Abstract
Music-based interventions and music lessons modulate cognitive functions, such as language or attention. However, the specific and differential effects of musical activities are a new focus of research. Therefore, this study aimed to investigate the effect of musical improvisation, a focal musical intervention, on the emotional memory of 4- and 5-year-old children. Each child individually looked at 24 neutral and emotional pictures and rated their valence and arousal. After that, the children were exposed to one of three interventions: musical improvisation (experimental intervention), musical reproduction (active control intervention), or rest (passive control intervention). Then, recall and recognition (immediate and deferred) were used to evaluate memory performance. The main results indicated that musical improvisation, compared with a reproduction music activity, improved memory. In addition, rest improved recognition compared with reproduction. Besides, children recalled more emotional than neutral images. Musical improvisation is a promising technique to be implemented in the educational field.
Finding different teaching strategies to improve children’s performance in the educational field has been the goal of several studies in recent years. Thus, different activities have been tested to improve the performance of cognition in children (Ballarini et al., 2013; Benítez et al., 2018). Learning and memory are cognitive functions of great relevance that have attracted researchers’ attention since they are capital abilities in education (Immordino-Yang & Damasio, 2007).
Different biological, environmental factors, or treatments affect learning and memory, either by improving or deteriorating them, which is known as modulation processes (Justel et al., 2013b). Among these factors, emotion is a key aspect (Courage & Cowan, 2008; Justel et al., 2013a). An event could be considered emotional according to two variables: valence (characterized as a spectrum that ranges from pleasantness to unpleasantness) and arousal (the degree of physiological activation that ranges from calmness to excitement; Diaz Abrahan, Benitez, et al., 2020; Kensinger & Murray, 2012; Osgood et al., 1957). Events that are highly pleasant or unpleasant are better consolidated than those that are in the middle of the spectrum. Regarding arousal, information causing a moderate level of excitement is better retained than that at the extremes, like an inverted U-shape. Generally, studies have shown that emotional stimuli (pleasant or unpleasant with a moderate to high level of arousal) are processed differently from non-emotional or neutral stimuli and are better remembered over time (Cordon et al., 2013; Goodman et al., 1994; Leventon et al., 2014; Ortega & Ruetti, 2014; Ruetti et al., 2019). Therefore, under certain circumstances, the emotions in the acquisition or consolidation memory periods improve storage, leading to better subsequent recall (Dolcos et al., 2005). This phenomenon is known as emotional memory (Marchewka et al., 2016; McGaugh & Roozendaal, 2009).
Besides emotion, another factor that has shown modulatory potential is music (Diaz Abrahan & Justel, 2019b). Music modulates cognitive functions and mediates human development (Benítez et al., 2017; Bilhartz et al., 2000; Dumont et al., 2017). Studies have shown the beneficial effects of long-term music training on functions related to music (near transfer, for example, better fine motor coordination due to piano training; Bergeson & Trehub, 2005; Cohrdes et al., 2018; Habibi et al., 2016) and to non-musical skills (far transfer; Benz et al., 2016; Schlaug et al., 2005). Far transfer effects refer to musical training affecting functions that are not easily related to music, such as memory, language, and executive functions (Hogan et al., 2018; Jaschke et al., 2018; Linnavalli et al., 2018). An example of far transfer was reported by Benitez et al. (2018), who showed that 4 weeks of music training with a school-based intervention enhanced memory in 4- and 5-year-old children. However, the options and activities that emerge from music are numerous (Chanda & Levitin, 2013). A specific type of research involves the use of music-based interventions (Särkämö, 2018; Särkämö et al., 2016), implemented only once during or after a task to improve cognitive performance (Diaz Abrahan, Bossio, & Justel, 2019). Because of their acute implementation, these interventions are called focal (Pontifex et al., 2015). In this line of research, the activities that involve music listening are called receptive interventions (Särkämö et al., 2008), and those that imply music exploration, playing, music-making, singing, dancing, among others, are called active music interventions (Schneider et al., 2018). An example of the implementation of a focal proposal is the study performed by Ballarini et al. (2013), in which, after the acquisition of information, children exposed to only one music class (based on games with homemade musical instruments) improved their visual and verbal memory.
