Abstract
There is limited empirical research devoted to exploring children's subjective responses to music, despite extensive examination in the literature of their ability to recognize emotions said to be expressed by music. To address this gap, we recruited 26 participants aged between 5 and 11 years with the aim of learning more about their capacity to report on their music-evoked experiences, and to test the suitability of a categorical measurement scale as an age-appropriate self-report format. Participants were invited to give free descriptions of how six musical stimuli made them feel and to rate their music-evoked experiences using the categorical scale. Their responses revealed an awareness of music-induced feelings, visual imagery, and imagined narrative, and the capacity to describe these in detail. All participants were able to use the categorical measurement scale; however, the richness of their free responses suggests that an open format might capture more comprehensive and meaningful data than a forced-choice format. We recommend further testing with a much larger sample of participants from a wider population.
Introduction
There is limited empirical exploration of young children's lived experiences with music. Of note is Herbert and Dibben's (2018) focus on the responses of older children and adolescents, and earlier interest in children's musical or aesthetic awareness (e.g., Parsons et al., 1978; Rodriguez & Webster, 1997) and their ability to recognize basic emotions purportedly expressed by music (as reviewed by Trehub et al., 2010). Infants have demonstrated affective responses to music influenced by its style and genre, notably lullabies or play songs (e.g., Cirelli et al., 2020; Kragness et al., 2023). There remains scope to explore how music is experienced subjectively by young children and to trace the development of their responses ontogenetically.
To support future longitudinal studies of children's music-evoked experiences, the aim of the present study is to establish a preliminary response format suitable for children of UK primary school age (five to 11 years). Studying this specific population's responses to music extends the work of Herbert and Dibben (2018) to some of the youngest listeners who are old enough to express themselves verbally. The small overlap with Herbert and Dibben's target population concerns children at a critical developmental stage as they are approaching puberty. Many UK schoolchildren also enter secondary education during the academic year in which they reach their 12th birthday. In the absence of a standardized format for recording children's verbal reports of their subjective responses to music, such as a pre-validated, age-appropriate music-specific psychometric instrument, we initially sought to adapt a categorical rating scale, a format commonly used in adult-oriented music-emotion research (e.g., Zentner et al., 2008). In order to capture music-induced experiences beyond those included in the categorical scale, as well as to learn more about children's capacity to report on these, we also asked children to give free descriptions of how six musical stimuli made them feel.
Gabrielsson (2002) highlighted the need in empirical research to distinguish between felt experiences induced by music and emotions that listeners recognize as being expressed by music—described elsewhere as internal (induced) or external (recognized) “locus of emotion” (Schubert, 2013). Gabrielsson (2002, p. 124) acknowledged, however, that distinguishing felt from perceived emotion might not be straightforward: In reality, the border between the two alternatives is somewhat blurred. We may think of them as opposite extremes on a continuum from “pure”, emotion-free perception at the one end to intense emotional reaction at the other end. Rather than being at any of these extremes, in most situations listeners are probably somewhere along this continuum, depending on many circumstances.
Schubert (2013, p. 4) agreed that “the differentiation between internal and external locus can be blurred, and… may not even be distinguishable or meaningful to some.” Several studies indicate that emotions perceived and felt in music often correlate, for example via contagion (Juslin & Västfjäll, 2008), but that felt responses tend to be more subtle than perceived responses (see Evans & Schubert, 2008; Hunter et al., 2010; Kallinen & Ravaja, 2006), supporting Gabrielsson's concept of a spectrum of response from perception to feeling. Individual differences in awareness of feelings and ability to express them (granularity) may also impact how listeners report their responses (e.g., Barrett, 2004). In designing our study and interpreting its results, we needed to be mindful of the potential for variability in children's interpretation of what we were asking them to describe.
What Does Music Really Make Us Feel?
Debate endures as to the specific nature of music-evoked feelings. Whether these may be considered only broadly in terms of their arousal and valence; whether some qualify as discrete emotions—and whether those constitute basic, complex, or blended emotions—or whether subjective responses to music are emotionally relevant at all (Cespedes-Guevara & Eerola, 2018; Cowen et al., 2020; Eerola & Vuoskoski, 2011) continues to provoke scholarly discussion. Nonetheless, associations of music and idiosyncratic feeling date back to Plato (ca. 375 B.C.E./1997) and have emerged cross-culturally (Spitzer, 2021). Listener-reported music-evoked feelings widely include happiness, joy, a desire or need to move, relaxation, calmness, soothing, peace, power, inspiration, motivation, excitement, chills, wonder, fear, tension, mystery, sadness, nostalgia, reflection, and loss (e.g., Asmus, 1985; Campbell, 1942; Coutinho & Scherer, 2017; Cowen et al., 2020; Farnsworth, 1954; Hevner, 1936; Schubert, 2003; Zentner et al., 2008), as well as specific bodily sensations (Putkinen et al., 2024). In addition to feelings, listener responses include music-induced visual imagery, narrative engagement, and episodic memories, among others (e.g., Margulis et al., 2022; McAuley et al., 2021).
Empirical research exploring the variety and diversity of children's subjectively felt responses to music is relatively scant, compared with the wealth of studies examining their ability to recognize or discriminate emotions said to be expressed by music, typically from a choice of discrete basic emotions (such as happiness, sadness, or fear) where “correct” responses are determined by researchers or a panel of experts (e.g., Adachi & Trehub, 1998; Cunningham & Sterling, 1988; Dalla Bella et al., 2001; Dolgin & Adelson, 1990; Gerardi & Gerken, 1995; Gregory et al., 1996; Morton & Trehub, 2007; Nawrot, 2003, and see Trehub et al., 2010 for a review). Testing children's music-emotion recognition might reveal useful information about their enculturation; however, understanding the full scope of experiences that music may induce in young listeners offers a potentially important contribution to the growing body of research interested in the bio-cultural evolution of musicality—the human capacity for engaging with and appreciating music (e.g., Honing et al., 2015; Tomlinson, 2015; van der Schyff & Schiavio, 2017) and may be of relevance in the fields of music therapy and education.
The present study therefore begins with a focus on the felt component of children's responses to music (Scherer, 2005). Carvalho and Damasio (2021), along with Colombetti (2014), have underscored the significance of bodily feelings within the process of meaning-making. Given that music's meaningfulness is widely acknowledged (e.g., Cross, 2014), it could be argued that that emotionality and bodily engagement with music may be fundamental aspects of musical experience and sense-making (van der Schyff et al., 2022). When Herbert and Dibben (2018) asked 10- to 18-year-olds about the meanings they attached to music, many described emotions and other feelings. Learning about how young children describe their music-evoked experiences and how these evolve ontogenetically might offer additional insight into why music has such ubiquitous appeal and carries meaning for many people.
