The relationship between phoneme awareness and reading proficiency, particularly in pre-reading stages, is a complex and debated topic. Phoneme awareness is thought to be the strongest predictor of reading skills, although its developmental nature is still not clear. The causal relationship between reading and phoneme awareness is said to be both predictive and reciprocal (Mann, 1993; Hulme et al., 2002; Clayton, et al., 2019). Longitudinal measures have shown that typical phoneme awareness tasks contribute differentially to this relationship (Cassady, Smith, & Putman, 2008). 

However, most studies so far explored the development of phoneme awareness in deeper orthographies, where phoneme awareness interacts with other factors like task characteristics and developmental status, making its predictive power less straightforward compared to more consistent orthographies (Landerl et al., 2018). On the other hand, children learning to read in shallow orthographies, such as Slavic languages, demonstrate better phoneme awareness compared to those learning in deep orthographies like English ​​(Caravolas, Volín, & Hulme, C., 2005; Spencer & Hanley, 2004). Studies on shallow orthographies have shown that they facilitate the development of phoneme awareness more effectively than deep orthographies. This is primarily due to the consistent grapheme-to-phoneme correspondences found in shallow orthographies. Studies suggest that children in shallow orthographies become accurate and fluent in reading before the end of the first school year, indicating that starting formal reading instruction around age 5 to 6 is effective (Seymour, Aro, & Erskine, 2003). This aligns with findings that emphasize the importance of an integrated approach to phonemic, orthographic, and morphemic knowledge during this period (McMurray, 2020). Having in mind that mastery of reading fluency before established benchmarks is a significant predictor of later reading skills, early intervention and development of reading fluency may be crucial for preventing reading difficulties (Park, Chaparro, Preciado, & Cummings, 2015). Moreover, early phonological processing skills scaffold later reading skills, with better early reading ability reducing the cognitive effort required for phonemic access (Wang, Pines, Joanisse, & Booth, 2021).

Longitudinal studies have shown that early language and pre-literacy skills are strong predictors of reading as far as PISA reading measurements (Eklund et al., 2018). Serbia shows stable below average PISA results in reading during the past 20 years, since its first participation (OECD, 2003-2023). Despite the one-to-one grapheme-to-phoneme correspondences, formal reading acquisition in Serbia doesn’t begin until first-grade elementary school, where most children are 6-7 years old. This study is part of a Ph.D. thesis that focuses on exploring phoneme awareness skills in preschoolers in Serbia to understand whether this benchmark requires revision and potential shifting of reading instruction onset to preschool in order to boost reading proficiency of Serbian children. Another aim of this study is to establish critical periods when preschool children, based on their phoneme awareness and synthesis development, would be capable of engaging in reading instructions.

This study focuses on exploring phonemic analysis and synthesis tasks in preschool, pre-reading children in a shallow orthography (Serbian) to determine the dynamics of its development in children aged 3.5 to 6.5 years (N = 250, 48% girls). Children were recruited from four private kindergartens in Belgrade. For phoneme awareness tasks, phoneme segmentation tasks were constructed due to being the strongest predictors of reading skills. Since at the time of data collection there were no official phoneme awareness instruments in Serbian language, we constructed five parallel forms that consisted of two groups of tasks - phoneme analysis and phoneme synthesis tasks. Each group consisted of 4 tasks (5 items per task), resulting in 40 items per form. Testing was performed by three examiners. For each item, the examiner would note if the child responded correctly (1) or incorrectly (0). Four typical phonemic analysis tasks were used - identifying the first, last, second, and all phonemes in a word. Another four tasks were used to test phonemic synthesis - synthesizing 3,4,5 and 6 phonemes. Five parallel forms were balanced and distributed equally (roughly N=50 per form), controlling for word length and familiarity. To balance word categories, three of the words in each task were nouns, one was a verb, and one was an adjective. For phonemic analysis tasks, the length of the words was varied so that each set had one word of 3 letters, one word of 4 letters, two words of 5 letters, and one word of 6 letters. The two words of 5 letters were never of the same word type (i.e. one would be a noun, while the other one would be either a verb or an adjective). All words had a CVCV or VCVC structure to control for phonological complexity. The internal consistency of this short-version phoneme segmentation test was high (Cronbach’s alpha = .96), as well as the reliability of phoneme analysis (Cronbach’s alpha = .92) and synthesis scales (Cronbach’s alpha = .94). There were no significant differences in correctness rates between forms neither in phoneme analysis tasks (p=.143), nor phoneme synthesis tasks (p=.578), so further analyses were conducted on all items together (100 items per scale) to improve statistical power.