Specifically regarding emotional memory, the effect of receptive or active music activities, such as focal interventions, has been investigated in adults. Participants who listened to (Justel et al., 2015, 2016) or made music after the acquisition of information (Diaz Abrahan et al., 2021; Diaz Abrahan, Shifres, & Justel, 2019, 2020) had better memory performance than different control groups (participants who listened to white noise, rested, or reproduced a rhythmic pattern). The literature indicates that music production is more effective than receptive treatments for cognitive modulation (Fancourt et al., 2014). Improvisation, an example of a focal active music intervention (Abrahan & Justel, 2015; Beaty, 2015; Limb & Braun, 2008), implies making music with a creative component, and it may be performed by musicians and non-musicians alike (Diaz Abrahan et al., 2021). Free improvisation has been shown to modulate neutral and emotional memory in young and older adults (Diaz Abrahan et al., 2021; Diaz Abrahan & Justel, 2019a; Diaz Abrahan, Shifres, & Justel, 2019, 2020). An important benefit of this intervention is that it is flexible enough to be employed with different populations, regardless of the participants’ formal knowledge of music (Diaz Abrahan, Shifres, & Justel, 2019). However, research on the effect of music-based interventions on children is a gap in the scientific literature.
The current study
The main goal of this study was to investigate the potential effect of a musical focal intervention (music improvisation) on children’s emotional memory in an educational environment. According to previous findings on the use of music-based interventions to modulate adults’ memory (Diaz Abrahan et al., 2021; Diaz Abrahan & Justel, 2019a; Diaz Abrahan, Shifres, & Justel, 2019, 2020), we expected children exposed to musical improvisation to remember more information than passive and active controls (participants not exposed to musical intervention or participants who reproduced a rhythmic pattern). In addition, we expected children to recall more emotional than neutral images (Benítez et al., 2017; Cordon et al., 2013; Leventon et al., 2014; Ruetti et al., 2019).
Method
Participants
A priori power analysis revealed that to detect a small- to medium-sized effect (η2 = .06), with 90% power, the study required a total sample size of 33 participants (Faul et al., 2007). Forty-eight preschool children, aged 4 and 5, participated in this study (43.73% female). They were attending two schools (n= 434 and 487) in Chubut Province (Argentina) and one school (Instituto José Hernandez) in Buenos Aires Province (Argentina; Table 1). The children recruited had no chronic diseases or birth defects, according to their parents, and therefore, they met the inclusion criteria.
Participants’ Age and Gender.
Their families were primarily from low socioeconomic backgrounds (according to school reports of sociodemographic characteristics of the area where the schools are located), with Spanish as the primary language and Latin American ethnicity. We recruited them by means of presentations at family school meetings. The children were randomized to a musical improvisation condition, an active control condition (music reproduction), or a passive control condition (without music, resting). All children were informed about the activity and asked whether they wanted to participate in the study. Informed consent was obtained from parents or legal guardians prior to the study, according to the current ethical standards, complying with the Helsinki Declaration, Convention of the Council of Europe on Human Rights and Biomedicine. No incentive or payments were given to participants.
Interventions
Improvisation (IMP)
The children performed a musical improvisation activity. This study adheres to the music therapy perspective of improvisation, in which music improvisation is a creative musical experience with great flexibility to adapt to people with and without musical skills (Diaz Abrahan, Shifres, & Justel, 2020). In this activity, the children performed a musically novel task that was not copied, repeated, or imitated, but invented (Tafuri, 2006). They participated in this activity for 3 min, using musical instruments, their voices, and/or their bodies. Considering the children’s level of comprehension and age, we gave them the following instructions: “Would you like to play with instruments? We are going to create and invent sounds and music. How about that? We can use musical instruments, sing, and move.” The researcher presented a rhythmic base-pattern (a rhythmic pattern of two 4/4 bars played as a loop [ostinato]), with a percussion instrument at a medium volume, to encourage creative production.
