Auditory Mismatch Response to Pitch and Duration Changes in Children with Developmental Language Disorder: A Longitudinal Approach

Auditory Mismatch Response to Pitch and Duration Changes in Children with Developmental Language Disorder: A Longitudinal Approach | Authorea try { document.documentElement.classList.add('js'); } catch (e) { } var _gaq = _gaq || []; _gaq.push(['_setAccount', 'G-8VDV14Y67G']); _gaq.push(['_trackPageview']); (function() { var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true; ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js'; var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s); })(); Skip to main content Preprints Collections Wiley Open Research IET Open Research Ecological Society of Japan All Collections About About Authorea FAQs Contact Us Quick Search anywhere Search for preprint articles, keywords, etc. Search Search ADVANCED SEARCH SCROLL This is a preprint and has not been peer reviewed. Data may be preliminary. 13 February 2026 V1 Latest version Share on Auditory Mismatch Response to Pitch and Duration Changes in Children with Developmental Language Disorder: A Longitudinal Approach Authors : Liis Themas 0000-0002-7615-0895 [email protected] , Gesa Schaadt , Pärtel Lippus , Marika Padrik , Liis Kask , Ulvi Vaher , Mairi Männamaa , and Kairi Kreegipuu Authors Info & Affiliations https://doi.org/10.22541/au.177100799.93641922/v1 213 views 96 downloads Contents Abstract Liis Themas, Näituse 2, 50409, Tartu, Estonia; [email protected] 1.5 Goals, research questions and hypotheses Supplementary Material Information & Authors Metrics & Citations View Options References Figures Tables Media Share Abstract Developmental language disorder is characterized by difficulties in auditory processing and prosody perception. Estonian word prosody features a unique three-way quantity distinction, characterized by changes in syllable duration and pitch contour, which differentiate word meaning and grammatical cases. We investigated auditory discrimination in 50 children (25 with developmental language disorder), aged 4-7 years, using the event-related potential Mismatch Response and behavioral tasks longitudinally at two assessment points. Here, we specifically focus on the results of the second assessment and compared these results with those reported previously (Themas et al., 2025a) In the duration change condition, we observed a shift from a positive at mean age of 5;8 years to a negative Mismatch Response at mean age of 6;8 years, which was less pronounced in the developmental language disorder group compared to their typically developing peers. Interestingly, we found a significant negative association between the change of the Mismatch Response amplitude and behavioral discrimination, suggesting a stronger change from positive to a more negative Mismatch Response to be associated with better behavioral discrimination performance for both groups. In the duration and pitch change condition, no Mismatch Response was observed in either group at the first or second assessment. This study sheds light on the maturation of auditory change detection processes and deepens our understanding of the interplay between neurophysiological and behavioral indicators in children with and without language impairments. Auditory Mismatch Response to Pitch and Duration Changes in Children with Developmental Language Disorder: A Longitudinal Approach Running title Auditory Perception in DLD Authors * these authors contributed equally *Liis Themas, Institute of Estonian and General Linguistics, University of Tartu; Institute of Psychology, University of Tartu, Tartu, Estonia; https://orcid.org/0000-0002-7615-0895 *Gesa Schaadt, Department of Education and Psychology, Freie Universität Berlin; Department of Neuropsychology, Max Planck Insitute of Human Cognitive and Brain Sciences Leipzig, Germany; https://orcid.org/0000-0002-3192-3698 Pärtel Lippus, Institute of Estonian and General Linguistics, University of Tartu, Tartu, Estonia; https://orcid.org/0000-0003-4407-811X Marika Padrik, Institute of Educational Science, University of Tartu, Tartu, Estonia ; https://orcid.org/0009-0003-5851-2084 Liis Kask, Institute of Psychology, University of Tartu, Tartu, Estonia; https://orcid.org/0000-0003-1852-0058 Ulvi Vaher, Department of Radiology, Institute of Clinical Medicine, University of Tartu, Tartu, Estonia; Children’s Clinic, Tartu University Hospital, Tartu, Estonia; https://orcid.org/0000-0002-2546-1079 Mairi Männamaa, Department of Radiology, Institute of Clinical Medicine, University of Tartu, Tartu, Estonia; Children’s Clinic, Tartu University Hospital, Tartu, Estonia; https://orcid.org/0000-0001-5234-6474 Kairi Kreegipuu, Institute of Psychology, University of Tartu, Tartu, Estonia; https://orcid.org/0000-0002-0953-7264 Corresponding author information Liis Themas, Näituse 2, 50409, Tartu, Estonia; [email protected] Word count 10086 (including abstract, keywords, and main body: Introduction, Methods, Results, Discussion, and Conclusion) Figure count 6 (main text) + 6 (appendices) Acknowledgements This work was supported by the Estonian Research