Significant developmental differences were found in phonemic awareness abilities among pre-reading children. Success rates varied significantly between age groups, with the most substantial differences observed between the oldest (5.5-6.5) and youngest (3.5-4.5) groups. To understand the developmental curve of phoneme awareness in Serbian language, multiple binary logistic regressions were run in R using the glm function with age (in months) as the predictor and binary results of each task as criteria. Two different criteria (liberal and conservative) were used to establish at what age the shift from no phoneme awareness to developed phoneme awareness was most likely to happen. The liberal criteria treated total task scores 0-2 as no phoneme awareness, while scores 3-5 were treated as developed phoneme awareness, while the conservative criteria treated total scores 0-4 as no phoneme awareness, and only all correct answers (5) were treated as developed phoneme awareness. The 0.5 logistic curve flexion point was determined as the likelihood shift. According to the liberal criteria, the likelihood shift in phoneme analysis happens between 53 and 69 months, while for phoneme synthesis this shift happens later, between 65 and 74 months. The conservative criteria shift the likelihood in both phoneme analysis and synthesis a couple of months later, ranging between 58 and 72 months and 69 and 78 months for phoneme analysis and synthesis, respectively. When both criteria are taken together, the likelihood shift in phoneme awareness development happens between 4.4 and 6.5 years. These results show for the first time that phoneme awareness in preschool children, whose native language is Serbian, may be well developed up to 1.5 years earlier than the official start of reading instructions in Serbia. Implications from these results can have a significant impact on the future of preschool literacy in Serbia and across the Balkans.

Mann, V. (1993). Phoneme Awareness and Future Reading Ability. Journal of Learning Disabilities, 26, 259 - 269. https://doi.org/10.1177/002221949302600406. Clayton, F., West, G., Sears, C., Hulme, C., & Lervåg, A. (2019). A Longitudinal Study of Early Reading Development: Letter-Sound Knowledge, Phoneme Awareness and RAN, but Not Letter-Sound Integration, Predict Variations in Reading Development. Scientific Studies of Reading, 24, 107 - 91. https://doi.org/10.1080/10888438.2019.1622546. Hulme, C., Hatcher, P., Nation, K., Brown, A., Adams, J., & Stuart, G. (2002). Phoneme awareness is a better predictor of early reading skill than onset-rime awareness.. Journal of experimental child psychology, 82 1, 2-28 . https://doi.org/10.1006/JECP.2002.2670. Cassady, J., Smith, L., & Putman, S. (2008). Phonological Awareness Development as a Discrete Process: Evidence for an Integrative Model. Reading Psychology, 29, 508 - 533. https://doi.org/10.1080/02702710802271966. Spencer, L., & Hanley, R. (2004). Learning a Transparent Orthography at Five Years Old: Reading Development of Children during Their First Year of Formal Reading Instruction in Wales.. Journal of Research in Reading, 27, 1-14. https://doi.org/10.1111/J.1467-9817.2004.00210.X. Caravolas, M., Volín, J., & Hulme, C. (2005). Phoneme awareness is a key component of alphabetic literacy skills in consistent and inconsistent orthographies: evidence from Czech and English children.. Journal of experimental child psychology, 92 2, 107-39 . https://doi.org/10.1016/J.JECP.2005.04.003. Landerl, K., Freudenthaler, H., Heene, M., De Jong, P., Desrochers, A., Manolitsis, G., Parrila, R., & Georgiou, G. (2018). Phonological Awareness and Rapid Automatized Naming as Longitudinal Predictors of Reading in Five Alphabetic Orthographies with Varying Degrees of Consistency. Scientific Studies of Reading, 23, 220 - 234. https://doi.org/10.1080/10888438.2018.1510936. Seymour, P., Aro, M., & Erskine, J. (2003). Foundation literacy acquisition in European orthographies.. British journal of psychology, 94 Pt 2, 143-74 . https://doi.org/10.1348/000712603321661859. Park, Y., Chaparro, E., Preciado, J., & Cummings, K. (2015). Is Earlier Better? Mastery of Reading Fluency in Early Schooling. Early Education and Development, 26, 1187 - 1209. https://doi.org/10.1080/10409289.2015.1015855. Wang, J., Pines, J., Joanisse, M., & Booth, J. (2021). Reciprocal relations between reading skill and the neural basis of phonological awareness in 7- to 9-year-old children. NeuroImage, 236, 118083 - 118083. https://doi.org/10.1016/j.neuroimage.2021.118083. Eklund, K., Torppa, M., Sulkunen, S., Niemi, P., & Ahonen, T. (2018). Early cognitive predictors of PISA reading in children with and without family risk for dyslexia. Learning and Individual Differences, 64, 94-103. https://doi.org/10.1016/J.LINDIF.2018.04.012.