Reproduction (REP)
The children performed a musical rhythmic reproduction activity. The researcher presented the rhythmic base-pattern used in the musical improvisation (with the same percussion instrument as that used for IMP, at a medium volume), and the children responded by imitating the pattern with their instruments for 3 min. The following instructions were given: “Would you like to play with instruments? We are going to listen to a sound and then you have to make the same sound. How about that? We can use musical instruments, sing, and move.” This intervention was included as active control for possible effects of movements, auditory perception, and musical instruments, among others, that could explain the findings.
Rest condition (REST)
The children remained silent for 3 min. The instruction used was “Do you want to play a game to see who of us can be quiet for a while?” The aim of this intervention was to take into account a passive control condition.
Materials
Instrumental setting
For the musical activities (improvisation and reproduction), the participants were allowed to choose percussion instruments (e.g., drums, maracas, bells, woodblocks, shakers, tambourine) or melodic or harmonic instruments (e.g., guitar, melodica, xylophone, flutes). These instruments were included because they were easy to handle.
Memory task
A set of 72 child-appropriate images was selected from the International Affective Picture System (IAPS; Lang et al., 1995). 1 Twenty-four of the pictures were pleasant images (positive valence and moderate to high arousal, for example, ice-cream), 24 were unpleasant (negative valence and moderate to high arousal, for example, a boy crying), and 24 were non-emotional (neutral valence and non-arousing, for example, a book). Twenty-four of the images were used as target pictures to rate valence and arousal, and the other 48 images were used for recognition tests (immediate and deferred).
Because it was reported (Solomon et al., 2012) that some children find it difficult to understand the self-assessment manikin (the figures that Lang et al., 1995 designed to evaluate arousal and valence), we decided to use a face system to facilitate the task and test the valence and arousal of the selected pictures. To evaluate valence, at the sight of each of the 24 images, the children were asked to point to the corresponding face (sad, neutral, or happy). To analyze emotional valence, the data were transformed into a three-point scale where sad, neutral, and happy faces were assigned the value of one, two, and three, respectively. The possible differences in the coding of the images were analyzed with the Chi-square test (χ2). The analysis indicated a significant effect, χ2 = 125.56, p < .0001. To further analyze the differences and compare the images, the Mann–Whitney U test was employed. This test indicated significant differences between neutral and positive images, Z = –5.63, p < .0001; neutral and negative, Z = –6.87, p < .0001; and positive and negative Z = –8.69, p < .0001.
To evaluate arousal, for each selected emotion, the researcher showed the child the face he or she had selected in five different sizes, from small to large, and the child was asked to select one of the sizes to indicate how happy (or sad) he or she felt about the picture. To analyze the results, we performed a one-way analysis of variance (ANOVA) with Image (positive, negative, or neutral) as the factor. This analysis yielded a main effect of Image, F(2, 143) = 6.56, p = .002. The post hoc analyses indicated that the emotional pictures (positive and negative) were more activating than the neutral ones, without differences between positive and negative images, p > .05. Table 2 shows the mean and standard deviations for both emotional dimensions.
Emotional Processing.
Procedure
Data were collected in a quiet room inside the children’s school during the academic year. The procedure consisted of three sessions. See Figure 1 for a scheme of the experimental procedure.
Scheme of the Experimental Procedure.
Session 1
Prior to testing, the researchers received the school’s approval and then held a parents’ meeting in which the study was explained. The parents were shown the images, and if they agreed with the study, they signed their informed consent and completed a sociodemographic questionnaire (Information Phase).
Session 2
The second session comprised three phases. In the first one (Acquisition Phase), each child was tested individually in a quiet room inside the school, accompanied by a researcher (three researchers evaluated the participants, one per child). The child sat in front of a computer monitor, and eight neutral, eight positive, and eight negative pictures were presented in random order and he or she rated their valence and arousal.
In the second phase (Treatment Phase), the children were blindly and randomly exposed to one of the conditions (musical improvisation, musical reproduction, or rest) in groups for 3 min.
In the third phase of the first session (Test Phase), the researcher asked each child individually which images he or she recalled from all the pictures shown before (immediate free recall test). The child responded by using a word or phrase that described the image recalled (e.g., three happy cats, a broken building, an ice-cream cone, a man crying, a comb). Next, the 24 images were mixed with 24 new ones; while looking at them, the child was asked to decide which images were new and which ones had been shown before (immediate recognition test).