Measuring Musical Feelings
Subjective responses to music are customarily analyzed from listener self-reports (Eerola & Vuoskoski, 2013; Hevner, 1936). In this context, physiological measurements have become possible in recent decades (e.g., Koelsch et al., 2006), but idiosyncratic neurological responses have yet to be associated with specific feelings (Clore & Ortony, 2013; Janowski & Chelkowska-Zacharewicz, 2019; Khalfa et al., 2002; Russell, 2009). Further, there may be considerable within-participant variability in physiological responses associated with reported feelings, depending on their context (Hoemann et al., 2020). Barrett (2022, p. 903) dismisses the notion of discrete, universally experienced Platonic essences, proposing that contextual factors “actually play a fully causal role in creating an instance of emotion or any mental event”. In this sense, any feeling concept is the product of a complex web of interrelated elements, from an individual's personality, mood, preferences, or level of engagement, to environmental or situational factors. Investigations of music-induced experience must therefore consider—and ideally measure—a myriad of potentially influencing factors. In music-evoked experiences, Schiavio and colleagues (2017) argue that subjectivity, personal agency, and engagement are fundamentally important factors, and highlight the need for researchers to consider the complexities of responses to music holistically. Self-reports reveal how individuals interpret their feelings in a given situation or construe meaning from them, which is of central interest in the present study.
In music-emotion research, many adult-directed self-response tools are designed with a view to allowing quantitative comparative data analysis. Adjective checklists, from which listeners select those that most closely describe their music-induced feelings (e.g., Campbell, 1942; Farnsworth, 1954; Hevner, 1935, 1936; Schubert, 2003) and measurement scales, which additionally include a feeling intensity rating (e.g., Coutinho & Scherer, 2017; Cowen et al., 2020; Zentner et al., 2008), require participants to categorize their responses from a range of experimenter-selected options. A potential limitation of this forced-choice approach is that how closely listeners identify with the experimenters' response suggestions might not adequately reflect their authentic experience (Cespedes-Guevara & Eerola, 2018; Clore et al., 1987; Gendron et al., 2018; Scherer, 2005). Dimensional models, alternatively, allow participants to evaluate their affective experience in terms of valence (pleasantness–unpleasantness) and arousal (high–low activation) (Asmus, 1985; Edmonston, 1966; Osgood et al., 1957; Wedin, 1972; Russell, 1980; Schimmack & Grob, 2000; Schubert, 1999; Watson & Tellegen, 1985), but not to describe specific feelings (Clore et al., 1987).
Given the availability of several recently developed measurement scales, along with the advantage that these allow participants to rate the intensity of their experience, the format appeared to be a good candidate for age-appropriate adaptation. Before testing the suitability of a music-specific categorical measurement scale for children, we explored existing tools designed to capture children's responses to non-musical stimuli, as well as some designed to test their music-emotion recognition.
Words or Pictures?
We first considered whether words or pictures would provide a more suitable response format for our target population. Pictorial response tools, featuring photographs or cartoon images of facial expressions of emotion, have been used in various child-directed emotion-recognition studies. Participants select which facial expression best matches the music's putative character, most commonly from a choice of basic emotions1 such as sadness, fear, and anger (e.g., Dalla Bella et al., 2001; Kastner & Crowder, 1990). However, the use of facial expressions as emotion stimuli may be problematic, since these are known to vary among individuals and across situations and cultures (Barrett et al., 2019; Cordaro et al., 2018), and may elicit a variety of inter-individual interpretations (Nelson & Russell, 2013). Children's recognition of facial expressions of emotion is known to develop at different rates (Thomas et al., 2007; Widen & Russell, 2003, 2010). Studies of recognition of facial expressions (e.g., Ekman et al., 1969) have not produced consistent results (Prinz, 2004) and Vicari et al. (2000) found recognition of facial emotional expression continuing to develop throughout primary school years.
There is evidence to suggest that words may provide a more reliable means for children to categorize their music-induced experience than pictures: according to Bretherton and Beeghly (1982), children are generally able to use verbal emotion labels by their third birthday. Widen and Russell (2004) found emotion labels the stronger cue for four-year-olds over facial expressions and stories describing the causes and consequences of an emotion, and later reported a “story superiority effect” whereby young children were more likely to correctly label a target emotion from a story of cause and consequence than from a facial expression (Widen et al., 2015).
On the basis that our target participants are likely to be verbally articulate, and given the potential ambiguity of pictures, we concluded that a word-based instrument would provide the more appropriate means to record children's music-evoked responses. However, the choice of terminology for the categorical scale may yet present limitations, according to some authors.
The Language of Feeling
Barrett (2004), among others, has highlighted the importance of language in subjective experience, arguing that the nature of experience is influenced by the vocabulary an individual possesses to describe it. Barrett coined the term “emotional granularity” to describe the level of nuance with which people are able to interpret and describe their feelings, highlighting the influence of individual differences in emotional articulacy on results in studies of subjective response. According to Barrett, we generate emotion concepts from sensory input based on life experience: no single emotion concept is common to all individuals; instead, emotions may be seen as interpretations of sensory input, unique to one's personal experience, culture, education, and language. In this view, a person's vernacular informs not only the expression of feelings but the feeling concept itself. On this basis, asking children to categorize their responses to music verbally requires an understanding of how children's language and emotion concepts develop and shape each other (Baron-Cohen et al., 2010).
With a view to developing a categorical scale, we considered both inductive and deductive methods (Boateng et al., 2018; Hinkin, 1995): eliciting a selection of words from the children themselves with which to construct a measurement tool from scratch; or adapting an existing format, such as simplifying a pre-validated adult scale by reducing the number of categories or modifying the language to suit our target population. Following recommended best practice (Boateng et al., 2018), we incorporated both methods by testing a simplified reduction of three scales designed for adults: the Geneva Music-Induced Affect Checklist (GEMIAC) (Coutinho & Scherer, 2017), Geneva Emotional Music Scales (GEMS) (Zentner et al., 2008), and the 13 dimensions of affect (Cowen et al., 2020); additionally asking children to give free responses to the stimuli to elicit additional or alternative terms. However, we felt it important to keep an open mind as regards putting all our response format eggs into the categorical scale basket, given certain limitations associated with emotion vocabulary.
