It is well established that Phonological Awareness, PA, is a complex meta-linguistic skill and it refers to an individual’s ability to reflect upon and manipulate the sound structure of spoken words. PA is a forerunner to learning to read in an alphabetic writing system (^{Alegría, 2006}; ^{Caravolas et al., 2012}; ^{Castles, Rastle, & Nation, 2018}; ^{Defior, 2014}; ^{Dehaene, 2017}; ^{Melby-Lervåg, Lyster, & Hulme, 2012}).

Many have argued that the development of phonological awareness in children follows a hierarchical pattern, progressing from the ability to isolate large sound units (words) to small abstract units (phonemes); which conform oral language (^{Defior & Serrano, 2014}; ^{Melby-Lervåg et al., 2012}; ^{Nohales & Giménez, 2014}).

Children’s ability to analyse smaller and smaller parts of oral language depends on adequate cognitive development, as well as their social and home literacy environment. Early forms of PA development prior to literacy instruction, suggests that experiences with early oral language play an important role in evolving PA. Children’s experiences with written language dramatically influence PA development, especially in the advancing of phoneme awareness (^{Defior, 2014}; Defior & Serrano, 2014; Defior & Serrano, 2011; ^{de la Calle Cabrera, Villagrán, & Guzmán, 2016}; ^{Domínguez, Nasini, & Teberosky, 2013}; ^{Melby-Lervåg et al., 2012}; ^{Nohales & Giménez, 2014}; ^{Villagrán, Consejero, Guzmán, Jiménez, & Cuevas, 2011}). In this sense, 4 year-old children, who grow up in an auspicious home literacy environment, are able to identify words in which the final sounds are the same, for example “sopa vs pipa” (“soup” vs “pipe” ). Rhyme awareness requires less conscious and deliberate manipulation of segments than the other phonological levels, which makes it less cognitively demanding when compared to other phonological awareness skills. Therefore, in most cases, it does not depend on the explicit instruction.

Succeeding children can detect or manipulate syllables before they can detect or manipulate onset and rime. In addition, they are able to perform in initial-final sound word identification and isolation tasks. When children are exposed to formal learning of written language, they can detect or manipulate individual phonemes within intra-syllabic word units. Phonemic awareness includes the ability to separate a word into the sounds that make it up and blend single sounds into words. It also involves the ability to add, omit, or substitute new sounds in words (^{Cuadro, 2010}; ^{Defior & Serrano, 2014}; ^{Castles et al., 2018}; ^{Torres, 2014}). Each one of these phonological tasks involve different levels of difficulties, it depends on their cognitive, linguistic and memory requirement.

The difficulty of a phonological awareness task depends on the size of the phonological unit and the nature of the manipulation that is required. Additionally, the complexity level is defined by unit positions in a word. Preliterate children recognize initial units easier than ending units (e.g., isolating initial sound vs isolating ending sound). On the contrary, if children must manipulate the units into a word, it will be easier to manipulate ending sound than initial sound (e.g., deleting ending sound vs deleting initial sound). Children’s experiences with written language dramatically improve phonological awareness tasks performance, especially in phonological units deletion and addition tasks, as a consequence of a reciprocal relationship between reading and phonological awareness. The higher level of complexity of any phonological task is perceived in phonological units that are located in the middle of a word (^{Bravo, Villalón & Orellana, 2011}; ^{de la Calle Cabrera et al., 2016}; ^{Nohales & Giménez, 2014}).

The positive effects of PA intervention programs are well established. Several studies have demonstrated that PA can be fostered through nursery rhymes, storytelling and oral languages activities. These studies have found that PA intervention programs are beneficial in terms of progress in reading abilities, both in infants with typically developing and young children who are at risk of literacy delay (^{Cuadro & Trías, 2008}; ^{Gordon, Fehd, & McCandliss, 2015}; ^{Jiménez & Ortiz, 2008}; ^{Kjeldsen, Karna, Niemi, Olofsson, & Witting, 2014}).

Over the last few decades, a growing number of researchers have suggested that musical abilities play an important role in linguistics abilities development, particularly in PA. It is important to highlight that these studies have been done in a diversity of languages (^{Anvari, Trainor, Woodside, & Levy, 2002}; ^{Bolduc & Montésinos-Gelet, 2005}; ^{Gordon et al., 2015}; ^{Hansen, Bernstorf, & Stuber, 2014}; ^{Kraus & Chandrasekaran, 2010}; ^{Lamb & Gregory, 1993}; ^{Lucas & Gromko, 2007}; ^{Peynircioğlu, Durgunoglu, & ÚneyKüsefogˇlu, 2002}; ^{Tierney & Kraus, 2013}). Several correlational studies have established a link between musical skills and PA, suggesting that music and speech may depend on many of the same basic auditory processes. Rhythmic and pitch patterns auditory analysis are similar to speech segmentation (^{Gordon et al., 2015}; ^{Goswami, 2012}a; Kraus & Chandrasekaran, 2010; ^{Patel, 2010}; ^{Tierney & Kraus, 2013}).