Session 3
A week later, deferred free recall and deferred recognition tests were carried out. The procedure was the same as that used in Session 2, but this time a different set of new 24 images were used for the recognition test.
Design and data analysis
An experimental randomized design was run to identify the effect of a focal musical activity on the children’s memory. Participants were randomly and blindly assigned to the different groups, and they were always tested individually.
The assumptions of normality and homogeneity of variance were tested through the Shapiro–Wilk and Levene’s tests, respectively. The results indicated that the assumptions were maintained, and therefore, we employed parametric analysis.
The dependent variables were the number of pictures recalled or recognized, immediately or deferred. Recall and recognition were independently analyzed through repeated measures ANOVA with Intervention (Improvisation, Reproduction, and Rest) as the between factor, while Image (Neutral, Positive, and Negative) and Time (immediate and deferred) were the within factors.
Post hoc least-significant difference pairwise comparisons were conducted to analyze significant main effects and interactions. Partial Eta-squared (η² p ) was used to estimate effect size. The alpha value was set at .05, and the software SPSS Statistics 23 was used to compute descriptive and inferential statistics.
Results
The impact of the three interventions on preschoolers’ memory was evaluated through two memory tasks. Data from six children were discarded because they were not present in the second session (7 days later).
Free recall
After the children performed the interventions (improvisation, reproduction, or rest in session 2), the researcher asked them what images they remembered (immediate free recall, Figure 2(a)). This task was repeated 7 days later (deferred free recall, Figure 2(b)). Children responded by using a word or phrase that described the images that they remembered (e.g., three happy cats, a broken building, a crying boy, a happy girl with a watermelon).
Free Recall Tasks.
The ANOVA yielded significant main effects of Time, F(1, 39) = 9.53, p = .004, η² p = .196; Image, F(2, 78) = 11.59, p < .0001, η² p = .229; and Intervention, F(2, 39) = 17.61, p < .0001, η² p = .474. The interactions Image × Intervention, F(4, 78) = 3.07, p = .021, η² p = .136; and Image × Intervention × Time, F(4, 78) = 2.78, p = .035, η² p = .123, were also significant. The corresponding post hoc indicated that more images were recalled in the immediate measure than in the deferred one, that more emotional than neutral images were remembered, that IMP remembered more images than REST, p = .014, and REP, p < .0001, and that REST remembered more images than REP, p = .001.
Regarding the three-way interaction (the two-way interaction was submitted to the three-way one, and therefore, those results are not shown), for IMP on the immediate and deferred measure, the emotional images were better remembered than the neutral ones. For REST, the emotional images were better remembered only in the deferred measures, and the neutral and positive images were better remembered in the immediate than the deferred measure. Immediately, positive images were better remembered by IMP than REP participants (p = .008), and by REST than REP participants (p = .014). Negative images were better remembered by IMP than REP participants (p < .001), and a trend was found for IMP remembering more negative images than REST (p = .06). As for deferred measures, the IMP group remembered more positive images than REST (p < .001) and REP groups (p < .0001). Besides, IMP and REST remembered more negative images than REP (p < .0001 and p = .006, respectively).
Recognition
The second memory task was recognition. To evaluate it, the 24 target pictures were mixed with 24 new pictures (immediate recognition, Figure 3(a)) and other 24 new pictures 7 days later (deferred recognition, Figure 3(b)). The children were asked to indicate if the images were new or old, by answering Yes/No. False recognitions were subtracted from the total recognition score. The statistical analyses indicated significant main effects of Intervention, F(2, 39) = 18.65, p < .001, η²p = .489, and Image, F(2, 78) = 30.08, p < .0001, η²p = 435; the interactions Image × Intervention, F(4, 78) = 26.26, p < .0001, η²p = .574, Image × Time F(2, 78) = 49.41, p < .0001, η²p = .559, and Image × Intervention × Time, F(4, 78) = 42.55, p < .0001, η²p = .686, were also significant.
Recognition Task.
The post hoc analysis for Intervention indicated that IMP and REST participants remembered more images than REP, without differences between them. Regarding Image, the post hoc indicated that negative images were better remembered than neutral ones, and the latter, better remembered than positive images.