Limitations of Emotion Vocabulary
Emotion concepts are linguistically and culturally informed (Barrett, 2006; Prinz, 2004) and the language of emotion is in constant flux (Warner, 2021). Some languages have many synonyms for some feelings (e.g., scared, afraid, terrified), whereas some concepts exist only in certain languages, such as the Japanese amae2 (Clore & Ortony, 2013). Language may influence the interpretation of emotion in facial expressions: pairing an ambiguous facial expression with an emotion word is likely to bias an observer's interpretation of the facial expression (Halberstadt & Niedenthal, 2001). In this view, experimenter-selected terms might guide participants to produce responses that are not necessarily authentic. Barrett (2017a) argues that language constrains meaning: the process of organizing emotion concepts into categories effectively reduces idiosyncratic experiences to generic prototypes. In Barrett's view, reductionist experimental methods such as the forced-choice paradigm “strongly constrain the emotional meanings that participants could experience” (Barrett, 2017a p. 897). The implication is that attempting to describe music-evoked experiences within a small selection of categorical terms risks ignoring important detail or nuance.
There are increasing calls to heed the full variety of listeners’ subjective experiences in order to fully understand responses to music (Barrett, 2022; Cowen et al., 2020). Gabrielsson (2010, p. 571), for example, advocates including a “free phenomenological report”3 when recording music-evoked responses. And Gendron et al. (2018) call for methodological diversity in exploratory research, to include unconstrained tasks that “provide a context for discovery” (p. 213). Following these insights, participants’ own words might reveal important data that a categorical scale could not capture. We therefore felt not only that our exploratory study would benefit from the inclusion of a free response option to inform further development of the categorical scale, but also that including an open question within our eventual response format might help to elicit richer data than a traditional ratings scale alone could afford.
With a view to developing a tool capable of capturing optimally nuanced and authentic responses, we invited children to describe feelings elicited by music and stories (as a control) in their own vernacular to explore the usefulness of a free response option in addition to testing a categorical measurement scale using six experimenter-adapted emotion terms.
This article reports on an empirical study in the pursuit of understanding how emotion vocabulary is used by children of different ages, and to what extent children are aware of and able to describe music-induced feelings. In what follows, we first present our study, and then report its findings and their implications for the development of an age-appropriate response instrument.
Methods
Participants
Participants were recruited through an email invitation sent to families by the headteacher. Parents or carers gave their written consent for their children to take part and completed a brief demographic survey for each child, detailing age, school year, languages spoken, and any hearing impairment or specific learning need (for which the interview procedure might need to be adapted). Children's and their family members' musical background and engagement were assessed using questions from Module 1 of the Music Use and Background Questionnaire (MUSEBAQ) (Chin et al., 2018). Their participation in musical activity was recorded as (5) most days; (4) weekly; (3) monthly; (2) occasionally; or (1) never.
Participants were 26 children (18 females) aged between five and 11 years (M = 7.88, SD = 1.604) who attended primary school in Berkshire, UK. Five children were in Year 1,4 seven in Year 2, one in Year 3, eight in Year 4, one in Year 5, and four in Year 6. All spoke fluent English; five spoke two languages fluently and two spoke three. In addition to English, languages spoken included French, Spanish, Italian, Turkish, Urdu, and Cantonese. For two participants, English was their second language. In terms of their musical engagement, 26.9% played a musical instrument or sang most days; 61.5% played or sang at least once a week, and 3.8% played or sang occasionally. Additionally, 15.3% of participants had a family member who was a professional musician; 3.8% had at least one family member who engaged in amateur music-making, with 11.5% taking part in musical activity most days. All families listened to music at home, with 76.9% listening most days, while 26.9% of families reported attending concerts regularly.
Stimuli: Two Tasks
We presented participants with two tasks: an open question (How does this music make you feel?) requiring a free response, and a categorization task, requiring them to select one or more categories from an age-adapted scale that best described their music-induced experiences.
Adapting the Categorical Measurement Scale
For the categorical scale, we adapted six emotion terms from three pre-validated scales designed for adult participants (Coutinho & Scherer, 2017; Cowen et al., 2020; Zentner et al., 2008). Contrasting terms relating to valence (happy, sad) and arousal (tense, relaxed), with the addition of one motivating (powerful) and one reflective concept (longing for something in the past), were chosen from among the most commonly featured emotion categories in the pre-validated scales. Tense was chosen to represent feelings of both fear and anger, since previous research has shown that young children may confuse the two (Nawrot, 2003; Robazza et al., 1994; Terwogt & van Grinsven, 1991; and see Cespedes-Guevara & Eerola, 2018 for a summary). A further reason for including tense is the important role that musical tension and release are considered to play in emotional response (e.g., Huron, 2006). Other terms common to pre-validated scales (such as tenderness and transcendence) were omitted to keep the initial listening experiment of a manageable duration for the children; however, we anticipated that the free response component of the study could serve to reveal any important terminology that the categorical scale had not captured, as well as any additional types of subjective experience, and that these could inform further development of the scale.
Table 1 shows the emotion categories chosen for this experiment alongside the three emotion scales used with adult participants: GEMS (Zentner et al., 2008), GEMIAC (Coutinho & Scherer, 2017), and the 13 dimensions of affect (Cowen et al., 2020).
Comparison of emotion categories in measurement scales.
Stimuli: Stories and Music
The stimuli for this experiment were (1) six stories, each describing a character experiencing one of the six emotions; and (2) 12 musical stimuli, two hypothetically evocative of each emotion category. Our rationale for using two different musical extracts ostensibly evocative of each feeling category is that a wider range of extracts might produce a greater variety of free responses overall and reveal richer data from which we could develop our scale. Whereas we expected the feelings experienced by the story characters to be relatively unambiguous, a single piece of music would be unlikely to convey or induce identical intersubjective responses, given music's ineffability, or “floating intentionality” (Cross, 2001, p. 38).
Stories
We used short stories to test children's understanding of the categorical terms. According to the theory of embodied simulation, imagining how a story character feels and how music makes us feel might recruit similar neural and bodily processes (Decety & Grèzes, 2006; Ruby & Decety, 2004). The theory holds that inferring other people's mental states and processes involves simulating these through our own cognitive system. Gallese (2011, p. 197) describes understanding another person's feelings as “a pre-rational, non-introspective process… generating a physical, and not simply ‘mental,’ experience of the mind, motor intentions, emotions, sensations, and lived experiences of other people, even when narrated.”
Six vignettes were composed or adapted, each describing a character experiencing one of the six emotions (Table 2). Four were based on stories used by Widen and Russell (2010) as examples of characters experiencing happiness, sadness, relaxation, and tension; an extract from Harry Potter and the Goblet of Fire (Rowling, 2000) was chosen as exemplifying a sense of power as Harry wins the Triwizard Cup; and another, conceived by the first author, describes a girl remembering her holiday with fondness and longing. Audio recordings were made of each, read by a 10-year-old actor, with the intention that these should appeal to and feel relevant to young participants.
