Developmental Characteristics of Phonological Awareness in Hearing-Impaired Preschool Children with Cochlear Implants in China

Abstract

This study analyzes the current state and developmental characteristics of phonological awareness in hearing-impaired preschool children with cochlear implants aged 3 to 5 years in China. The phonological awareness development of hearing-impaired preschool children is assessed via a comparison with normal-hearing preschool children of the same age, utilizing a number of key metrics and statistical analyses to determine any differences in the developmental characteristics between the two groups. The results show that the phonological awareness development of Chinese-speaking hearing-impaired preschool children with cochlear implants follows, for the most part, the progression of their normal-hearing preschool counterparts, albeit at a lower level. Identifying phonological awareness profiles in children with cochlear implants helps improve the accuracy of assessment and supports the development of targeted intervention strategies. This study aims to provide a clearer understanding of their phonological processing abilities.

1. Introduction

Phonological awareness is a foundational skill in early childhood that plays a critical role in literacy development, particularly in reading and writing. In contemporary early childhood education, letters and visual symbols are often introduced in kindergarten, providing children with early exposure to multiple forms of communication. These experiences support the gradual acquisition of reading and writing abilities, with phonological awareness serving as a key prerequisite. Phonological awareness broadly refers to the capacity to recognize and manipulate the sound structures of spoken language—including phonemes, syllables, and words—without relying on written text [1,2]. This ability develops progressively, beginning with sensitivity to larger phonological units such as words and syllables, advancing to rhyme awareness, and eventually reaching phonemic manipulation [3]. The preschool years and early primary school stages represent a sensitive period for acquiring phonological awareness. Phonological awareness has been identified as a key linguistic predictor of early reading success [4,5,6]. Early mastery of this skill strongly predicts future reading proficiency [7]. Higher levels of phonological awareness facilitate more efficient mapping between spelling and sound at the sub-lexical level, aiding in decoding unfamiliar words and supporting the development of orthographic knowledge, which enables faster and more accurate word recognition [8,9].

While much research on phonological awareness has focused on alphabetic languages, the Chinese phonological system has distinct characteristics that shape the development of phonological awareness in Chinese-speaking children. Unlike alphabetic languages, where phonemes correspond directly to letters, the fundamental phonological unit in Chinese is the syllable, which typically corresponds to a single character composed of three parts: an initial consonant (onset), a final (vowel plus sometimes nasal ending), and a tone. Tones are a unique and essential phonological feature in Chinese—pitch variations that distinguish meaning among syllables sharing the same consonants and vowels—and are critical for accurate pronunciation and comprehension. Mandarin Chinese has four primary tones [10]. Although Chinese characters do not directly encode phonemes, phonological awareness remains essential, particularly through Pinyin, a romanization system that connects characters to pronunciation via letter–phoneme correspondence [11].

Building on the unique features of the Chinese phonological system, researchers have examined how typically developing children acquire phonological awareness (PA) skills throughout early childhood. Children first develop the ability to distinguish between similar and dissimilar-sounding words before mastering the manipulation of specific sounds. Generally, blending skills develop before segmentation skills, even when both have similar linguistic complexity [12]. Recent studies have revealed distinct developmental trajectories across different PA components in Chinese-speaking children. Shu et al. (2008) [13] conducted two empirical studies from preschool through early schooling. Their first study (N = 146) showed that phoneme onset awareness was at chance level between ages 3 and 5 but improved to about 70% accuracy after Grade 1 following formal Pinyin instruction. Tone awareness appeared above chance from age 4 and improved significantly after school entry, reaching 74% accuracy. Syllable and rime awareness developed more gradually and steadily from age 3. These findings suggest that PA components are influenced by factors such as cognitive maturation and formal literacy instruction. In their second study (N = 202), the authors found that syllable awareness, tone awareness, and rapid automatized naming significantly predicted Chinese character recognition, highlighting their critical role in early literacy development. Although these findings have deepened our understanding of PA development in typically developing children, far less is known about how these skills emerge in children with hearing loss, particularly during the critical preschool years.

To further deepen the understanding of how different components of phonological awareness contribute to literacy acquisition over time, researchers have conducted longitudinal studies tracking children’s development across multiple years. A recent eight-year longitudinal study analyzed related variables in depth. The present study reports data on phonological awareness, morphological awareness, and Chinese literacy skills of 294 children from this eight-year longitudinal study. The results showed that preliterate syllable awareness at ages 4 to 6 uniquely predicted morphological awareness at ages 7 to 10 and directly contributed to character reading and writing at age 11. However, post-literate phonemic awareness only predicted character reading at age 11. Moreover, early syllable and morphological awareness indirectly affected later reading fluency and comprehension via morphological awareness. These findings emphasize the unique relationship between syllable and morphological awareness in Chinese, both of which are focused on the syllable unit, which maps directly to the character—the fundamental unit of Chinese reading. This highlights the significant role of Chinese-specific phonological structures in literacy and higher-level reading skills [14]. Early syllable awareness uniquely predicts later morphological awareness and directly supports Chinese character reading and writing, highlighting the crucial role of Chinese-specific phonological structures in literacy development.

Beyond typical developmental trajectories, children with hearing loss—particularly cochlear implant (CI) users—face unique auditory challenges that affect their language development. While many existing studies on Mandarin-speaking CI users have primarily focused on speech perception, phonological awareness (PA) has received comparatively less attention. However, PA plays a critical role in early literacy acquisition and is closely linked to both speech perception and higher-order language processing [15]. Recent studies have begun to fill this gap, providing emerging evidence on PA development in Chinese children with hearing loss. For instance, Chan (2023) [16] tracked 28 Grade 1–2 CI users, assessing PA and morphological awareness (MA) in early Grade 1, vocabulary knowledge at the end of Grade 1, and reading comprehension at the end of Grade 2. She found that both PA and MA predicted vocabulary growth, and that vocabulary fully mediated their impact on reading comprehension—highlighting that even when PA catches up, its benefits are exerted indirectly through lexical development. This underscores the foundational role of PA in broader literacy outcomes.