The post hoc analysis for the three-way interaction (two-way interactions were submitted to the three-way one, so they are not reported) indicated that the REP group immediately remembered more neutral than negative images, and in deferred measures, they remembered more negative than neutral images and more neutral than positive images. Besides, positive images were better remembered in the immediate than the deferred measures, and negative images were better remembered in the deferred than the immediate measures. The three types of images were better remembered by IMP and REST groups than by REP in the immediate measures. Regarding the deferred measures, neutral images were better remembered by IMP and REST compared to REP (p = .011 and p = .004, respectively). The same pattern of results was found for the positive images, that is, IMP and REST participants remembered more images than REP (ps < .0001). No differences were found for negative images (p > .05).
Discussion
Different strategies that improve the cognitive functions of children have been identified (Ballarini et al., 2013; Benítez et al., 2017, 2018). This notion represents a potential field of research that may provide strategies for enhancing school learning. In line with this research line, this study aimed to test the effect of a specific music-based intervention on the consolidation of the visual emotional information of preschoolers.
The principal prediction, and the fundamental hypothesis of this work, was that children exposed to the musical improvisation would remember more information than those reproducing a rhythmic pattern or resting in the treatment phase. This hypothesis was partially corroborated because the two control activities generated different effects on memory. Music improvisation improved memory when compared to one of the two control conditions (active control). In other words, the results showed that musical improvisation, performed after information acquisition, improved memory in comparison with the children who reproduced the rhythmic pattern, in the immediate and deferred measures. However, music improvisation, in comparison with the rest condition, only improved the deferred free recall of positive images. The general finding of this study is in line with the results of the modulatory effect of music, specifically musical production, on the visual and verbal memory of adults (Diaz Abrahan et al., 2021; Diaz Abrahan, Shifres, & Justel, 2019, 2020).
The difference between the two types of musical activities used in this study generates an interesting discussion regarding memory modulation through musical interventions because it provides information about the need for specificity of the musical proposals. Music, sound, movement, and the use of musical instruments were present in both activities, but the difference lay specifically in the creative process involved in improvisation.
From a neuroscientific perspective, musical improvisation is considered an example of complex creative behavior. Regardless of the degree of improvisation complexity, some studies have shown that there is wide neural activity when people generate new sounds and/or musical structures (Abrahan & Justel, 2015; Limb & Braun, 2008) and a great cognitive compromise that involves attention, working memory, autobiographical memory, executive functions, among others (Biasutti, 2017). However, further investigation, focused on neurocognitive aspects, is necessary in this regard.
In this work, the definition of musical improvisation was adopted from a music therapy perspective, in which participants make use of the resources available at a given time to create novel sounds and melodies, minimizing the importance of aesthetic musical results (Bruscia, 1998). In this discipline, musical improvisation is a tool that generates a positive emotional state (Punkanen, 2011). This emotional induction could influence memory consolidation. Some research proposes that emotion, along with the activation of neural structures involved in emotional processing with neuronal projections to the hippocampus, modulates memory acquisition and consolidation (Diaz Abrahan, Benitez, et al., 2020; Shields et al., 2017; Tambini et al., 2017).
The reproduction condition seems to interfere with memory formation. During rhythmic imitation, children are asked to listen to and repeat a pattern by synchronizing with an external rhythm. This intervention could diminish cognitive resources and lead to mnemonic deterioration (Diaz Abrahan, Bossio, & Justel, 2019; Miendlarzewska et al., 2013). In this sense, the children who were playing a rhythmic pattern may have been concerned with faithfully replicating or sharply adjusting to the pattern, leading to memory impairment. Besides, this condition may have interfered with rehearsal and storage in the phonological loop and may have transferred to long-term memory (Baddeley, 2012; Baddeley & Hitch, 2019), but this hypothesis needs further research.