Among the pioneers in this field were ^{Lamb and Gregory (1993}), who assessed 16 4 and 5-year-old children with pitch and timbre discrimination tasks, phonemic awareness, and a simple reading test. Results showed that phonemic awareness correlated with simple reading ability. They also reported that pitch discrimination was significantly correlated with phonemic awareness. Subsequently, ^{Anvari et al. (2002}) examined the relations among PA, music perception skills, and early reading skills in a population of 100 4 and 5-year-old children. Subjects were given a battery of music tasks and a set of phonemic awareness tasks. Regression analyses indicated that music perception skills contributed unique variance in predicting reading ability, even when variance due to phonological awareness and other cognitive abilities (math, digit span, and vocabulary) had been accounted for. Interestingly, in 4-year-old children both rhythm production and rhythm discrimination predicted PA. In contrast, for the 5-year-old, only the pitch abilities were correlated with PA.

^{Peynircioğlu et al. (2002}) carried out a study on Turkish and North American preschoolers. Their study showed that perceptive melodic and rhythmic abilities were correlated with PA and recognition of pseudowords abilities, regardless of the children's mother tongue. In a subsequent study, ^{Bolduc and Montésinos-Gelet (2005}) examined the performance between PA, melodic perception, and rhythmic perception tasks in a group of French speaking Canadian kindergartners. Results indicated a significant correlation between scores obtained in the perceptive melodic tasks and the syllables and rhymes identification tasks. More precisely, they noticed that melodic perceptive abilities were by far more correlated with syllabic and rhythmic identification tasks. ^{Lucas and Gromko (2007}) found a significant correlation between rhythmic and musical pattern discrimination skills and phonemic awareness in first-grade North American 6-year old children.

Furthermore, there is evidence that children who participate in musical training programs not only enhance their musical skills, but also extend to their language skills. In fact, musical activities promote the development of auditory perception, and involve manipulating temporal and nonlinguistic sound structures unfolding in time. A similar operation is required when a phonological task is solved. This implicates speech sound analysis, as well as handling efficiently the inherent temporal structures of the linguistic units. Therefore, some children who are reading disabled exhibit sensory deficits associated with poor temporal processing. In addition, they show difficulties with musical rhythmic skills (^{Bolduc, 2009}; ^{Degé & Schwarzer, 2011}; ^{Dellatolas et al., 2009}; ^{Flaugnacco et al., 2014}; ^{François, Chobert, Besson, & Schön, 2012}; ^{Goswami, 2012}b; ^{Gromko, 2005}; ^{Moritz, Yampolsky, Papadelis, Thomson, & Wolf, 2013}; ^{Peretz, 2018}; ^{Slater at al., 2014}; ^{Tierney & Kraus, 2013}).

^{Herrera, Defior, and Lorenzo (2007)}, undertook a research to evaluate the effectiveness of two training programs on PA. One of the training programs included musical activities. Participants were divided into two groups. One group involved Spanish speaking preschoolers and the other Tamazight speaking children (a Berber dialect that is transmitted orally and is spoken in Morocco's Rif area). Overall, the results revealed that after the intervention phase the two experimental groups performed better than the control group on all phonological awareness tasks, regardless of their mother tongue. Moreover, the intervention combining phonological and musical activities was found to improve rhyme awareness.

^{Gromko (2005)} conducted a study comparing two groups of preschoolers, before and after they participated in a sixteen-week music program. Children who were involved in the musical programs showed better performance in letter word recognition and segmentation phonological tasks than the control groups.

^{Bolduc (2009}) analyzed the effect of a music training program on the development of PA in a group of French-Canadian kindergarten children. The treatment group participated in a specific musical training program, while the control group took part in the curricula music program. Results have shown that both music programs contributed similarly to the development of tonal and rhythmic perceptive skills. However, the experimental music training program proved to be more effective when it came to developing PA skills.

^{Degé & Schwarzer (2011}) compared the effects of a music program and a phonological skills program on PA in a group of preschoolers. Children were randomly assigned to a phonological skills program, a music program, or a control group that received sports training. Participants in both the phonological skills group and the music group showed significant increases in PA after 20 weeks of training. On contrary, children in the sports group maintained their initial phonological awareness tasks performance.

