Familiarity as a groove-inducing factor

Among various predictors of groove, familiarity has been identified as one of the most influential factors [1,2,6,7]. This may result from its facilitation of rewarding behaviours and its effect on enjoyment. For example, familiarity enhances our ability to anticipate a song’s rhythm and synchronize our movements to it, which humans find rewarding [8,9]. Additionally, the mere exposure effect finds that repeated exposure to a song can increase enjoyment of it [10], which in turn may enhance groove ratings, as enjoyment correlates with perceived groove [6]. Enjoyment may, in turn, increase listening frequency (i.e., familiarity), creating a positive feedback loop that further amplifies the groove experience.

Nostalgia as a Groove-Inducing Factor

Unlike familiarity, nostalgia’s effect on groove is unclear. Nostalgia involves a yearning for the past [11], and it is most strongly felt for memories formed during late adolescence and early adulthood [12,13]. This phenomenon is attributed to the ‘reminiscence bump’, in which memories from these formative years are encoded with greater intensity [14]. Despite inducing feelings of loss or pain, nostalgia remains a predominantly positive emotion [15]. Nostalgia activates reward-related areas [16] that are also implicated in groove perception [17]. Sensory inputs, particularly music, are primary triggers for nostalgia [13,18–20]. Research suggests that music-induced nostalgia is stronger when a song is more familiar [13], however, nostalgia and familiarity are distinct, with nostalgia sometimes triggered by elements reminiscent of the past, even if unfamiliar.

Movement types and groove experience

While studies have examined various factors that affect the urge to move, little work has systematically examined whether groove ratings are affected by considering specific types of movements. Movements like tapping your foot or hand along to the beat, generally moving along, or dancing, have been used to assess groove and are generally treated as equivalent metrics [21]. However, wanting to tap to a song may not be the same experience as wanting to dance, and asking about specific types of movements can affect groove ratings [22]. This distinction may be important as different musical characteristics can influence what parts of the body we move and can elicit different movement characteristics [23,24]. In general, emotions can influence bodily movements in unique spatial, temporal, and energetic ways [25] and this is also true of music-induced emotion [26]. The emotion attached to nostalgia may affect the type of movement one chooses to engage in. The urge to move to music is also related to rhythm [7] and rhythm perception involves interactions between the auditory and motor systems, including the premotor cortex, supplementary motor area, and basal ganglia [27–30], which can be activated even without actual movement. The experience of groove may be considered a precursor to music-induced movement, as high-groove stimuli have been shown to elicit spontaneous rhythmic movements [2]. By analyzing different movement types in relation to groove, we can investigate if these different movements index the same experience, or if they differ.

The present study explored whether groove ratings are affected by nostalgia and whether previously observed differences across movement descriptors (i.e., tap, move, dance) could be reproduced. Specifically, we assessed ratings related to the desire to tap, move, and dance. We hypothesized that 1) nostalgia would significantly predict groove ratings 2) groove ratings would differ across movement descriptors, with differences between tapping, moving, and dancing.

Participants

Participants were recruited using the Prolific Research database [31], and ranged in age from 23 to 28 years old. This age range was selected to ensure a defined participant group with a high likelihood of familiarity with the selected songs, as well as to align the nostalgic songs with participants’ reminiscence bump period, when music-evoked nostalgia is particularly strong [13,14]. Additionally, participants were required to either be born in North America or relocated there prior to age five to ensure they would be culturally familiar with the selected songs. Self-reported English fluency was also a requirement, and individuals with self-reported hearing impairments were excluded. For the data analysis, 102 participants were included (M = 26 years, SD = 1.8 years). This sample size was chosen to align with prior groove studies [17,32] and to provide sufficient power to detect medium-sized effects. Initially, 104 individuals signed up, but two did not complete the study. All participants provided informed written consent before participating. After completing the study, all participants received financial compensation. The non-medical research ethics board at Western University reviewed and approved the experiment.

Stimuli

The music stimuli were chosen based on a piloting procedure. The following were selected: the top ten songs from the Billboard charts during the participants’ adolescent period (2010–2015) and the top ten songs from more recent years (2018–2021), at the time of data collection (August 14–24, 2021). These release periods were selected based on the targeted age of the participants to try to create lists of nostalgic and non-nostalgic songs, respectively. Pilot participants (N = 4) in the targeted age range of 23–28 years rated these 100 songs for familiarity and nostalgia. Based off the ratings, 20 early release (nostalgic) songs and 20 late release (non-nostalgic) songs were selected. Both lists had similar overall familiarity ratings. Due to a technical error one of the non-nostalgic songs was played twice, resulting in ratings for only 19 of those songs. All music clips were 25 seconds in length and included the chorus to maximize recognizability. Participants evaluated these forty music excerpts using three 100-point scales gauging their desire to tap, move, and dance, as well as their liking, familiarity, and nostalgia (full song lists and averages for each rating can be found in S1 and S2 Tables).