Extending this inquiry to preschoolers, Zhang et al. (2022) [17] compared 16 Mandarin-speaking CI children aged 3–5 with 16 age-matched normal-hearing (NH) peers, using syllable-and-prosody detection tasks and auditory digit-span tests. Their results showed that early-implanted CI users performed on par with their hearing peers in PA and they demonstrated a strong correlation between PA and working memory. These findings suggest that once a certain threshold of perceptual access is achieved, cognitive resources such as memory capacity become critical to phonological skill acquisition. In a Cantonese context, Tse and So (2012) [18] provided further insights into the tonal challenges faced by CI users. While these children showed typical performance on syllable, phoneme, and rhyme awareness, they exhibited significant delays in tone awareness and phonological knowledge. These delays were attributed to the limited pitch resolution of cochlear implants, a constraint that is especially consequential in tonal languages such as Cantonese. Recognizing these challenges, researchers have begun testing structured interventions aimed at enhancing PA. Chen and Yi (2014) [19], for example, conducted a 26-week PA training program for 20 deaf and hard-of-hearing preschoolers in northern Taiwan. Compared with a group receiving standard aural rehabilitation, the PA-trained children made significant gains in both phonological tasks and early reading skills. These results strongly support the inclusion of systematic PA training in early language intervention programs for Mandarin-speaking children with hearing impairments. Taken together, these studies shift the focus from basic auditory perception to higher-level phonological processing, emphasizing the importance of PA in shaping long-term literacy outcomes. While speech perception provides the auditory input necessary for language acquisition, it is PA that enables children to consciously analyze and manipulate speech sounds. Understanding the interaction between these domains is essential for designing effective aural rehabilitation programs that go beyond perception to target literacy and cognitive-linguistic development. While intervention studies demonstrate that structured PA training pays off, optimizing these programs requires understanding why they work. Zhao et al. (2019) [20] examined Grades 3–6 and showed that vocabulary knowledge consistently predicts reading fluency across hearing statuses, whereas PA’s predictive power varies by age and auditory experience—older CI users and younger NH children benefit most directly. This aligns with Chan’s mediation finding and suggests that vocabulary building should go hand-in-hand with PA training. More broadly, PA also underpins higher-level language skills: Tan and Song (2024) [21] showed its role in Chinese focus constructions, a task demanding cognitive flexibility. Finally, Zhang et al.’s [17] PA working-memory link highlights that children’s capacity to hold and manipulate sounds influences how readily they internalize phonological patterns. Together, developmental comparisons, intervention trials, and mechanism studies converge on one clear prescription: effective aural rehabilitation for Mandarin-speaking CI users should integrate perceptual (segmental and tonal) training, structured PA exercises, vocabulary enrichment, and working-memory supports. This multifaceted approach promises the greatest gains in both early literacy and long-term language outcomes.

Despite a growing body of research on speech perception in children with hearing impairments, the number of studies specifically investigating phonological awareness (PA) in this population remains extremely limited. This scarcity is particularly acute among preschool-aged children (3 to 5 years old), a critically sensitive period for phonological and language development, yet one that has been largely overlooked. Most existing research focuses on school-aged children (6 years and above) and predominantly examines speech perception abilities such as tone and segmental sound recognition. However, phonological awareness is a distinct construct that involves the explicit recognition and manipulation of sound units, which are foundational for early literacy acquisition. This distinction is crucial, as PA reflects deeper phonological processing abilities and is a moderate-to-strong predictor of speech perception and reading success. Unfortunately, the majority of current assessment tools are adapted from English-language models and fail to capture the unique phonological characteristics of Mandarin Chinese—particularly its syllable-based structure, including initials, finals, and lexical tones. As a result, there is a pressing need for localized, systematic evaluation methods tailored specifically for Mandarin-speaking preschoolers with cochlear implants (CIs). To address these critical gaps, the present study focuses on the development of phonological awareness—especially tone and syllable awareness—in Mandarin-speaking preschool children with CIs. It also examines influential factors, such as auditory experience, age at implantation, and the quantity and quality of language input, using assessment tools designed to align with Mandarin’s phonological system. This research aims to fill the significant void in early PA research within this underserved population, advancing both theoretical understanding and practical approaches for early intervention.

This study seeks to achieve the following objectives: (1) To examine the current status of phonological awareness in infants with CIs by comparing their performance to that of NH infants of the same age, focusing on four key aspects: awareness of initial sounds, awareness of medial sounds, awareness of final sounds, and the ability to discriminate the four lexical Chinese tones. (2) To elucidate the developmental characteristics of phonological awareness in infants with CIs by analyzing age-related patterns in comparison to those observed in NH infants. By addressing these objectives, the study is expected to generate valuable insights into the developmental profile of phonological awareness in children with hearing impairments. These findings will serve as foundational data for early literacy intervention and support programs tailored to the needs of this population.

2. Methods

2.1. Participants

The participants in this study consisted of two groups: preschool children with cochlear implants (CI preschool children) and preschool children with normal hearing (NH preschool children). The CI preschool children were randomly recruited from five different rehabilitation centers for special education programs for deaf children in China, provided that all children were cochlear implant users and capable of spoken communication. NH preschool children were randomly recruited from preschools for general education in China. All children were aged between 3 and 5 years. Each group included a total of 90 participants, with 30 selected from each age group (3, 4, and 5 years old) (see

Table 1. Participants’ information.

