes and higher order themes.
Note. Negative valence codes are shown in grey. Code descriptions can be found in the extended data (see OSF repository).
Inter-coder reliability was substantial, with κ = 0.62, z = 104, p < .001 across all codes and phases. Agreement was similarly high for the attraction (κ = 0.62, z = 60.27, p < .001), building (κ = 0.60, z = 58.65, p < .001), and maintenance (κ = 0.63, z = 61.65, p < .001) phases. Tables S4 and S5 in the Extended Data (see OSF repository) further show that all Kappas for second-order themes were consistently above 0.40, indicating at least moderate agreement (M = 0.57, SD = 0.12, range = 0.40 - 0.65). Of the 55 first-order codes, 76% showed at least moderate agreement (> 0.40), and 96% achieved at least fair agreement (> 0.20) (M = 0.58, SD = 0.21, range = 0.12 - 1.00).
Given the high level of agreement between the coders, we present the qualitative responses based only on those codes on which both coders agreed, to increase the robustness of the findings. Consequently, only the codes that both coders identified as present in a participant's response were included in the analysis.
Aggregate dimensions: Action and Outcome. Regarding the two aggregate dimensions of action and outcome (the latter subdivided into bonding, signalling, and emotions), 72.3% of the identified codes corresponded to musical actions, which included activities such as listening to music, sharing music, and creating music. Bonding as an outcome, involving (re)connection and self-disclosure, was present in 12.6% of the responses, signalling as an outcome, indicating attraction or compatibility, in 11.8%, and emotion as an outcome, including (co)regulation of emotions in terms of setting the mood and processing emotional experiences, in 3.3%.
Across all relationship phases, musical actions show a consistent presence, accounting for 68.6%, 73.0%, and 75.7% in the attraction, building, and maintenance phases, respectively. Similarly, bonding was observed across all phases with 10.3%, 14.0% and 13.7% in the attraction, building and maintenance phases respectively. Signalling was most pronounced in the attraction phase (16.5%), followed by the building phase (9.6%), and maintenance phase (8.9%). Finally, emotions were less pronounced in all phases, with 4.7%, 3.4%, and 1.7% in the attraction, building, and maintenance phases, respectively.
Second-order themes. The heatmap in Figure 3 provides an overview of theme frequencies overall and across phases. Overall, the most frequent theme was listening to music in public and private contexts, followed by sharing and discussing music and engaging in active musical activities. We also observed variations in the prominence of themes across phases. For example, 'listening to music' increased in frequency from the attraction phase to the maintenance phase, while 'sharing and discussing music' was more common in the attraction and building phases and decreased in the maintenance phase. The theme of 'memories' appeared in all phases, but was most prevalent in the attraction and maintenance phases, where participants referred to creating shared memories early on and recalling them later. These participant responses include quotes such as, “Having that song as a reminder of all the good in the relationship and why you love the other person, having positive memories tied to that and other songs that are brought up when you listen to them[...]” (Participant 116), and
Figure 3. Frequencies of second-order themes across relationship phases.
“Certain songs discovered and consumed during that period (especially when they were relationship pop [songs]) connect with that time and reinforce the feeling of attraction. I've always felt that the songs I've consumed a lot tie in with important events in my life during the listening period, as well as the seasons.” (Participant 168)
Feelings of attraction through music were the most prevalent in the attraction phase, whereas references to compatibility through music were consistently present in all phases. The themes 'Self-disclosure and communication' and 'Emotion / affect & regulation' were more prominent in the early stages of the relationship. Regarding self-disclosure, participants detailed instances in which they revealed personal information to their partners, including their musical preferences and experiences, and used songs or song lyrics to communicate their feelings and intentions to their partners. Quotes exemplifying this theme include, “we shared playlists of music that we like. also we exchanged playlists with music in our native languages, as we are an international couple (finnish and spanish).” (Participant 17), “You can also ask which songs you listened to as a child and find out something about your family environment.” (Participant 326), Regarding emotion / affect & regulation, participants wrote about experiences in which music was used either intentionally or unintentionally to set the mood or affective tone while together or independently. Examples of this theme in participants’ responses include quotes such as,
Many participant responses in which this theme was observed discussed the importance of music before, during, and/or after sex. These participants’ responses shared ideas such as:
“I think that the choice of music is extremely important during sex. In my experience, the type of music has a significant effect on sex and therefore also determines whether it vibes. What I've also noticed personally in this context is that "more relaxed" and "romantic" music only feels good with people you really like. So music can also be an indicator of how comfortable I feel with a person and how vulnerable I can be.”(participant 238)
First-order codes. In the action dimension, three of the most frequent codes were related to listening to music: listening to music together (2F, n = 139, 15.3%), listening to music in public (2A, n = 71, 7.8%) and listening to music in private (2B, n = 48, 5.3%). In addition, sharing and talking about music was represented in 7.6% of the codes (n = 69), and making music together in 5.1% of the codes (n = 46). In the outcome dimension, only one code appeared in nearly five percent of the responses: compatible musical tastes (6A, n = 45, 4.96%). The second-most prevalent code was strengthening bonds/deepening connections (8A, n = 42, 4.63%).