Story stimuli.
Music
The 12 extracts included nine excerpts of film soundtracks pre-validated as evocative of happiness, sadness, tenderness, relaxation, and fear (Ali & Peynircioglu, 2006; Eerola & Vuoskoski, 2011; Warrenburg, 2021). Three additional extracts of film music were selected by the experimenters: two sections of the track Remember (Zimmer, 1994) which accompanies a nostalgic scene in The Lion King (Allers & Minkoff, 1994) where the main character remembers his childhood and his long-lost father; and the opening title track (Williams, 1992) from JFK (Stone, 1991), featuring a military-style solo trumpet fanfare accompanied by snare drum tattoos, described by Broxton (2021) as “patriotic Americana”. The stimuli were selected as likely to convey one or more of the emotions in our categorical scale: the pre-validated extracts ostensibly happiness, sadness, relaxation, tension, and power, and the experimenter-added extracts longing and power. It should be noted that here we were not seeking to retest what these extracts conveyed; rather, they were chosen as good examples of emotionally charged music and therefore suitable means to investigate whether children would describe subjective experiences induced by hearing them. We were also mindful that any subjective experiences induced in listeners by music may not necessarily match the quality or intensity of feeling they believe the music expresses (e.g., Gabrielsson, 2002). The variety chosen was also with a view to assessing whether children distinguished a different quality and intensity of subjective experience on hearing each one. Detailed information about each extract is given in Table 3.
Musical stimuli.
Experimenter-selected
Eerola and Vuoskoski’s rationale for using film soundtracks was that film music “is composed for the purpose of mediating powerful emotional cues, and could serve as a relatively ‘neutral’ stimulus material in terms of musical preferences and familiarity” (p. 23). In selecting their examples, Eerola and Vuoskoski (2011, p. 23) avoided music with lyrics, dialogue, or sound effects that might influence listeners’ responses. The duration of the excerpts varies somewhat, “depending on the natural phrasing of the excerpt. … [to] convey the target emotion to the general listener in an optimal way”. We used the same criteria in selecting our additional stimuli: excerpts of instrumental film music cut at natural phrase boundaries.
Procedure
The study was conducted in person inschool by the first author. Each participant was tested individually, following the same procedure. For each participant, the experimenter-selected words were displayed on a tray, which was kept out of sight until the participant had heard each extract. The stimuli were presented in Qualtrics XM, using the experimenter's laptop. First, participants heard the six stories and were asked to describe how they thought the character in each one would feel. They were then asked to choose any of the experimenter-selected terms they thought matched this feeling (and equally to indicate if none did). This was to help us establish whether participants understood the meaning of the terms we had selected; this being the case, we could be reasonably confident that they would be able to apply the terms in other contexts, such as describing musical feelings. After hearing the stories, participants heard the musical extracts. All participants heard the same six stories and one group of six musical extracts (either version A or B). The test took place in a quiet classroom away from other lessons.
The experimenter greeted the participant and engaged in a brief informal conversationto ensure they were comfortable and happy to proceed with the listening experiment. The experimenter introduced and explained the procedure as follows: Have you ever heard some music that made you feel really happy or excited or sent shivers down your spine, or made you remember something special? Today we are going to hear some little pieces of music. I’d like you to tell me how the music makes you feel. We’ll also hear some stories and I’ll ask you how you think the characters in the stories feel. Is that OK?
The process was repeated for the musical extracts, which were also presented in random order. For each extract, the participant was asked, “How does this music make you feel?” and their free response and intensity rating were recorded. The tray of emotion words was presented again, and the participant was asked, “Here are the words again, which might describe how music makes us feel. Which do you think goes well? You can choose more than one.” For each word selected by the participant, the experimenter asked, “Does the music make you feel that a lot, or just a bit?” The experimenter recorded the participant's response against each emotion word as (2) a lot or (1) a bit. For words that the participant did not select, (0) not at all was recorded. At the end of the experiment, the participant was thanked for taking part.
Results
To test whether participants’ emotion categorization was significantly different depending on the stimuli heard, a cross-tabulation was conducted in SPSS with the extract type as the independent variable and the identification of the target emotion as the dependent variable.
Story Stimuli
Table 4 shows the percentage of participants who chose the target emotion for each story. These results demonstrate that participants identified the target emotion well above chance for all emotion categories (100/6 = 16.66%), indicating that the children taking part in our study largely understood the emotion category words. On this basis, we can be reasonably confident that they could potentially use them in the context of describing music-induced feelings. However, the fact that some children did not choose the target word as the best fit for some scenarios is grounds for revisiting our choice of terminology, as will be discussed later. It should be noted that, during the experiment, participants under eight years of age reported that they did not know the word tense; nevertheless, following explanation of its meaning, several selected tense in the categorization task.
Cross-tabulation: Percentage of participants who selected target emotion for the story stimuli (n = 26).
Given our small sample, we conducted a Fisher's exact test to determine any association between the stimuli and participants’ responses. The results (p < .001) indicate a significant association between the stories and the children's selection of the target category. As can be seen in Table 5, some stories elicited a more reliable identification of the target emotion than others. To understand these differences, we reviewed the categories chosen by participants who did not select the target emotion.
Percentage of participants who selected each emotion category for each story.
Shaded boxes represent target emotion. The highest percentage for each category is in
Table 5 is a confusion matrix showing the percentage of participants who selected each emotion category for each story. The target emotions are shaded. Participants were allowed to select one or more categories for each story.
The confusion matrix shows that all our participants chose the target emotion for the happy and relaxed stories. A large proportion (61%) additionally chose the happy category in relation to the “relaxed” story (some commenting that the protagonist would feel happy because they were relaxed). The majority of participants identified the target emotion for all stories; however, more than half chose happy for the “onging” story and 38.5% chose sad; more than half the sample chose tense for the “sad” story. The “tense” story elicited the lowest percentage categorization of the target emotion. Most importantly, the target emotion was the most frequently selected emotion for all story stimuli, strongly suggesting that most participants understood the six emotion terms. Next, we examined how well participants could apply these to the musical stimuli.
Whereas identifying a target emotion in the stories was important for us to establish whether participants had understood the words, their choice of emotion category for the musical stimuli is not a question of identifying a correct target but rather an indication of whether they distinguished differences in how each piece made them feel—in other words, that their choice of words was not random but related to their music-evoked experience. Nonetheless, any correlation between the children's felt responses and the emotional expression defined by Eerola and Vuoskoski's (2011) participants in Eerola & Vuoskoski (2011) (and us, in the case of the “powerful” and “longing” extracts) could be interesting to examine.