		CI Preschool Children (n = 90)	NH Preschool Children (n = 90)
Age	3	30	30
	4	30	30
	5	30	30
Gender	Girls	43 (47.78%)	44 (44.89%)
	Boys	47 (52.22%)	46 (51.11%)
Average Hearing level	Profound hearing loss (90 dB or more)	57 (63.34%)	--
	Severe hearing loss (71 dB to 90 dB)	21 (23.37%)
	Moderately severe hearing loss (56 dB to 70 dB)	11 (13.29%)

Note: Hearing levels in China: mild hearing loss: 26 dB–40 dB; moderate hearing loss: 41 dB–55 dB; moderately severe hearing loss: 56 dB–70 dB; severe hearing loss: 71 dB–90 dB; profound hearing loss: over 90 dB. CI: cochlear implant; NH: normal-hearing.

). None of the participants had received any form of emergent literacy instruction prior to this study. All preschool children in the CI group were able to communicate by spoken language.

Among the CI preschool children, 57 had bilateral cochlear implants, while 32 had unilateral implantation. Of those with unilateral implantation, 21 CI preschool children wore a hearing aid in the non-implanted ear. Bilateral implantation was performed sequentially. For those with sequential implantation, the interval between the two surgeries ranged from 6 to 12 months. Most participants had symmetric hearing loss prior to implantation. Participants received cochlear implantation at an average age of 2 years. At the time of testing, all participants had a minimum of one year of cochlear implant experience. At the rehabilitation center, CI preschool children received daily one-on-one speech and pronunciation training and guidance from their teachers; all participants communicated primarily through spoken language.

Among the NH preschool children, hearing status was based on parental reports and school records indicating no known hearing concerns. However, no formal audiological screening was conducted, which is acknowledged as a limitation of this study.

All CI preschool children were raised in predominantly monolingual Mandarin-speaking environments with no reported exposure to sign language. The cochlear implant systems used included devices from Cochlear Limited (Sydney, Australia), Nurotron (Hangzhou, China), and MED-EL (Innsbruck, Austria), with regular mapping and device maintenance performed by clinical audiologists. In addition to daily speech and pronunciation training, a subset of children participated in optional music therapy sessions, which have been shown to support auditory and phonological development.

Using G*Power 3.1, an a priori power analysis was performed for a 2 × 3, between-subjects ANOVA (α = 0.05; medium effect size, f = 0.25; targeting the interaction effect). The results indicated a required total sample size of approximately N = 158, corresponding to at least 26 participants per cell, to achieve ≥ 0.80 statistical power. In this study, 180 participants were recruited, with 90 in each group and 30 per age level (3, 4, and 5 years).

The study was conducted in accordance with the Declaration of Helsinki and was approved by the Ethics Committee of Tsukuba University (Approval No. [Tsukuba NO. 2023-175A]), and written informed consent was obtained from the parents or legal guardians of all participants.

2.2. Developmental Screening Procedure

To exclude general developmental delays among participants, all preschool children underwent assessment using the Chinese version of the Ages & Stages Questionnaires, Third Edition (ASQ-3), a standardized developmental screening instrument designed for children aged 1 month to 5 years. The ASQ-3 encompasses five developmental domains: communication, gross motor skills, fine motor skills, problem-solving, and personal–social development. Each domain comprises six items, scored on a 10-point scale, yielding a maximum domain score of 60 and a total possible score of 300. Parents completed the questionnaire based on systematic observation of their children’s daily behaviors.

Psychometric evaluation of the ASQ-3 (Chinese version) within a normative Chinese sample has demonstrated satisfactory reliability and validity. The internal consistency reliability, as indicated by Cronbach’s alpha, was 0.80, reflecting adequate homogeneity across items. Inter-rater reliability, assessed via Pearson’s correlation coefficient, was 0.80. Furthermore, independent completion of the ASQ-3 by parents and professionals yielded a significant correlation in total scores (r = 0.80, p < 0.001). Domain-specific correlations between parent and professional ratings were robust: communication (r = 0.80), gross motor (r = 0.90), fine motor (r = 0.80), problem-solving (r = 0.80), and personal–social (r = 0.80) all attained p < 0.001 [22].

Consequently, the ASQ-3 (Chinese version) was deemed a reliable and valid instrument for assessing the developmental status of the participants in the present study.

2.3. Assessment of Phonological Awareness

The assessment included four subtests commonly used in Chinese language research: (1) initial sound awareness, which refers to the ability to perceive and identify the initial syllables of a word; (2) medial sound awareness, which refers to the ability to perceive and identify the syllables of a word; (3) final sound awareness, which refers to the ability to perceive and identify the ending syllables of a word; and (4) tone discrimination (four tones). Each subtest consisted of 30 items, with each item scored as either correct (1 point) or incorrect (0 points), yielding a maximum score of 30 points per subtest and a total maximum score of 120 points for the full assessment. The stimuli for the test were three-syllable words represented by pictures. For each item, the preschool children were first asked to identify the number of sounds in the words, and then state the initial sound, medial sound, and final sound of the word shown in the image, as well as indicate the four tones associated with the word. The experimental setup is shown in Appendix A.

The four phonological awareness tests used in this study were adapted from established instruments [23] to accommodate the linguistic characteristics of Chinese preschoolers. To assess the reliability of the phonological awareness tasks used in this study, internal consistency (Cronbach’s α) and test–retest reliability analyses were conducted. The results showed that the overall internal consistency of the measures was high, with a Cronbach’s α coefficient of 0.83, indicating good internal consistency. Additionally, the α coefficients for specific subtasks (e.g., initial sound awareness, media sound awareness, ending sound awareness, four tones) ranged from 0.66 to 0.92, further demonstrating strong consistency among the subcomponents.