Table 2 provides an overview of the top three codes in each relationship phase, separately for the action and outcome dimensions. Across all phases, the most common action codes consistently involve listening to music together, whereas the outcomes of musical activities vary according to the relationship phases. In the attraction and building phases, partners primarily focus on (in)compatibility, shared musical tastes, and deepening bonds through music. In the maintenance phase, coping with differences becomes more prominent, alongside topics such as incorporating music into future plans and commitments. Interestingly, negative experiences were more frequently mentioned in the building and maintaining phases, namely the realisation of incompatibility and loss of connection. Quotes from participants' responses exemplifying these negative experiences shared ideas such as,
Table 2. Most prevalent (top three) first-order codes across relationship phases.
| Code | Frequency | Percentage
(by phase) |
|---|
| Attraction Phase | | |
| Action | | |
| 2F - Together (music listening) | 40 | 12.4 |
| 3A - Sharing or talking about music | 38 | 11.8 |
| 2A - Public (music listening) | 17 | 5.3 |
| Outcome | | |
| 6A - Compatible musical tastes | 25 | 7.8 |
| 8A - Strengthening bonds / deepening connection | 17 | 5.3 |
| 5A - Attracted by musical ability & 5B - Attracted by shared taste | 8 | 2.5 |
| Building Phase | | |
| Action | | |
| 2F - Together (music listening) | 46 | 15.7 |
| 3A - Sharing or talking about music | 20 | 6.8 |
| 2A - Public (music listening) & 4A - Making music | 19 | 6.48 |
| Outcome | | |
| 8A - Strengthening bonds/deepening connection | 17 | 5.8 |
| 6A - Compatible musical tastes | 14 | 4.8 |
| 6E - Incompatibility | 8 | 2.7 |
| Maintaining Phase | | |
| Action | | |
| 2F - Together (music listening) | 53 | 18.2 |
| 2A - Public (music listening) | 35 | 12.0 |
| 2B - Private (music listening) | 22 | 7.53 |
| Outcome | | |
| 6D - Coping with differences | 12 | 4.11 |
| 8A - Strengthening bonds/deepening connection | 8 | 2.7 |
| 6E - Incompatibility, 8D - Future plans & commitment, & 8H - Losing connection | 7 | 2.4 |
“Different music preferences led to us growing apart (he never really wanted to listen to music, but it was important to me) - even with the songs we listened to, when we listened to music we became estranged (I think) because we only ever listened to his songs.” (Participant 370).
A more detailed heat of the frequency of all 55 first-order codes across the relationship phases is provided in Figure S1 in the Extended Data (see OSF repository).
Co-occurrence of codes. To explore whether certain action and outcome codes were frequently mentioned together, we conducted a thematic co-occurrence analysis following the method outlined by Scharp (2021). As shown in Figure 4, outcome codes belonging to compatibility (Theme 6) and bonding (Theme 8) mostly co-occurred with action codes. Of the compatibility codes, compatible tastes (6A) showed the most co-occurrences, specifically listening to music codes (2F together, n = 19; 2C active, n = 7; 2A in public, n = 7), sharing music (3A, n = 15), and actively making music (4A, n = 5). Participant responses illustrating these co-occurrences included descriptions of actions such as, “Discussing musical preferences, and finding common interests, setting up a date location based on mutual musical interests (e.g. jazz bar), meeting up at a festival.” (Participant 110). The codes belonging to bonding (Theme 8), specifically strengthening bonds (8A), mostly co-occurred with listening to music (2F, n = 8) and sharing music (3A, n = 8). Responses such as, “Going to concerts together and generally listening to songs that both partners like was a nice experience. I feel more connected to my partner when he has at least some idea of what music I like [...].” (Participant 245- 8A and 2F) demonstrate the co-occurrence of these codes.
Figure 4. Co-occurrences of action and outcome codes across all relationship phases.
Note. Circle sizes indicate the frequency of individual codes, and line thicknesses indicate how often the two codes were mentioned in the same responses.
A full matrix of co-occurrences overall and separately by relationship phase can be found in the Extended Data (see OSF repository).