Musical Stimuli
Table 6 shows the overall percentage of children who said they experienced each emotion for the musical stimuli. The highest percentage for each category is highlighted in bold. Separate results for stimuli versions A and B (Tables S1 and S2 in the supplementary material) can be seen in Appendix A in the supplementary material.
Percentage of participants (n = 26) who chose each term to describe their experience induced by each musical stimulus.
The highest percentage for each category is in
A Fisher's exact test revealed a significant association (p = .001) between the musical stimuli and the categories that participants said best described their music-induced experience, indicating that they responded differently depending on what they heard—in other words, that they distinguished differences in how different pieces of music made them feel.
Estimated marginal means (Figure 1) indicate differences in children's emotion intensity ratings, depending on the music they heard. As Figure 1 illustrates, the highest mean intensity ratings for both versions A and B of the “happy”, “powerful”, “relaxed”, “sad”, and “tense” music are within the corresponding emotion categories. The highest mean intensity rating for the “longing” music is within the happy and relaxed categories, also for both the A and B stimuli.
(a) Estimated marginal means of participants' emotion intensity ratings for Music A stimuli. (b) Estimated marginal means of participants' emotion intensity ratings for Music B stimuli.
We conducted a repeated-measures analysis of variance (ANOVA) to examine differences in participants’ responses to the individual stimuli. Mauchly's test indicated that the assumption of sphericity had been violated for both the Music A stimuli, Χ2(14) = 33.701, p = .002 (ε = 0.618), and Music B stimuli, Χ2(14) = 49.568, p < .001 (ε = 0.413); therefore, Greenhouse–Geisser corrected results are reported. The results revealed a statistically significant difference in the quality and intensity of response between at least two Music A stimuli, F(4.215, 303.480) = 5.337, p < .001, partial η2 = .069, with a significant interaction between stimulus and response category intensity rating, F(21.075, 303.480) = 3.797, p < .001, partial η2 = .209; and a significant difference between at least two Music B stimuli, F(4.092, 294.602) = 5.527 p < .001, partial η2 = .071, with a significant interaction between stimulus and response category intensity rating, F(20.458, 294,602) = 5.570 p < .001, partial η2 = .279, indicating that participants distinguished some variation in the quality and intensity of their subjective response, depending on the stimuli. The means and standard deviations for the intensity ratings are given in Table S3 in the supplementary material and the mode intensity ratings are given in Table S4 in the supplementary material.
Post hoc pairwise comparisons for the Music A stimuli revealed significantly different scores between ratings of happiness and longing, p = .015; longing and power, p < .001; and longing and relaxation, p < .001; and for the Music B stimuli between ratings of happiness and longing, p < .001; longing and power, p < .001; longing and relaxation, p = .003; longing and sadness, p = .007; and longing and tension, p = .009.
Figure S1 in the supplementary material presents mean intensity ratings for each of the words that children selected from our proposed scale to best match their music-induced feelings for each extract. Interestingly, for most of the musical extracts, the children's highest mean responses align with the pre-validated emotion category (Eerola & Vuoskoski, 2011), as do those for the “powerful” music we selected ourselves. The “longing” music, which had not been pre-validated, evoked more happiness, relaxation, and sadness than longing in the children's responses. These mean ratings do not tell us any details about how children of different ages rated the music; however, developmental differences will be discussed later. For a view of participants’ perspectives expressed in their own vernacular, we can turn to their free responses.
Free Responses
We invited participants to give free responses to the stimuli to reveal vocabulary they would choose unprompted. “Scared” was a common response to the “tense” stimuli. In the context of the “longing” stimuli, “wanting to go back” and “wishing they could go back there” were used in relation to the stories, and “thinking back” in relation to the music. Several children recognized longing as a mixed, bittersweet, emotion, with feelings of both happiness and sadness. One child described longing Music A as, “Like a happy ending. Also sad, like they’ve lost something to get there.” Another child commented, “It's just amazing how music does that—makes you feel different things at the same time.”
Table 7 shows the most commonly occurring words in free responses to each of the musical stimuli. For the “happy”, “powerful”, “relaxed”, and “sad” extracts, participants used either our proposed term or a synonym most often. The “longing" extracts most often evoked a “relaxed”, followed by “sad” response, and only fourth-most often “longing", “making me remember", or “looking back”. The powerful extracts evoked second-most often “happy", “proud", or “excited” responses. For the “tense” extracts, children responded with “scared” most often, and “tense” second-most often. The complete list of free responses to each musical stimulus (Tables S5 to S16 in the supplementary material) is given in Appendix A in the supplementary material.
Descriptive terms used by participants.
We can see that many of the responses are very similar to our proposed emotion words, suggesting that our choices were age-appropriate. However, we need at this point to acknowledge that the procedure we used—asking participants to listen to each extract, give a free response and then use the rating scale before repeating this for the next extract—meant that they became familiar with the rating scale terms during the course of the experiment. We cannot therefore rule out the possibility that those terms might have influenced their subsequent responses. Nevertheless, conceptually different vocabulary did emerge in the free responses, and this included amazed, brave, excited, heroic, proud, royal, invincible, something about to happen, triumphant, and unstoppable.
Through thematic analysis, we found that the free responses could be grouped into 13 conceptual categories (Table 8), largely reflective of categories that have emerged in research with adult listeners (e.g., Cowen et al., 2020; Zentner et al., 2008).
Children's free responses to the musical stimuli.
To examine the extent to which the quality and intensity of participants’ free responses differed for each musical stimulus, a repeated-measures ANOVA was conducted, treating the stimulus target category and the 13 response categories (Table 8) as independent variables and our participants’ intensity ratings—a lot (2); a bit (1); or not at all (0)—as dependent variables. Estimated marginal means (Figure 2) indicate differences in the nature of children's subjective experiences according to the music they heard.
Estimated marginal means of participants’ intensity ratings of their free responses to the musical stimuli.
As can be seen in Figure 2, the highest intensity rating for the “tense” music is within the FEAR/TENSION category; for the “happy” music, the HAPPINESS category; and for the “relaxed” music, the RELAXATION category, although there is also a moderate peak in the SADNESS category. The highest intensity rating for the “powerful” music is shared between the POWER and FEAR/TENSION categories. These results correspond closely to our findings in the forced-response task.