Council, Tallinn, Estonia [grant number PRG1151] and Estonian Center of Excellence for Well-Being Sciences (EstWell), funded by grant TK218 from the Estonian Ministry of Education and Research. Abstract Developmental language disorder is characterized by difficulties in auditory processing and prosody perception. Estonian word prosody features a unique three-way quantity distinction, characterized by changes in syllable duration and pitch contour, which differentiate word meaning and grammatical cases. We investigated auditory discrimination in 50 children (25 with developmental language disorder), aged 4-7 years, using the event-related potential Mismatch Response and behavioral tasks longitudinally at two assessment points. Here, we specifically focus on the results of the second assessment and compared these results with those reported previously (Themas et al., 2025a) In the duration change condition, we observed a shift from a positive at mean age of 5;8 years to a negative Mismatch Response at mean age of 6;8 years, which was less pronounced in the developmental language disorder group compared to their typically developing peers. Interestingly, we found a significant negative association between the change of the Mismatch Response amplitude and behavioral discrimination, suggesting a stronger change from positive to a more negative Mismatch Response to be associated with better behavioral discrimination performance for both groups. In the duration and pitch change condition, no Mismatch Response was observed in either group at the first or second assessment. This study sheds light on the maturation of auditory change detection processes and deepens our understanding of the interplay between neurophysiological and behavioral indicators in children with and without language impairments. Keywords EEG, MMN, auditory perception, developmental language impairment, longitudinal study, child development 1 Introduction 1.1 Developmental language disorder Developmental language disorder (DLD) emerges during development and is characterized by persistent difficulties in acquiring and using language across different domains, such as phonology, vocabulary, and morphology (Bishop et al., 2017; Gillam et al., 2021), independent of intellectual disabilities, hearing impairments, or primary physical or neurological disorders (Bishop et al., 2017). One of these children’s core deficits is an atypical auditory processing (Kujala & Leminen, 2017; Näätänen et al., 2014), including prosody perception (Calet et al., 2021; Cumming et al., 2015; Leonard & Kueser, 2019; Richards & Goswami, 2019; Sundström et al., 2019). 1.2 The Estonian three-way quantity distinction Estonian is one of several quantity languages where phoneme length, as a prosodic feature, plays a crucial role in differentiating between word meanings and grammatical forms. Both vowels and consonants in the stressed syllable exhibit a three-way length distinction: they can be short (quantity 1 – Q1; e.g., [sɑtɑ], [kɑlɑ]), long (quantity 2 – Q2; e.g [sɑːtɑ], [kɑllɑ]) or overlong (quantity 3 – Q3; e.g., [sɑːːtɑ], [kɑlːlɑ]; Lippus et al., 2013). Quantity distinction is marked primarily by the duration of the stressed initial syllable, which is longer in the case of higher degrees of quantity, while the following unstressed syllable is compensatorily shortened. In addition, the quantity is marked with the fundamental frequency contour: in Q1 and Q2, the pitch fall is aligned with the end of the stressed syllable, whereas in Q3, the fall takes place in the first half of the stressed syllable (Lippus et al., 2013). To date, only one study (Themas et al., 2025a) has been conducted on auditory quantity perception with three-way quantity distinction in Estonian children. 1.3 Auditory discrimination and its maturation 1.3.1 Measuring the ability of auditory discrimination Auditory perception across maturation can be objectively and non-invasively investigated with electroencephalography (EEG) using event-related potentials (ERPs). An ERP reflecting the brain’s ability to discriminate different characteristics of speech, such as different quantity degrees, is the Mismatch Negativity (MMN), a well-studied fronto-central negativity typically elicited 100-200 ms after the onset of a deviating stimulus within a stream of standard stimuli (Näätänen et al., 1978). The MMN is typically tested with an oddball paradigm (Squires et al., 1975) but can also be elicited with the multi-feature paradigm (Optimum-1; Näätänen et al., 2004), in which more than one deviant stimulus can be tested, such as different quantity degrees in the Estonian language. In studies investigating children, the MMN is often referred to as the Mismatch Response (MMR) because it can be of positive or negative polarity and its latency is delayed compared to the adult’s MMR latency (Campos et al., 2023; Cheng et al., 2015; Leppänen et al., 2004; Maurer et al., 2003a; J. L. Mueller et al., 2012). The amplitude of the MMR is also influenced by the nature of the eliciting stimuli, particularly by whether they are linguistic or nonlinguistic. Evidence for this distinction comes from several studies in children, which show enhanced MMNs for linguistic compared to nonlinguistic stimuli (Kuuluvainen et al., 2016), as well as domain-specific maturational trajectories towards adult-like negative MMNs (Chen et al., 2022; Paquette et al., 2013). 