^{Moritz et al., (2013}), carried out a longitudinal study which analysed the relationship between kindergartners’ music rhythm skills and their PA in kindergarten and second grade. Children were assigned to different musical training programs that differed in range of activities administered, and workload. The experimental group received the Kodaly method, while the control group attended a musical workshop for singing nursery rhymes and learning basic musical vocabulary such as slow, fast, low and high, among others. Results indicated that children who received more music training during kindergarten showed improvement in a wider range of phonological awareness skills at the end of kindergarten. Twelve of them were followed up at the end of second grade. Results showed that rhythmic abilities developed in kindergarten were correlated with phonological skills in second grade.

^{Slater et al., (2014}), using a longitudinal design, compared the reading ability of 42 low-income, Spanish-English bilingual children aged 6-to-9. Children from low-socioeconomic backgrounds tend to fall progressively further behind their higher-income peers over the course of their academic careers. In this study, control variables such as reading abilities, intelligence, gender, age at which English was acquired, as well as English reading abilities and educational level of their parents were assessed. The experimental group comprised 23 children, who immediately started the music program once they were assessed. The remaining subjects integrated the control group. Children who received music training maintained their age-normed performance on the composite reading measure after 1 year, whereas children in the matched control group deteriorated over the same period of time, consistent with expected declines in this population. Researchers argued that while the extent of change is modest, the outcomes nonetheless provide evidence that music programs may have value in helping to counteract the negative effects of low-socioeconomic status on child literacy development. Further, early bilingual experience might provide advantages in PA development.

For the most part, these researches have been pursued in languages other than Spanish. A considerably lesser number of studies about PA stimulation through music have been carried out with Spanish-speaking children. It is worth noting that these studies have seldom focused on the development of rhythmic skills. This is so, even when it has been suggested that speech rhythm perception facilitates the identification of spoken words and also prompts the broadening of vocabulary in children, thus promoting the development of PA (^{Holliman, Wood, & Sheehy, 2010}; ^{Wood, Wade-Woolley, & Holliman, 2009}). In fact, children who exhibit low PA levels, or difficulties in the process of reading acquisition, also show difficulties in rhythmic skills and synchronization of musical tasks. (^{Goswami, 2012}b; ^{Goswami, Wang, Cruz, Fosker, Mead, & Huss, 2010}; ^{Flaugnacco et al., 2014}; ^{Lundetræ & Thomson, 2018}; ^{Moritz et al., 2013}; ^{Tierney & Kraus, 2013}; ^{Tierney, White-Schwoch, MacLean, & Kraus, 2017}; ^{Thomson, Huss & Goswami, 2006}).

To our knowledge, there are no studies in Spanish language that assess the results of a Program of PA Stimulation through musical activities focused on rhythm in pre-school typically developing children. It is therefore of the utmost relevance to produce empirical evidence for this population. The main objective of this study was to analyze the effects of a Program of PA Stimulation through a musical intervention focused on the development of rhythmic skills in a group of 5-year old Uruguayan children, so as to provide some additional evidence to Uruguayan Spanish language.

Results

Since the sample size was small, we used non-parametric techniques. This allows the groups to be compared without making the assumption that values are normally distributed. For comparisons between the groups Mann-Whitney U test and average ranges were used (^{Bologna, 2012}). As a central tendency, median is reported due to it being less affected by outliers. Percentiles 25 and 75 are reported (^{Shaughnessy et al., 2007}).

At the Time 2 (posttest), group comparisons were calculated using the gains in scores from LST (Linguistic Segmentation test) between T1-T2. Effect size from r de Rosenthal is reported.

Table 1 shows statistically significant differences between the groups in the gains in scores from LST (r= 0.69; p=.001). Significant differences in a set of phonological tasks were also observed: segmentation (r=0.80; p=.001); isolation (r=0.47; p=.029); deletion (r=.057; p=.008) and syllable blending (r=0.45; p=.043). No statistically significant differences were observed between the groups in recognizing words with the same final syllable sound (p=.085)

Table 1: Gains in scores from LST between T1-T2 and effect size 

When children entered first grade, they were reassessed with the task. This allowed us to observe whether differences between groups held over the long-term (Time 3). 

Tabla 2: LST scores and phonological performance tasks 

As observed in Table 2 statistically significant differences between the groups remained intact in LST total score (p=.003), segmentation (p=.002) and deletion tasks (p=.003). In this phase, there were no significant differences between the groups in isolation (p=.973), and syllable blending tasks (p=.173).

A year later the music program had finished (Time 4), 19 children (10 experimental group; 9 control group) were assessed with a reading word test (adapted from ^{Serrano, 2005}). Children were given 60s to read aloud as many words as possible from a list of 26 high-frequency words. The list began with words consisting of bisyllabic words and grew to include more complex words. At this time, a measure of word reading rate was obtained. Table 3 shows statistically significant differences between the groups in word reading rate (p=.035)

Table 3: Word Reading Rate