The repeated-measures ANOVA revealed a statistically significant difference in the quality and intensity of response between at least two stimuli, F(5.844, 146.111) = 15.240, p < 0.001 (since sphericity is violated (ε = 0.487) Greenhouse–Geisser corrected results are reported), and a significant interaction between previously rated stimulus category and response category intensity rating, F(12.169, 304.235) = 8.057, p < 0.001, indicating that participants distinguished some variation in the quality and intensity of their subjective response, depending on the stimuli. The means and standard deviations for the intensity of response within each conceptual category to the musical stimuli are given in Table S17 in the supplementary material and the mode intensity ratings in Tables S18a and S18b in the supplementary material.
Post hoc pairwise comparisons revealed significantly different scores between several conceptual categories, as shown in Table 9.
Conceptual categories receiving significantly different scores in post hoc comparisons.
Visual Imagery and Narrative
Our participants’ free responses additionally produced 35 reports of music-evoked imagery, such as “It makes me feel like I’m outside Buckingham Palace celebrating the Queen's Jubilee, like I’m a soldier” (P5, 8yo F); “Really nice, like I’m sitting on a feather—a pink one. My head and feet are on a cloud” (P13, 7yo F), and narrative imaginings (McAuley et al., 2021), for example: “It's snowing, people are walking by themselves on the street and everyone around them is having fun” (P7, 11yo F); “I imagine lots of ballerinas doing a superior dance—it makes me want to dance” (P7, 11yo F); “Like a passed-away thing, a memory about someone, a rainy day” (P9, 9yo M).
No Feeling
No feeling was reported for seven stimuli and bored for six—all by the same participant (P2, 6yo M), who also reported at the beginning of the interview that they did not enjoy listening to music.
Developmental Differences
Children's own choice of vocabulary, therefore, provided categorical information about their responses, as did the forced-response task, but with an additional level of nuance that might be informative of developmental differences. Looking at the responses by age (Table S19 in the supplementary material), we can see a general pattern of development in the complexity of the participants’ vocabulary, from mostly single-word responses in the younger children to more sophisticated and nuanced descriptive responses in older children. The younger children generally gave single-word answers, predominantly using the experimenter-suggested words, although one five-year-old's responses were quite detailed: “Like a violin someone's not playing properly; like a storm; dangerous”; “Funny, really funny, like it could go on forever”; “Peaceful, but it might get dangerous”; “Like marching Romans—a big attack, dangerous, like swords bashing on to shields” (P4, 5yo M).
Some children explicitly described what they felt the music expressed, as opposed to something they felt themselves: “Like the part at the end of a movie where something sad happens but it’s also happy” (P25, 6yo F); “Funny, like it's meant to be scary” (P10, 7yo F); “Feels like a Christmas song”; “Feels like happy music from a fantasy film”; “Feels like music from a war film” (P18, 11yo M). One 11-year-old male made it clear that the music did not make him feel anything, but that he recognized its expressive qualities: “No feeling, but it sounds like a horror movie”; “No feeling, but a jolly sea shanty” (P18, 11yo M).
Some older children noted how their response evolved as the music changed: “A bit sad at the beginning, more powerful at the end” (P24, 8yo F); “Longing, kind of sad, more relaxed than happy—you can hear the breeze—but lively in some parts” (P8, 10yo M). Comments by older children aligned with ratings of the “longing” music as partly happy, partly sad, and reflective: “Sorrowful, reminds me of things, sad but also happy—remembering what was good about it” (P14, 9yo F).
Discussion
The purpose of this study was to explore whether children aged five to 11 years were able to describe subjective responses to music, and whether they could use six experimenter-adapted emotion terms to describe feelings evoked by situations and musical stimuli. Our findings indicate that children aged five years and older understood and could use our adapted terms in relating how a character in a story felt, and could apply them within a musical context. Furthermore, their free responses offer evidence of a capacity and willingness to describe music-induced experiences in considerable detail.
With respect to the experimenter-adapted terms, the youngest participants were unfamiliar with the word tense at the start of the experiment, and more participants used “scared” than “tense” in their free responses to the “tense” stimuli. Longing for something in the past also required clarification for some participants; nevertheless, once again they were able to select this term with respect to the target story. This suggests that with sufficient explanation, all six of our proposed emotion terms may be used by children aged five and older. For the purpose of an age-adapted categorical scale, scared may be more intuitive for young children to use than tense; however, since tense may also describe feelings of anger or frustration, scared should perhaps be added, rather than replace tense.
An alternative choice of methodology could have been to develop the categorical scale based on children's authentic responses in the first instance, since these would have generated age-appropriate terms from the outset. Nonetheless, considering that any categorical reduction necessarily excludes some potentially relevant concepts, including the free response option alongside the categorical task offers the possibility of capturing unlimited additional detail and nuance as well as revealing developmental and other individual differences.
The free responses endorse findings from the forced-response task, conceptually reflecting the target feelings, as well as categories of feeling reported in research with adult listeners. The free responses included synonyms of our proposed terms as well as some additional concepts, including amazed, brave, excited, heroic, proud, royal, invincible, something about to happen, triumphant, and unstoppable. Proud, heroic, and triumphant appeared as adjective markers for the “Power” factor early in the preparation of the GEMS (Zentner et al., 2008). Excited featured in the “Activation” factor, which would later be combined with joy to form the umbrella term “Joyful activation”, and amazed appears on the "Wonder” factor. Triumphant/ heroic is included among the 13 dimensions of music-evoked feeling identified by Cowen et al. (2020); and both excited and proud featured among terms tested in determining the 13 dimensions. The widespread use of the above terms in relation to the stimuli in this study suggests that participants considered them additions, rather than alternatives, to the six terms we presented them with, and possibly therefore relevant additions to the categorical scale—with the caveat that a larger selection of musical stimuli than we offered might elicit a greater diversity of responses and variety of descriptive terms.
We found significant differences in the quality and intensity of responses reported, depending on the music participants heard, results from the free responses endorsing our findings from the analyses of responses to the experimenter-adapted terms: that variation distinguished by participants in the quality of their subjective responses may be attributable to the music itself. The musical extracts elicited a greater diversity of responses than did the stories, which is perhaps intuitive, given music's ineffability, or “floating intentionality” (Cross, 2001) and the importance of individual differences and contextual factors (e.g., Barrett, 2017b; Cespedes-Guevara, 2023). In trying to understand the mechanisms behind music-induced feeling, examining subjective responses to specific expressive cues or musical structures may be more informative than testing the effects of whole pieces of music, even short extracts of which comprise a multitude of complex interacting expressive features that might elicit an infinite variety of sensory and psychological responses. Future research might attempt to measure the relative contribution of different musical structures to listeners’ holistic subjective experience.