1.3.2 Auditory discrimination of duration and pitch The perception of different quantity degrees in the Estonian language requires the auditory discrimination of duration and pitch cues that can be measured by using the MMR. Both abilities emerge early in life and undergo maturational changes through childhood, as reflected by the MMR becoming more negative and exhibiting shorter latency with increasing age (Chobert et al., 2014; Friederici et al., 2002; Friedrich et al., 2004; He et al., 2007; Linnavalli et al., 2018; Polver et al., 2023; Shafer et al., 2000; Virtala et al., 2022; Werwach et al., 2022). Conclusions can also be drawn from studies on Mandarin lexical tone perception using linguistic stimuli, also requiring the perception of pitch contour information (e.g. Wu et al., 2025). Perception of lexical tone differences emerges in early infancy and continues to mature through childhood and adolescence (Y. Chen et al., 2016; A. Chen et al., 2022; Cheng et al., 2013; Cheng & Lee, 2018; Cheng et al., 2021; Lee et al., 2012; Liu et al., 2014; Wu et al., 2025). Unfortunately, this research topic remains understudied in the context of Estonian. To date, only two adult studies have investigated the discrimination of falling versus level pitch contours in meaningful words, which is relevant for distinguishing between Q2 and Q3. The results were contradictory - Kask et al. (2021) did not observe a significant mismatch negativity (MMN), whereas Lyu et al. (2024) did. However, in the context of Estonian quantity degrees, in addition to differentiating pitch contour the listener must also rely on the duration cue to distinguish between Q2 and Q3, such that the difference in pitch contour gradually appears starting from the beginning of the word, followed by the durational contrast. Adult studies have shown that in some cases the presence of multiple differences between the standard and deviant stimuli modulates the MMN. Deviants differing by two features from the standard elicit a stronger MMN compared to deviants differing by one feature (Paavilainen et al., 2001; Ylinen et al., 2005). Nevertheless, in our previous study (Themas et al., 2025a) we found no significant MMR in children discriminating combined pitch contour and durational differences (Q2 vs Q3) in meaningful words. In contrast, Lyu et al. (2024) reported significant MMNs in Estonian adults during discrimination of combined pitch contour and duration changes in meaningful Q2 and Q3 words, but no enhanced MMN due to multiple differences between the standard and the deviant. 1.3.3 Auditory discrimination of meaningful words As prosodic features like duration and pitch contour are not perceived in isolation in naturalistic speech, but rather contribute to lexical distinctions in Estonian, the interplay between the discrimination of prosodic features and word meaning needs to be considered. Interestingly, the MMN has also been shown to reflect lexical processing in adults (Aleksandrov et al., 2017; Hasting et al., 2008; Pettigrew et al., 2004; Pulvermüller et al., 2001; Zora et al., 2015; Zora et al., 2016; Zora et al., 2023), such that prosodic features activate memory-based predictions about word meaning in the brain as evidenced by enhanced MMN amplitudes and shorter latencies compared to pseudowords with similar prosody (Pulvermüller et al., 2001; Zora et al., 2015; Zora et al., 2016; Zora et al., 2023). There is also evidence of higher sensitivity to prosodic changes in meaningful words as the brain leverages prior to lexical knowledge alongside the incoming acoustic information (Zora et al., 2015). One might also assume that in words with a high frequency in a language, the processing of prosodic features is more facilitated. This assumption is consistent with the view that differences between high-frequency and low-frequency words resemble those observed between high-frequency words and pseudowords and words with higher frequency are processed more faster and more automatically (Alexandrov et al., 2011; Aleksandrov et al., 2017). In studies involving children, evidence is scarce. Strotseva-Feinschmidt et al. (2015) found a more mature MMR in response to high frequency function words compared to low frequency function words in 3-year-olds. Campos et al. (2023) found more robust MMNs in adults when differentiating meaningful words compared to nonwords but did not observe this effect in children. Further, they found positive MMRs in children in the nonword condition, but negative MMRs in the meaningful word condition, suggesting a more mature processing of meaningful words compared to nonwords. However, they did not find any processing differences between TD children and children with DLD (Campos et al., 2023). 