To verify the validity of the measurement tool, content validity was first ensured through expert review, confirming that the test content aligned with the research objectives and theoretical framework. Subsequently, the Kaiser–Meyer–Olkin (KMO) measure of sampling adequacy (KMO = 0.70) and Bartlett’s Test of Sphericity (χ² = 322.98, df = 6, p < 0.001) were used to assess the suitability of the data for factor analysis. Exploratory factor analysis revealed a single-factor structure with all item loadings above 0.60, supporting the structural validity of the test.

Phonological awareness was assessed individually in a quiet room to minimize distractions. The testers were professionally trained and possessed the relevant qualifications.

Testing was conducted individually, with each child’s session lasting approximately 30 min, varying according to individual differences. Standardized software and hardware were used to present audio stimuli in a randomized order. Scoring was performed manually by assessors blinded to participant groupings. Inter-rater reliability was assessed, yielding a satisfactory Cohen’s kappa of 0.848.

2.4. Data Analysis

The analysis aimed to compare phonological awareness between CI preschool children and NH preschool children across total and subtest scores. First, scores from the ASQ-3 were used to confirm that the developmental status of the CI and NH groups fell within the typical range. Next, the normality of the data was assessed using a Shapiro–Wilk test. The results indicated that the data did not meet the assumptions of a normal distribution. Additionally, a Levine’s test was conducted to assess the homogeneity of variance, which showed that the variances were unequal across groups (initial sound awareness: p = 0.036 < 0.05; medial sound awareness: p < 0.001; ending sound awareness: p < 0.001; four-tone awareness: p < 0.001).Therefore, for subsequent analyses, a Generalized Linear Mixed Model (GLMM) were conducted to examine group differences in the total phonological awareness scores as well as in each of the four subtests in to determine the current developmental status of phonological awareness in the CI group relative to their NH peers. To investigate age-related developmental patterns, GLMM tests were performed to analyze trends in phonological awareness across the three age groups (3, 4, and 5 years) within each group. Comparisons with NH children allowed for a clearer understanding of the developmental characteristics of phonological awareness in the CI group, including the identification of which specific subskills (e.g., initial sound, tone discrimination) emerge earlier or later with age.

3. Results

3.1. Preschool Developmental Screening Test of CI and NH Preschool Children

Descriptive analyses were conducted to summarize the sample characteristics and performance across tasks. Table 1 presents the means and standard deviations for each phonological awareness task by group and age.

Regarding the results of the ASQ-3,

Table 2. Scores of the Infant Development Test Screening for NH preschool children and CI preschool children.

	Age	NH Preschool Children (n = 90)	CI Preschool Children (n = 90)	p
		Average Score (SD)	Average Score (SD)
Communication	3	52.50 (5.21)	25.86 (2.09)	0.000 *
	4	54.50 (4.42)	28.16 (5.17)	0.000 *
	5	55.33 (5.24)	30.83 (5.83)	0.000 *
Gross motor skills	3	54.33 (6.40)	52.41 (7.86)	0.435
	4	55.16 (4.04)	53.50 (3.97)	0.225
	5	55.67 (5.04)	53.83 (4.85)	0.094
Fine motor skills	3	54.50 (7.35)	51.55 (8.67)	0.082
	4	54.66 (5.07)	54.50 (5.14)	0.147
	5	54.67 (4.34)	54.68 (4.34)	0.425
Problem-solving abilities	3	53.66 (5.56)	52.58 (9.31)	0.589
	4	56.16 (3.87)	53.33 (5.13)	0.369
	5	57.17 (4.49)	54.16 (4.37)	0.356
Personal–social development	3	54.83 (4.99)	53.27 (8.05)	0.187
	4	56.00 (4.23)	54.33 (5.83)	0.527
	5	57.50 (4.31)	55.16 (4.45)	0.364
Total scores	3	269.83 (21.71)	235.67 (30.79)	0.000 *
	4	276.50 (17.62)	243.83 (13.94)	0.000 *
	5	280.67 (13.88)	248.67 (14.97)	0.000 *

Note: * p < 0.05. CI: cochlear implant; NH: normal-hearing.

presents the descriptive statistics for all participants, showing that preschool children with cochlear implants (CIs) demonstrated age-appropriate development in all domains except communication. Specifically, the average language communication score for 3-year-old children with normal hearing (NH) was 52.5, whereas CI children scored significantly lower at 26. This substantial difference suggests a clinical delay in language development among CI children. Although cochlear implants provide auditory input, the quality of auditory perception differs markedly from that of NH peers, particularly during the critical sensitive period for language acquisition, where insufficient auditory stimulation may hinder rapid language development [24].

Moreover, this deficit in language communication appears to impact phonological awareness development, a well-established predictor of language comprehension and literacy skills [17]. Consequently, the reduced language abilities in CI children may constrain their phonological awareness performance, potentially affecting their future reading and writing development [25].

Importantly, our findings indicate that CI preschool children do not exhibit significant developmental delays outside the communication domain. However, as the ASQ-3 is a parent-reported screening tool, its results should be interpreted with caution and cannot substitute for in-depth clinical assessments in other domains. This conclusion aligns with prior research, such as Geers and Nicholas (2013) [26], who reported that with adequate language-rich environments and early intervention, CI children show no necessary delays in cognitive or social development. Similarly, Kral and Sharma (2012) [27] emphasized the importance of early implantation for optimal language outcomes and noted that other developmental areas may be comparable to those of hearing peers.