Previous research has widely found that children's emotion categorization becomes gradually more nuanced throughout the primary school years (Widen & Russell, 2008), with complex emotions developing later. Our results reflect these findings: unsurprisingly, children's responses became more sophisticated and complex with age. Most (although not all) children aged seven years and older used the experimenter-selected terms in some responses but added their own thoughts and expressions, using a wider range of vocabulary and concepts.
Our results suggest that, although as children get older their vocabulary becomes more sophisticated, whether they experience music-induced feelings and are able to describe them is not necessarily age-dependent. Insightful responses were given by one of the youngest children, such as “Peaceful, but it might get dangerous” (P4, 5yo M), while one 11-year-old gave only single-word answers in response to the music.
The detailed responses given by children younger than seven years were almost exclusively about how the music sounded, rather than to do with their own feelings (except perhaps for “…it might get dangerous,” as reported). From age seven, responses included detailed feelings and imagined scenarios, such as “Like I need to dance and do something sharp, like wild dancing” (P13, 7yo F). This mirrors findings from developmental research that at five years children may have some capacity for introspection but are “generally much less aware than older children and adults of their own spontaneous, ongoing ideation” (Flavell et al., 2000).
With regard to the intensity ratings within each free response conceptual category, it is perhaps not surprising to see significant differences between pairs such as CELEBRATION and FEAR/TENSION, FEAR/TENSION and NOSTALGIA/REFLECTION, or IMPULSE TO MOVE and RELAXATION. Some results might appear surprising, such as the lack of significant difference between HAPPINESS and SADNESS, until we consider that a passage of music typically comprises a multitude of features and structures, any of which may evoke an affective response that, in turn, may be altered by the presence of another musical element and throughout the evolution of a musical work. Comments such as “a bit relaxed but also a bit sad” (P15, 6yo F); “Like the part at the end of the movie where something sad happens but it's also happy” (P25, 6yo F); “A bit scary and relaxing” (P25, 6yo F); “Relaxed, happy, and sad” (P10, 7yo F); “A bit sad, a bit scared, and a little bit happy” (P11, 7yo F) help to explain the lack of statistically significant difference found. A further possible explanation is that our rudimentary Likert scale with only three levels (a lot, a bit, or not at all) may limit how much information we can draw from comparisons.
The term longing for something in the past needed explanation for some younger participants and was interpreted by at least one (who volunteered this explanation to the experimenter) as wanting to go back in time not to re-experience a happy occasion but more specifically to escape danger [“Tense, sad, and longing for something in the past” (P26, 9yo F)], highlighting the need to be mindful of individual differences in interpretation of categorical terms. Longing was also selected much less often as the best fit to describe music-evoked feelings than happy, sad, or relaxed, which might suggest that longing for something in the past is not an age-appropriate term. An alternative conclusion might be that children do not often associate music with nostalgia or experience nostalgia through music. Nevertheless, comments in response to the “longing” music such as “Calm, happy, I like it, relaxed, thinking back” (P17, 9yo F) and “Making me remember something” (P14, 9yo F) suggest that the concept of nostalgia or longing had been understood in relation to the music as well as to the story. One child described the experience as “Like a wave, like a relaxed feeling, a smiley feeling” (P9, 9yo M). Nostalgia is often described by adults as a “wave”. That the “longing” music appeared to evoke happy, sad, and relaxed feelings in almost equal measure might possibly reflect the bittersweet nature of a feeling of loss or looking back.
Barrett et al. (2010, p. 390) found that, for adult listeners, music-induced nostalgia “was stronger to the extent that a song was autobiographically salient” suggesting that familiar music is more likely to induce nostalgia than unfamiliar music. Lahdelma and Eerola (2015) additionally showed that certain harmonic elements, such as major and minor seventh chords, contributed to a feeling of nostalgia in adult listeners. On this basis, it may be that such structural elements present in unfamiliar music could be enough to evoke a sense of nostalgia, or, that smaller elements of familiar music appearing in unfamiliar music might evoke a memory of the original, familiar music. Our participants appeared to understand the feeling of reflecting on a memory, some expressing this as “wanting to go back” (P15, 6yo F) or “thinking back” (P17, 9yo F) and one, very eloquently, “Reminds me of things—sad but also happy—remembering what was good about it” (P14, 9yo F).
That children were all able to use the terms we proposed in relation to the stories indicates that these terms could be used in other contexts. However, the detail our participants provided in their free responses is evidence that an unconstrained task has the potential to reveal much more nuanced information than a categorization task from a developmental perspective. We therefore recommend that the free response option should remain in future experiments with this population. How best to balance the open question with the categorization task needs to be considered further.
The free responses revealed a range of music-induced experiences beyond pure feeling sensations, which included visual imagery and narrative imaginings. Taruffi and Küssner (2023) have highlighted the prevalence of visual mental imagery as integral to many people's music-listening experience and identified three main categories of music-evoked visual imagery: storytelling, associations, and references. Hutto's (2007) Narrative Practice Hypothesis illustrates how children use narrative through play to understand and explain feelings and behavior—to construct meaning. Lüdemann (2022, p. 9) draws a connection between narrative in play and narrative in music, noting: “In musicking, the intentionality to invest sounds with significance intersects with humans’ impulse to play.” Encouraging children to articulate their music-induced narrative imaginings might potentially offer an ecologically natural means to collect rich authentic data that could inform how humans begin to construct meaning from musical sounds or, conversely, the motivations for and mechanisms that inspire musical composition.
Margulis et al. (2022) found that listeners with a shared culture described remarkably similar narratives imagined in response to musical stimuli, compared with those from a different culture, suggesting intersubjectivity in music-evoked imagery and narrative that is dependent on a shared culture. Extending the study of music-evoked imagery and narrative to our target population presents an intriguing avenue for future research.
Limitations
An important limitation of our experimental procedure is that participants in our study were introduced to the experimenter-adapted emotion words after hearing the first stimulus. Therefore, only their response to the first extract heard was genuinely authentic. Responses to all subsequent stimuli might have been influenced by seeing the set words. Although older children's vocabulary was, unsurprisingly, generally richer than younger children's, there was evidence in many responses that children of all ages had been influenced by the words that we had presented to them. Some participants who had not initially understood tense or longing for something in the past, for example, nevertheless used these in their “free” responses. Using terms that children would not naturally use without guidance risks biasing our understanding of how music genuinely makes them feel and thus limits what we can draw as evidence from our results here. Further, suggesting words to the participants means imposing adult emotion concepts on them, even if the vocabulary has been simplified and meanings explained. Asking children to provide authentic descriptions in the first instance or collecting free responses to all the musical stimuli before retesting them using the rating scale would obviate potential priming. In any replication study, therefore, either free responses to all stimuli should be collected before presenting participants with the categorization task, or the two tasks should be counterbalanced.