1.3.4 Auditory discrimination abilities in children with DLD In contrast, other studies have found that children with DLD have brain-level differences in auditory processing compared to TD children. Much research has been conducted on studying perception of pitch changes in non-linguistic tone stimuli, where children with DLD showed attenuated MMN/MMR amplitudes compared to TD children (for review see Bishop, 2007 and Kujala & Leminen, 2017). Group differences between children with DLD and TD concerning their MMRs to lexical tones were also found, such that children with DLD show less prominent or less mature MMRs (Chen et al., 2016; Cheng et al., 2021). Few studies have researched duration perception in DLD, and the one investigating participants with comparable age to the current study found no group differences in MMNs in response to duration changes of non-linguistic tone stimuli (Uwer et al., 2002). Research on children with developmental dyslexia (DD) using linguistic stimuli can provide additional insights into expected group differences. DLD and DD seem to stem from similar language processing difficulties (Leonard, 2014) and studies on duration change detection in school-aged children (older than the current sample) with DD showed reduced negative MMR amplitudes compared to typically developing (TD) children (Männel et al., 2017; Schaadt & Männel, 2019). There is also limited research investigating the maturation of brain-level differences in auditory perception, as reflected in the MMR, either longitudinally or by age group comparisons in the DLD population (Bishop et al., 2010; Y. Chen et al., 2016; Choudhury & Benasich, 2011). To better understand DLD and its manifestations, it is, however, crucial not only to measure auditory processing at a single time point, but to also observe its maturation, as DLD itself evolves over time. Longitudinal studies offer valuable insights into several key aspects of DLD. The MMR measured at infancy and toddlerhood has been shown to be predictive of future language abilities (Cantiani et al., 2016; Y. Chen et al., 2016; Choudhury & Benasich, 2011; Friedrich et al., 2009; Guttorm et al., 2005; Schaadt et al., 2015; Schaadt et al., 2023; Weber et al., 2005) and therefore could also be suitable for predicting future outcomes of DLD at later ages. 1.4 Relationship between neurophysiological and behavioral findings of auditory discrimination Currently, however, there is an ongoing discussion regarding the use of the MMR as a diagnostic tool and as a measure for evaluating treatment efficacy. For the MMR to be regarded as suitable for diagnostic purposes as well as for evaluating treatment efficacy, evidence should prove the MMR to be correlated with behavioral discrimination abilities. However, findings on the relationship between auditory discrimination denoted by ERPs and behavioral auditory discrimination are contradictory so far. Some studies have found significant associations between some neurophysiological and behavioral measures of auditory discrimination (McArthur & Bishop, 2005; Nan et al., 2018; Pihko et al., 2006), while others have not (Bishop et al., 2010; Halliday et al., 2014; McArthur & Bishop, 2005; Nan et al., 2018; Pihko et al., 2006; Uwer et al., 2002). Different task designs of behavioral auditory discrimination tasks and type of contrast tested in the MMR paradigm might explain these contradictory findings. Pakarinen et al. (2007) reported MMN amplitudes and latencies in response to different contrasts (frequency, duration, intensity, and location) to differently predict behavioral discrimination. For example, MMNs elicited by duration and intensity deviants differed from each other, but the behavioral discrimination rate did not. Further, neurophysiological and behavioral discrimination impose different cognitive demands on participants. While the MMR reflects the pre-attentive ability to discriminate auditory and/or phonological features, behavioral auditory discrimination requires the participant to understand the task, sustain attention throughout, consciously decide whether the stimuli differed, and respond appropriately. Evidently, clarification is needed, as in clinical settings, understanding this link helps to verify if behavioral auditory discrimination tasks objectively measure these abilities. Finally, examining ERPs alongside behavioral tasks helps to validate each measure and would ensure that they accurately assess auditory discrimination. 1.5 Goals, research questions and hypotheses As discrimination of quantities is crucial for understanding the Estonian language, we aimed at investigating the maturation of duration and pitch contour perception, the two main cues for distinguishing between quantity degrees, in Estonian children with DLD compared to their TD peers across one year. Further, we were interested in whether behavioral discrimination abilities are reflected in the MMR’s amplitude in both groups. To address the aims of our study, we investigated 5;6 to 7;5 year old children and compared these results with those reported previously (Themas et al., 2025a), based on the same sample of children tested one year earlier. We studied participants’ auditory processing both behaviorally as well as neurophysiologically, by