In summary, our data-driven conclusion, consistent with previous studies, suggests that preschool children with cochlear implants experience language-specific developmental delays but do not show additional delays in other developmental domains. This highlights the critical need for early and continuous language intervention, with particular focus on phonological awareness training, to support comprehensive language and literacy development in this population.

3.2. The Current Status of Phonological Awareness Development

To examine the differences in phonological awareness scores between preschool children with cochlear implants (CIs) and normal-hearing (NH) preschool children, a generalized linear mixed model (GLMM) analysis was first conducted based on formal evaluations of phonological awareness and its four subdomains. The results are presented in

Table 3. Comparison of phonological awareness task performance between children with hearing impairment and children with normal hearing.

	Ages	CI Preschool Children (n = 90)	NH Preschool Children (n = 90)	p
		Average Score (SD)	Average Score (SD)
Total scores	3	27.89(7.45)	63.62(9.53)	0.000 *
	4	45.90(17.09)	82.81(15.91)	0.000 *
	5	64.82(21.06)	104.84(7.98)	0.000 *
Initial sound awareness	3	5.93(1.93)	20.33(4.08)	0.000 *
	4	10.95(6.35)	25.23 (4.05)	0.000 *
	5	17.87(6.89)	27.63(2.54)	0.000 *
Medial sound awareness	3	3.10(1.53)	11.37(3.05)	0.000 *
	4	7.03(5.85)	17.23(5.86)	0.000 *
	5	11.53(6.96)	25.13(2.74)	0.000 *
Ending sound awareness	3	9.21(3.96)	14.77(3.22)	0.000 *
	4	15.77(4.94)	21.10(5.46)	0.000 *
	5	18.67(6.89)	28.40(2.16)	0.000 *
Four-tone discrimination	3	9.65(1.43)	17.15(1.80)	0.000 *
	4	12.15(1.89)	19.25(2.15)	0.000 *
	5	16.75(1.95)	23.68(1.43)	0.000 *

Note: * p < 0.05. CI: cochlear implant; NH: normal-hearing.

.

The GLMM revealed significant main effects of group (CIs vs. NH), age, and task type on phonological awareness performance. Specifically, children with normal hearing showed significantly higher accuracy than children with cochlear implants (β = 1.85, SE = 0.45, p < 0.001). Age was positively associated with performance (β = 0.75, SE = 0.20, p < 0.001), indicating improvement with increasing age. Significant two-way interactions were observed between group and task type (β = −0.62, SE = 0.18, p = 0.001), suggesting that the difference in performance between groups varied depending on the phonological task, with the negative coefficient indicating a reduction in the group difference in certain tasks. Additionally, a significant three-way interaction of group × age × task type was found (β = 0.40, SE = 0.15, p = 0.01), indicating that group differences in specific tasks changed with age. Covariates, including age at implantation and duration of CI use, were also included in the model, and showed significant effects (p < 0.05), with earlier implantation age and longer CI use associated with better phonological awareness performance.

Normality tests (e.g., Shapiro–Wilk) were conducted in the preliminary analysis; however, the data did not meet the assumption of a normal distribution. Therefore, nonparametric tests were initially used, and generalized linear mixed models were applied to better handle the data structure. Descriptive statistics, including the means and standard deviations for all groups and tasks, are presented in Table 3 to provide a clear overview of the data distribution. When comparing age groups, multiple comparisons were conducted using appropriate post hoc tests with correction for multiple testing; the detailed results are presented in

Table 4. Analysis of changes in phonological awareness with increasing age.

Participants	Items	Ages	N	p	Changes
CI preschool Children	Total scores	3	30	0.000
		4	30
		5	30
	Initial sound awareness	3	30	0.000
		4	30
		5	30
	Medial sound awareness	3	30	0.000
		4	30
		5	30
	Ending sound awareness	3	30	0.000
		4	30
		5	30
	Four-tone discrimination	3	30	0.000
		4	30
		5	30
NH Preschool Children	Total scores	3	30	0.000
		4	30
		5	30
	Initial sound awareness	3	30	0.000
		4	30
		5	30
	Medial sound awareness	3	30	0.000
		4	30
		5	30
	Ending sound awareness	3	30	0.000
		4	30
		5	30
	Four-tone discrimination	3	30	0.000
		4	30
		5	30

Note: * p < 0.05.

. Regarding p-value calculation, all p-values reported from the GLMM analyses were derived using likelihood ratio tests comparing nested models, ensuring robust statistical inference.

The total score for phonological awareness was 120 points, comprising four subdomains: initial sound awareness (30 points), medial sound awareness (30 points), ending sound awareness (30 points), and four-tone discrimination (30 points).

Across all age groups (3, 4, and 5 years), preschool children with cochlear implants (CI group) scored significantly lower than their normal-hearing peers (NH group) in overall phonological awareness. Results from the generalized linear mixed model (GLMM) revealed a significant main effect of hearing status, with children with CIs scoring, on average, 37.37 points lower than NH children on the 120-point scale (β = −37.37, SE = 3.67, p < 0.001). This substantial difference underscores a marked delay in phonological development in children with CIs.

Preschool children with cochlear implants (CIs) scored significantly lower than their normally hearing (NH) peers on initial sound awareness across all age groups (ages 3, 4, and 5; p < 0.05). GLMM analyses revealed a significant main effect of hearing status (β = −13.75, SE = 1.37, p < 0.001), indicating that CI children consistently lagged behind NH children in the development of phonological awareness.

Preschool children with CIs scored significantly lower than their NH peers on medial sound awareness across all age groups (ages 3, 4, and 5; p < 0.05). GLMM analyses revealed a significant main effect of hearing status (β = −12.84, SE = 1.61, p < 0.001), indicating that the development of phonological awareness in children with CIs lagged behind that of NH children.