Procedure

Participants completed a 45-minute Qualtrics survey consisting of a demographics evaluation, music/dance background assessment, song listening and rating, and debriefing. Within the demographics section, participants reported their age, gender, education level, and ethnicity. The music/dance background section queried the time dedicated to music listening, preferences for music conducive to movement, the completion of the Goldsmiths Dance Sophistication Index (DSI) [31], and the training subscale of The Goldsmiths Musical Sophistication Index (MSI) [33]. The DSI, comprising 26 items, employs a 7-point Likert scale to assess participatory dance experience (20 items) and observational dance experience (6 items), with the former being further separated into four factors: body awareness (6 items), social dancing (6 items), urge to dance (5 items), and dance training (3 items). In the end, only the urge to dance and dance training subscales were used from the DSI.

Within the listening and ratings section, participants were first given a sample music clip and asked to put on headphones and set the volume to a comfortable level. Then the descriptions of what participants were rating (desire to tap, desire to move, desire to dance, enjoyment, familiarity, and nostalgia), were provided along with an example rating scale (see S3 Table). Each scale ranged from 0–100. The definition given for nostalgia was ‘a sentimental longing or wistful affection for the past’ [34]. Afterward, the main task started and participants listened to and rated the 40 music clips in randomized order. Midway through the task, an attention check was included, in which participants were presented with only rating scales and asked to set all values to 100. After all ratings were completed, participants were given an open text box to write anything they thought would be important for the researchers to know (e.g., any technical issues) and were provided with a debriefing form.

Data analysis

Statistical analyses were conducted in Jeffrey’s Amazing Statistics Program (JASP Team, 2024). For each participant, averages were calculated for the three movement rating types (desire to tap, desire to move, desire to dance), liking, familiarity and nostalgia for the early and late songs. To assess the song type manipulation on the non-groove ratings a 3 (rating type: liking vs. familiarity vs. nostalgia) x 2 (release period: early vs. late) repeated measures ANOVA on the average values for each rating type was conducted. To determine if groove ratings differed across movement type and release period a 3 (movement rating type: desire to tap vs desire to move vs desire to dance) x 2 (release period: early vs late) repeated measures ANOVA on the average values for each rating type was used. To compare the size of the differences between early and late release periods for each rating type, difference scores were calculated for each of the six rating types, by subtracting the late release period average for each rating type from the early release period average. Two one-way repeated measures ANOVA on the difference scores were used to compare 1) the three non-groove ratings and 2) the three groove ratings. When necessary, Bonferroni-Holm post-hoc tests were completed. Additionally, if the assumption of sphericity was violated then Greenhouse-Geiser corrections were used.

Pearson correlations indicated associations between the six rating scales. Follow-up multiple linear regressions were conducted with the three movement rating types as outcome variables, with familiarity and nostalgia as potential predictors. Liking was not included as a predictor because pleasure is an inherent part of the groove experience, as it is the ‘pleasurable urge to move’. Groove and liking are often highly correlated [2,3,35] and including it as a predictor would likely dilute other effects. Additionally, Pearson correlations between the three movement rating types, the music training subscale, the dance training subscale, and the urge to dance subscale assessed relationships between music and dance experience and the different groove ratings.

The above analyses were conducted on participants’ average ratings across songs to understand the general relationships between the three movement types, nostalgia, and familiarity. To further explore the strength of these relationships, Pearson correlation values were also calculated within each participant for the ratings of the three movement types with nostalgia and familiarity across songs. The correlation coefficients for each of the six correlations (i.e., tap-nostalgia, move-nostalgia, dance-nostalgia, tap-familiarity, move-familiarity, dance-familiarity) for each participant were put into a 3 (movement rating type: desire to move vs desire to tap vs desire to dance) x 2 (familiarity vs nostalgia) repeated measures ANOVA on the individual correlation coefficients for the six correlations with Bonferroni-Holm post-hoc tests. Three participants were excluded from this analysis as they had constant ratings for one of the rating categories and therefore correlation coefficients could not be calculated.

Limitations and future directions

While this study provides insight into the relationship between nostalgia and groove, several limitations could be considered. First, the study focused exclusively on young adults (23–28 years old), limiting the generalizability of the findings to older populations. In particular, older populations would be of interest to study, as the sensation of nostalgia generally increases over the lifespan [39]. Future research could investigate whether different age groups exhibit varying movement responses to nostalgic music, as understanding the properties of music that enhance movement could inform applications of music to gait rehabilitation [40]. An additional limitation is that, while familiarity and nostalgia were measured separately, their measurement at similar points in time may have influenced ratings, making it difficult to fully disentangle their individual effects. Despite efforts to control for familiarity, the newer songs were still rated as less familiar than the older ones. As a result, additional analyses were necessary to account for familiarity, but the current study design cannot fully isolate nostalgia from familiarity. The study was also conducted remotely, meaning listening conditions could vary across participants and were not under experimental control, potentially affecting responses. Finally, as this study did not comprehensively quantify musical features across stimuli (e.g., differences in syncopation, micro timing, tempo range, etc.), the songs may differ in other characteristics that contribute to groove.

Further research is needed to determine whether nostalgia alone drives the observed effects, or if differences in familiarity, listening conditions, musical properties, and age contribute. Future studies could isolate nostalgia’s unique relationship to groove by controlling for these factors, as well as examining nostalgia’s specific role in the urge to dance across different populations.