measuring the MMR in the ERPs. We used a multi-feature paradigm to elicit MMRs, in which the standard stimulus was a meaningful word in Q2, the first deviant was the same word as the standard but in Q1 (syllable duration ratio change) and the second deviant was the same word as the standard but in Q3 (both, syllable duration ratio and pitch contour change). The few studies investigating duration change detection in linguistic stimuli at preschool age show that the ability to detect duration changes is still developing, as evidenced by neurophysiological measures (Linnavalli et al., 2018; Themas et al., 2025a). This is why we expected the ability to detect changes in syllable duration to mature over the course of one year in both groups as indicated by a more negative MMR at T2 compared to T1 (Themas et al., 2025a). As there is limited and contradictory evidence on the detection of the combined changes in duration and pitch contour (Kask et al., 2021; Lyu et al., 2024; Themas et al., 2025a; Wu et al., 2025), we did not formulate specific hypothesis concerning the maturation of the MMR in response to duration and pitch contour changes in linguistic stimuli over the course of one year. The one study investigating duration discrimination differences in participants with DLD and TD children with comparable age as in the present study, used tone stimuli and reported no group differences in MMR amplitudes (Uwer et al., 2002). As several studies have shown differences in the MMRs to non-linguistic and linguistic stimuli (Chen et al., 2022; Kuuluvainen et al., 2016; Paquette et al., 2015), comparability to the present study is not warranted. Literature on DD provides additional insights on the expected group differences, showing reduced negative MMR amplitudes compared to TD children (Männel et al., 2017; Schaadt & Männel, 2019). This is why we expected the group differences in the MMR amplitude to be significant and that children with DLD would show a less pronounced maturation of the MMR over the course of one year, reflected in a less negative MMR at the second measuring time point. For clinical settings, understanding the relationship between behavioral and neurophysiological measures of auditory discrimination is essential for determining whether behavioral auditory discrimination tasks objectively assess these abilities. Moreover, examining ERPs alongside behavioral measures can help to validate each approach and ensure that both accurately capture auditory discrimination. To our knowledge, longitudinal investigations of this relationship are rare (e.g., Nan et al., 2018) and cross-sectional studies provide mixed evidence on the relationship between behavioral and neurophysiological measures of auditory discrimination (Bishop et al., 2010; Halliday et al., 2014; McArthur & Bishop, 2005; Pihko et al., 2006; Uwer et al., 2002). Therefore, without formulating any specific hypothesis, we investigated the association between behavioral and neurophysiological auditory discrimination exploratorily, anticipating small effect sizes given that we are working with noisy child data, potentially lacking reliability at the individual level. 2 Materials and Method The current research was preregistered on the Open Science Framework, including the study description, goals, hypotheses, and a condensed version of the methods section at https://doi.org/10.17605/OSF.IO/PCW98 referenced as Themas et al. (2025b)11. All procedures performed in this study involving human participants were conducted in accordance with the ethical standards of the institutional and/or national research committee and with the 2013 revision of the World Medical Association Declaration of Helsinki, as published on the website of the Journal of the American Medical Association. The study was approved by the Ethics Committee of Human Research of the University of Tartu (approval number: 383/M-12). Written informed consent was obtained from the parents or legal guardians of all participating children prior to their inclusion in the study referenced as Themas et al. (2025b)22Compared with the preregistration, the goals and hypotheses presented in the paper have been refined to provide a clearer focus in line with the reviewers’ recommendations. However, the fundamental aim of the study—and, critically, the materials and methods—remain unchanged. All procedures performed in this study involving human participants were conducted in accordance with the ethical standards of the institutional and/or national research committee and with the 2013 revision of the World Medical Association Declaration of Helsinki, as published on the website of the Journal of the American Medical Association. The study was approved by the Ethics Committee of Human Research of the University of Tartu (approval number: 383/M-12). Written informed consent was obtained from the parents or legal guardians of all participating children prior to their inclusion in the study. This longitudinal project included a sampling phase, a first assessment (T1), and a second assessment (T2; see Figure 1). The sampling procedure and assessment methodology have been described in Themas et al. (2025a), where the results of T1 are reported. As identical assessment procedures were also used for our previous work, published in Themas et al. (2025a), we will give a concise overview of the materials, method, and procedure in the following. During the sampling phase, psychometric assessments of language and intellectual abilities were conducted. Children with language impairments meeting the diagnostic criteria were officially diagnosed with DLD and assigned to the clinical group. At the two assessments (at T1 and T2), neurophysiological and behavioral measures of auditory discrimination were administered. 