Preschool children with CIs scored significantly lower than their NH peers on ending sound awareness across all age groups (ages 3, 4, and 5; p < 0.05). GLMM analyses revealed a significant main effect of hearing status (β = −8.60, SE = 1.49, p < 0.001), confirming that the development of phonological awareness in children with CIs lagged behind that of NH children.

Finally, preschool children with CIs scored significantly lower than their NH peers on four-tone discrimination across ages 3, 4, and 5 (p < 0.05). GLMM analyses revealed a significant main effect of hearing status (β = −7.56, SE = 1.51, p < 0.001), confirming that phonological awareness development in preschool children with CIs lagged behind that of NH preschool children.

3.3. The Developmental Characteristics of Phonological Awareness in CI Preschool Children

To examine the age-related changes in phonological awareness between CI and NH preschool children, The generalized linear mixed model (GLMM) was also conducted to evaluate the differences across age groups. The results are presented in Table 4.

For the total phonological awareness scores (see Table 4), GLMM analyses revealed significant age-related differences among NH preschool children. Specifically, relative to age 5 (the reference group), children at age 3 exhibited significantly lower scores (β = −36.22, SE = 4.80, p < 0.001), as did those at age 4 (β = −16.93, SE = 4.47, p < 0.001), indicating marked improvement in phonological awareness with increasing age. Notably, the most substantial gains were observed between ages 3 and 4. Similarly, preschool children with CIs showed significant developmental progress across the same age range. Compared with age 5, children at age 3 had significantly lower scores (β = −39.85, SE = 2.82, p < 0.001), as did those at age 4 (β = −23.60, SE = 2.68, p < 0.001). However, unlike NH children whose growth was concentrated between ages 3 and 4, the improvement in children with CIs appeared more gradual and continuous across ages 3 to 5, as indicated by the GLMM results.

For initial sound awareness, GLMM analyses revealed significant differences among NH preschool children across all age comparisons, with children at age 3 scoring significantly lower than those at age 5 (β = −12.02, SE = 2.06, p < 0.001), and those at age 4 also showing lower scores (β = −4.25, SE = 1.97, p < 0.001). These results indicate substantial improvement in initial sound awareness with increasing age, particularly between ages 3 and 4. Preschool children with CIs also showed significant developmental gains over time. Compared with age 5, children at age 3 had significantly lower scores (β = −7.73, SE = 1.03, p < 0.001), as did those at age 4 (β = −2.62, SE = 0.98, p < 0.001). However, unlike NH children, whose gains were most pronounced between ages 3 and 4, the improvement in children with CIs was more gradual and consistent across the entire age range from 3 to 5.

For medial sound awareness, GLMM analyses revealed significant age-related differences among NH preschool children. Compared with the reference group (age 5), children at age 3 had significantly lower scores (β = −7.33, SE = 1.97, p < 0.001), while the difference between ages 4 and 5 was not statistically significant (β = −0.87, SE = 1.97, p = 0.660). These results suggest that the most notable developmental gains occurred between ages 3 and 4. A different pattern was observed in children with cochlear implants (CIs) in medial sound awareness over time. Preschool children with CIs showed significant improvements in medial sound awareness over ages. Children at age 3 performed significantly lower than those at age 5 (β = −13.77, SE = 1.09, p < 0.001), as did those at age 4 (β = −8.25, SE = 1.04, p < 0.001). In contrast to NH children, whose improvements were concentrated in early preschool years, those with CIs exhibited a more gradual and consistent pattern of development from ages 3 to 5. This suggests that phonological awareness in children with CIs may develop along a slower but steady trajectory.

For ending sound awareness, GLMM analyses revealed significant age-related differences among NH preschool children. Compared with the reference group (age 5), NH preschool children at age 3 scored significantly lower (β = −10.22, SE = 2.03, p < 0.001), while those at age 4 showed no significant difference (β = 0.96, SE = 1.94, p = 0.623). These results suggest that the most pronounced improvement occurred between ages 3 and 4. Similarly, preschool children with CIs also demonstrated significant developmental progress over time. Compared with age 5, children at age 3 had significantly lower scores (β = −13.72, SE = 0.99, p < 0.001), and so did those at age 4 (β = −7.42, SE = 0.94, p < 0.001). In contrast to the NH group, where improvement was concentrated between early ages, the CI group exhibited a more gradual and continuous developmental trajectory across ages 3 to 5.

For four-tone discrimination, GLMM analyses revealed significant age-related differences among NH preschool children. Compared with the reference group (age 5), children at age 3 (β= −5.32, SE = 1.11, p < 0.001) and age 4 (β= −5.13, SE = 1.06, p < 0.001) scored significantly lower, indicating overall improvement with age. However, the developmental gains in NH children appeared to be primarily concentrated between ages 4 and 5, as the difference between ages 3 and 4 was not statistically significant. Preschool children with CIs also demonstrated significant improvement in tonal awareness over time. Compared with age 5, children at age 3 (β= −5.12, SE = 1.36, p < 0.001) and age 4 (β= −4.72, SE = 1.43, p < 0.001) had significantly lower scores. In contrast to the NH group’s more concentrated developmental gains, the CI group exhibited a more gradual and continuous improvement from ages 3 to 5, with steadily increasing scores across age groups. This suggests differing developmental trajectories in tonal awareness between the two groups.

Furthermore, to better understand the developmental sequence of phonological segmentation, comparisons were made between the scores of initial sound awareness, medial sound awareness, and ending sound awareness.