An interesting finding concerning the rest condition is that the results were similar to those of the music improvisation intervention, especially in regard to recognition. These results may be explained by the latest innovative research on the effect of rest (Humiston & Wamsley, 2018; Wamsley, 2019) and sleep (Payne et al., 2009, 2015) on memory consolidation. This line of research suggests that, after the acquisition of information, no new sensory stimulus should be encoded to facilitate memory formation. Following this logic, the children who participated in the rest condition did not perform any type of learning activity after seeing the images, thus decreasing the input of sensory information. This information could explain the improvement of memory by this group, and it is a strategy to be considered in educational environments. In this era of omnipresent music and sound, silence could be an intervention for educators in general.
According to previous studies about the processing of emotional memory (Cordon et al., 2013; Ortega & Ruetti, 2014), we expected the children to recall more emotional than neutral images. Our results show that, both in the immediate and deferred free recall, the positive and negative images were more activating and better remembered than the neutral images by the improvisation group, and better remembered only in the deferred measure by the rest condition. These results are in agreement with other studies that found that events and experiences with emotional content are better recalled than neutral ones (Benítez et al., 2018; Justel & Diaz Abrahan, 2012). Regarding recognition, negative images were better recognized than the neutral ones and the latter better recognized than positive images, but this pattern was only found in the reproduction group, as the improvisation and rest groups recognized each type of picture similarly. This could be due to a ceiling effect in these groups, since both groups recognized a high number of pictures. This finding could be tested in future studies employing a larger number of images to be encoded by children to avoid the ceiling effect.
Limitations and future directions
Our study had some limitations, the number of participants being one of them. Future studies could extend the sample number and include a cognitive state and/or IQ baseline. Nonetheless, it is worth noting that, even with a low sample number, the results are significant and have a large power effect regarding the intervention factor. Besides, it would be interesting to study a different type of sample, such as dissimilar socioeconomic levels. It remains to be determined whether the results would replicate in a sample with moderate or high socioeconomic level, in which children are more exposed to musical activities since they have more musical material resources (Dumont et al., 2017) than children with low socioeconomic level, and therefore, the impact of improvisation may not be significant. In this sense, the effects of stimulation or interventions may be stronger in groups of children with low socioeconomic levels (Hermida et al., 2019). Thus, further research is needed in this area. It would also be relevant to evaluate the effect of musical improvisation on other school participants (e.g., adolescents), or the effect of focal musical proposals on classroom academic skills, using indicators of IQ, for example.
General conclusion
There are pioneering studies on the use of creative musical activities with preschoolers and their influence on non-musical factors. Our findings are in line with the results of the modulatory effect of music, specifically musical perception (Justel et al., 2015, 2016) and production (Diaz Abrahan et al., 2021; Diaz Abrahan, Shifres, & Justel, 2019, 2020), on adults’ memory. The results obtained in our study contribute to the field in two ways. First, they provide information about the specific effect of musical activities on cognitive functions, specifically memory. Second, they allow us to consider the potential of rest periods as an educational strategy.
Musical improvisation is a unique psychological phenomenon, available to most people, but different from other areas of musical activity, which require specific technical skills or knowledge (MacDonald & Wilson, 2014). This availability makes it a potentially wide-range intervention, which was tested in the current work as a modulator of a cognitive function in an educational context. According to social, economic, and political factors, the educational system throughout the world is different (Crossley, 2020). In this sense, Latin America has the highest levels of inequality in income distribution in the world, which affects the resources available for education (Briasco, 2010). This study proposes the use of ecological interventions, which are flexible in their use, accessible to institutions, and teams of professionals, and easy to be implemented between regular curricular activities to reinforce the educational content learned. Therefore, these aspects have implications for the design of educational public policy and intervention programs for our specific Latin American educational context and beyond.
Footnotes
Acknowledgements
The authors thank schools that participated in this study (N° 434, 487, Instituto Jose Hernandez), from Chubut, and Buenos Aires provinces, Argentina, and Ana Abrahan, music teacher, for her invaluable collaboration in this study.
Author contributions
VDA and NJ contributed to the conception and design of the studies. VDA, MBo, and MBe conducted the studies. VDA and NJ contributed to data analysis. VDA wrote the first draft of the manuscript. All authors reviewed and edited the manuscript and approved the final version of the manuscript. NJ supervised the study.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by CONICET and UNSAM and grants PICT 2014-1323. 2015-2017 and PICT 2017-0558. 2018-2021 to NJ.