The sample size and choice of stimuli also present limitations. Because our sample was small, we used Fisher's exact tests to explore relationships between the stimuli and participants’ responses; however, it should be noted that our participants were allowed to choose more than one response category in the categorical task, meaning that the assumption of mutual exclusivity is violated in this case. Future experiments could examine pairwise differences between response categories, or limit participants to one best-fit category choice for each stimulus.
Playing only six musical stimuli to each listener here provided us with some initial insight into children's capacity to describe music-evoked experiences; however, in future research, this can be explored in much greater depth by testing a wider variety of musical styles and genres. With this in mind, this study might provide a framework for a longitudinal study exploring developmental aspects and aged-related differences.
An additional consideration is that participants were randomly assigned to hear either version A or B of the musical extracts (originally with a larger sample in mind) and these were therefore not shared equally between children of all age groups. For example, all but one of the five- and six-year-olds were randomly assigned to version B stimuli. This imbalance makes it difficult to comment on developmental differences in responses to the version A stimuli and further highlights that, with a sample this small, the decision to split our musical stimuli into two groups has proved to be a limitation of our methodology.
Our stimuli varied in duration from 15 seconds to just under a minute; it is possible that this difference might have had a bearing on the intensity of response experienced. For future experiments, it may be worth searching for a range of extracts of closer uniformity in duration, also considering that longer excerpts may also have the potential to elicit a greater variety of feelings than shorter excerpts.
Kallinen and Ravaja (2006, p. 206) suggest that the laboratory environment might mitigate listeners’ subjective involvement in the music and thus the intensity of their felt responses: “The results suggest that the emotions perceived and felt should be studied also in various natural environments (e.g., in a concert or at home).” On this basis, running future experiments online would allow participants to listen to the stimuli in a more private setting, allowing them to engage with the music in a more ecologically typical way. Further, distributing the experiment online to attract a much larger and diverse participant sample could facilitate comparison of responses from a cultural as well as developmental perspective.
Conclusions
The purpose of this exploratory study was twofold: to test the suitability of a music-specific categorical measurement scale for children of primary school age; and to examine the extent to which they could describe subjective responses to music in their own words, informing the development of an age-appropriate format for children to report on their music-induced experiences.
Our results indicate that children aged five years and older could use the categorical scale and the six experimenter-adapted terms, although longing for something in the past and tense required some explanation to children younger than eight years. Our statistical analyses found that the story stimuli elicited more predictable responses than did the musical stimuli; however, the purpose of this experiment was primarily to determine whether children could use our selected words to describe music-evoked feelings. Our results suggest that they could, and their free responses provided further evidence that talking about subjective responses to music comes naturally to children in this age group. Exploring the relationship between specific musical stimuli, styles, or structures and children's responses is a topic for future research.
The ubiquity of reports by adults of nostalgia induced by music, along with the frequency of its appearance in our participants’ free responses, suggests that this is an important concept to investigate in developmental research and that its inclusion in the categorical scale is warranted; however, replacing the word longing with wishing or wanting might render our term more accessible to young participants. The inclusion of additional concepts that emerged in participants’ free responses, such as amazed, excited, proud, heroic, brave, and triumphant, is equally worth considering.
The variety of expressive terms that emerged in the free responses highlights the importance of allowing participants to speak for themselves when conducting exploratory research. Given the quality of free responses from even our youngest participants and recommendations from the literature (e.g., Cowen et al., 2020; Gabrielsson, 2010), we consider the option to contribute free descriptions of great value to a children's response tool. Allowing children to use their own language additionally offers the potential to elicit data that can inform about their emotional granularity, as well as more detail about their music-evoked feelings. Depending on their age or level of literacy, children may prefer to write their own responses or have these recorded by an adult supervisor.
Including both open questions and a categorization task, allowing us to examine responses to music both qualitatively and quantitatively, stands to elicit more detailed and nuanced data than either method alone. However, the use of both has implications for the duration of the experiment and the number of musical stimuli, bearing in mind our target age group. Counterbalancing the two tasks or assigning different participants to each task would eliminate the risk of priming discussed earlier. The wording of the open question should be clear enough to ensure that participants understand that they are being asked to describe their own experiences, rather than what they believe the music expresses.
Additionally, the notable number of responses that involved visual imagery and narrative, particularly in the light of recent work with adult listeners (e.g., Margulis et al., 2022; McAuley et al., 2021) warrants further research to explore children's music-evoked imaginings. The free response task could therefore include two questions: for example, “How does this music make you feel?” and “Does this music make you think of or imagine anything?”
What began as an effort to simplify has produced results that imply a need to expand on, rather than reduce, existing formats to optimize the detail and depth of data we hope to capture. Our findings—that children as young as five years can produce detailed and nuanced descriptions of feelings, imagery, and narrative imaginings in response to music—present an intriguing range of elements ripe for future developmental research.
Supplemental Material
sj-docx-1-mns-10.1177_20592043241265318 - Supplemental material for Putting Musical Feelings Into Words: Children's Verbal Descriptions of Music-Evoked Experiences
Supplemental material, sj-docx-1-mns-10.1177_20592043241265318 for Putting Musical Feelings Into Words: Children's Verbal Descriptions of Music-Evoked Experiences by Caroline Owen, Hauke Egermann and Andrea Schiavio in Music & Science
Footnotes
Acknowledgments
We extend special thanks to Mabel Collier and Dora Collier for commenting on and contributing to the story stimuli, Mabel Collier for narrating the stories, and Hannah Gibbs for technical support.
Action Editor
Adam Ockelford, University of Roehampton, School of Education.
Peer Review
Scott Bannister, University of Leeds, School of Music.
Julian Céspedes Guevara, Icesi University, Estudios Psicologicos.
Contributorship
CO and HE were involved in designing the study, gaining ethical approval, and analyzing the data. CO researched the literature, recruited participants, and conducted the listening experiments. CO wrote the first draft of the manuscript. CO, AS, and HE reviewed and edited the manuscript and approved the final version of the manuscript.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Ethical Approval
Ethical approval was granted by the University of York Arts and Humanities Ethics Committee in 2022.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Data Availability Statement
The datasets generated and/or analyzed during this study are available in the OSF repository: https://osf.io/sgz73/?view_only=a280daa440b94d5e9ce3cc6e2d49eac7.
Supplementary Material
All supplementary material for this article is available online.
Notes
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