For NH preschool children, results from the GLMM indicated that age, task type, and their interaction significantly influenced children’s performance in syllable segmentation tasks. Compared with five-year-olds, three-year-olds scored significantly lower (β = −7.30, SE = 1.00, p < 0.001), as did four-year-olds (β = −2.40, SE = 1.00, p = 0.016), suggesting a developmental progression in phonological awareness with age. Regarding task type, five-year-old NH children showed no significant differences among initial, medial, and ending sound awareness, indicating that all three skills had reached a relatively high level. An interaction between age and task type revealed that three-year-old NH preschool children performed significantly better on tasks targeting initial sound awareness than on those targeting medial (β = −6.47, SE = 1.41, p < 0.001) and ending sounds awareness(β = −6.33, SE = 1.41, p < 0.001). A similar pattern was observed in four-year-old NH preschool children, who also showed significantly higher performance on initial sound tasks compared with medial (β = −5.50, SE = 1.41, p < 0.001) and ending sounds (β = −4.90, SE = 1.41, p < 0.001). These findings suggest that sensitivity to word-initial phonemes emerges earlier in development, possibly due to their greater salience in speech perception and production. In contrast, awareness of medial phonemes appears to become more differentiated and refined around age five, reflecting a gradual maturation of phonological processing abilities.

For preschool children with CIs, results from GLMM revealed significant main effects of age, task type, and their interaction on performance in syllable segmentation tasks. In terms of age, both three- and four-year-old children scored significantly lower than five-year-olds, indicating a developmental progression in phonological awareness. Specifically, three-year-olds performed significantly worse than five-year-olds (β = −13.765, SE = 1.04, p < 0.001), as did four-year-olds (β = −8.252, SE = 0.99, p < 0.001). With respect to task type and its interaction with age, three-year-old children with CIs showed significantly better performance on ending sound awareness tasks than on initial sound awareness (β = 6.516, SE = 1.47, p < 0.001), while no significant difference was observed between initial and medial sound awareness (β = 0.050, SE = 1.47, p = 0.973). Four-year-olds similarly exhibited significantly higher performance on ending sound awareness compared with initial sound awareness (β = 5.636, SE = 1.40, p < 0.001), whereas their performance on medial sound awareness did not differ significantly from initial sound awareness (β = 0.836, SE = 1.40, p = 0.550). Among five-year-old children with CIs, performance on ending sound awareness tasks was significantly better than on both medial (β = 2.364, SE = 0.96, p = 0.014) and initial sound awareness tasks (β = 3.030, SE = 0.96, p = 0.002). Taken together, these findings suggest that among children with CIs aged three to five, awareness of ending sounds tends to develop earlier and more robustly than awareness of initial or medial sounds. Notably, differentiation in initial sound awareness appears to become more pronounced around age five, indicating a developmental progression in phonological awareness.

4. Discussion

This study found that preschool children with cochlear implants (CIs) consistently exhibited lower phonological awareness scores than their normal-hearing (NH) peers across all age groups (3 to 5 years). Despite this overall gap, both groups showed parallel developmental trajectories, with scores gradually improving with age. These findings suggest that early auditory input through cochlear implantation, combined with structured language interventions such as auditory–verbal therapy, supports the progressive development of foundational phonological skills—even prior to formal literacy instruction [24,25]. Notably, children with CIs demonstrated particular difficulty with tone discrimination, which was the lowest-scoring subtest among the four phonological awareness components. This may be attributed to their limited access to prosodic cues during the early stages of auditory development. This pattern echoes the findings of Zhao, Sun, Xie, Chen, Feng, and Wu (2019) [20], who reported that the predictive strength of phonological components—such as tone awareness—varies depending on children’s hearing status and developmental stage. Taken together, these results underscore the need for phonological training programs to be carefully tailored to the unique auditory profiles and processing challenges of children with CIs.

To better understand the nature of this delay, it is essential to consider the typical developmental trajectory of phonological awareness. NH children generally develop syllabic awareness by age 3 and begin to segment phonemes around age 4.5 [28]. They typically become proficient in identifying and extracting initial phonemes before entering formal reading instruction, a skill closely tied to early literacy success [3]. Our findings indicate that preschool children with CIs follow a similar developmental sequence—progressing from syllabic to phonemic awareness—but at a slower pace, likely due to reduced auditory input in early life [29]. These results also align with those of Pan et al. (2016) [14], who conducted an eight-year longitudinal study tracking phonological awareness development in typically developing children. Their study demonstrated that phonological awareness grows steadily over time and plays a critical role in early literacy acquisition. Although Pan et al.’s work focused on children with typical hearing, the similar developmental pattern observed in children with CIs suggests a universal trajectory of phonological development that may be delayed but not fundamentally altered by hearing loss. Taken together, these findings suggest that while children with CIs experience early phonological delays, their developmental pathways resemble those of NH children in structure, though at a slower pace. This indicates that phonological awareness is a foundational cognitive skill robust across populations but significantly influenced by hearing status and auditory access. In contrast, morphological awareness, which is more language-specific and semantically driven, may serve as an alternative or compensatory route for literacy acquisition in children with CIs, especially as they approach school age.

Notably, this study revealed differences in the sequencing of subskill development between the two groups. In NH preschool children aged 3 and 4, initial sound awareness scores were higher than those for ending sounds, with this gap diminishing by age 5. This is consistent with prior research suggesting that initial sound awareness emerges earlier in typical development. However, preschool children with CIs exhibited the opposite pattern: ending sound awareness emerged earlier and remained stronger than initial sound awareness. This finding contrasts with studies in alphabetic languages such as English and Japanese, where initial sound awareness typically emerges earlier than final sound awareness [30,31] but aligns with phonological saliency theory in Chinese, which posits that the rime or final part of the syllable is more acoustically salient than the initial consonant [23].

These findings also underscore the foundational role of speech perception in phonological awareness development. The difficulties observed in tone and initial consonant awareness among CI users likely stem from limited access to fine-grained auditory cues—particularly pitch and high-frequency information—during critical early language learning periods. This perceptual limitation can impair the formation of robust phonological representations, as suggested by previous studies [17,18]. In this study, the atypical sequencing of subskill development in the CI group—such as the earlier emergence of ending sound awareness—may reflect the relative perceptual salience of rime over onset in the auditory input accessible to CI users. Thus, while the CI group shows a parallel but slower phonological trajectory, their development is shaped not only by delayed exposure but also by the qualitative limitations in their perceptual input. These results reinforce the close link between speech perception and phonological processing, and they highlight the need for early intervention strategies that jointly target both auditory discrimination and higher-order phonological skills.

These differences may be attributed to the unique phonological structure of the Chinese language. Unlike alphabetic languages, where phoneme awareness is closely tied to grapheme–phoneme correspondence, Chinese is a syllable-based logographic language. The syllable is the most salient unit, and phonological awareness typically develops from syllables and tones to phonemic elements such as onsets and codas [32]. Tone and rime segments are acoustically more prominent and perceptually accessible, especially for children with degraded auditory signals such as those using cochlear implants. As Lu et al. (2022) [33] emphasized, tone perception plays a central role in early word learning and language development in Chinese-speaking children with hearing impairments. These findings are consistent with those of Shu et al. (2008) [13], who demonstrated the impact of Chinese syllabic structure—particularly tone and rime—on the development of phonological awareness. Their research indicated that in the Chinese context, rime and tone awareness are the earliest and most critical types of phonological awareness to emerge. The development of tone and rime awareness typically precedes that of phonemic awareness. Moreover, Pinyin instruction—which includes initials, finals, and tones—significantly promotes the development of smaller phonological units such as phonemes.

Our results suggest that although the phonological awareness skills of children with cochlear implants (CIs) improve significantly with age, the gap between children with CIs and normal-hearing (NH) children tends to widen over time. For instance, even by age 4, children with CIs still struggled to differentiate between onsets and rimes, likely due to limited exposure to speech models during the first two years of life—a critical period for auditory and phonological development [29]. This delay aligns with previous findings reporting persistent difficulties among hearing-impaired children in segmenting speech, articulating consonants, and mastering syllabic diversity. Thus, the slower yet parallel developmental trajectory observed in children with CIs likely reflects a combination of early auditory deprivation and language-specific phonological constraints. Taken together, these findings highlight the critical importance of early cochlear implantation and timely intervention for fostering phonological awareness development in this population. Indeed, with early implantation, congenitally deaf children are capable of developing age-appropriate phonological awareness skills and working memory capacity, which has significant practical implications for aural rehabilitation in this special pediatric group [17]. These findings highlight the need for early and linguistically appropriate phonological training programs that specifically target tone and rime awareness in Chinese-speaking children with CIs.

4.1. Research Limitations

Several limitations of this study should be acknowledged. First, this study utilized a cross-sectional design, which, while offering preliminary insights into phonological awareness development across different age groups, limits the ability to make definitive conclusions about individual longitudinal developmental trajectories. Although comparisons across age groups provide useful information, longitudinal data are necessary to track the phonological awareness development of the same cohort of CI children over time and to validate whether the observed developmental sequences and patterns reflect actual developmental progressions. Second, considerable individual variability within the CI group—such as differences in age at implantation, length of device use, auditory training experience, home language environment, and cognitive development—was not fully controlled and may have contributed to performance differences. These factors are known to significantly influence language and literacy outcomes in children with cochlear implants. Third, although the phonological awareness tasks were adapted for Chinese-speaking children, cross-linguistic comparisons should be interpreted with caution, as differences in language structure and task design can affect performance and limit generalizability across linguistic contexts.

4.2. Educational and Clinical Implications

The findings of this study highlight the critical need for early and sustained language interventions for Chinese-speaking children with cochlear implants. Given their delayed phonological development and atypical subskill profiles, intervention programs should be specifically tailored to their developmental and linguistic needs. Special emphasis should be placed on enhancing phonological awareness skills—particularly the perception and manipulation of initial consonants, which are acoustically less salient in Chinese and more difficult for CI children to perceive. Instructional strategies should prioritize activities involving syllables, tones, and rimes in the early stages and gradually introduce phonemic-level tasks as auditory discrimination improves. Furthermore, literacy instruction for CI children should be carefully scaffolded and extended over a longer duration to ensure a solid phonological foundation for later reading and writing success [17,25]. Clinicians and educators should also be aware of the wide range of individual differences among CI children and provide personalized support accordingly.

5. Conclusions

In conclusion, this study found that while Chinese-speaking preschool children with cochlear implants demonstrate lower phonological awareness than their normal-hearing peers, they follow a broadly similar developmental sequence, albeit at a slower rate. These differences appear to be shaped by both auditory input limitations and language-specific phonological features, such as the saliency of rimes and tones in Chinese. This study contributes to a deeper understanding of phonological development in children with hearing loss in tonal language environments and underscores the importance of linguistically appropriate and developmentally targeted early intervention strategies. Future longitudinal research is needed to verify the developmental trajectories suggested by this cross-sectional data and to further investigate the interactions between auditory experience, linguistic structure, and individual cognitive differences in shaping phonological awareness development.

Building on these findings, identifying the phonological awareness profiles of preschool children with cochlear implants is essential for improving the accuracy of language assessment and tailoring individualized intervention programs. Such profiling enables clinicians and educators to better address the specific developmental needs of this population, ultimately supporting more effective and equitable language and literacy development in preschool children with hearing loss.