2.1 Participants The initial cohort recruited for our previous study (Themas et al., 2025a) consisted of 50 monolingual Estonian-speaking children (Males = 27 and Females = 23; 25 children with DLD and 25 age and sex matched TD peers) aged 4;5-6;5 years. Exclusion criteria were intellectual disability, sensory impairment, and other neurological and psychiatric diagnoses (except for DLD). The DLD participants were recruited either through the Children’s Clinic of Tartu University Hospital or via kindergarten speech therapists, who forwarded study invitations to parents of children suspected of having DLD. TD children were recruited by contacting kindergarten administrators, who then forwarded the study invitation to parents. The participants were enrolled between 2021 autumn and 2023 spring from the southern counties of Estonia. After the sampling phase, the experimental testing was conducted approximately two weeks later. For the current study, all participants were re-invited one year after the initial experimental testing. Three participants did not follow the invitation (including one that already had been excluded from T1 due to noisy EEG data). All excluded participants belonged to the DLD group (see Figure 1). Therefore, for the present study, we included 47 participants (Males = 25 and Females = 22; DLD group = 22 and TD group = 25) aged 5;5-7;5 years. T2 testing took place between autumn 2022 until spring 2024. At T1, most of the recruited children attended regular kindergarten groups, except for four in the DLD group who attended special groups for speech- and language-impaired children, and one who did not attend kindergarten. By T2, a total of eight children (4 in the DLD group) had started school, and of these, two attended a special school for speech- and language-impaired children. All the children in the DLD group received speech and language therapy, although the amount of therapy varied in terms of monthly frequency and session length. Neither the parents nor the specialist who tested the children reported any hearing difficulties. 2.2 Psychometrical testing and background information To assess all children’s language abilities as well as to assess DLD, tasks of the standardized language battery for 5-6-year-old Estonian speaking children (Padrik et al., 2013) testing phonology, vocabulary, grammar usage, pseudoword repetition, sentence repetition, and sentence comprehension, were conducted. The above-mentioned tasks were selected based on their sensitivity in differentiating atypical from age-appropriate language development, while ensuring coverage of all major language domains. To control for children’s intellectual abilities, the Wechsler Preschool and Primary Scale of Intelligence fourth edition of the United Kingdom version (WPPSI-IV UK; Wechsler, 2013) was used for DLD and TD children at T1. As children with DLD are disadvantaged by tasks requiring language abilities, we assessed only the non-verbal intelligence and used those subtests that cover domains of visuospatial abilities, working memory, processing speed and fluid reasoning (Wechsler, 2013). A parental questionnaire was developed for this study to collect information on family background, home language environment, caregiver emotional state, attitudes toward language impairment, and attendance in speech and language therapy. The questionnaire was organized into five sections: (1) family background, (2) home language use, (3) caregiver well-being, (4) attitudes toward language impairment, and (5) therapy attendance. The last two sections were included only for parents of children in the DLD group. Items were a mix of Likert-scale ratings, multiple-choice questions, and short open-ended questions. The questionnaire was piloted with a small group of parents to ensure clarity. It was completed in an online format by the main caregiver at both T1 and T2. 2.3 Neurophysiological experiment 2.3.1 Paradigm and stimuli The method and stimuli used in the current study were identical to those in our previous work (Themas et al., 2025a). Below, we provide a brief description of the paradigm and stimuli for clarity. To test for discrimination abilities of quantity degrees, a passive multi-feature paradigm (Näätänen et al., 2004; Pakarinen et at., 2007; Pakarinen et al., 2009; see Figure 2A) was applied. The detailed description of the stimuli can be found in Table 1 and Figure 2B. The standard was an Estonian word [sɑːtɑ] ( ‘send!, imp. sg2’ ) which is pronounced in the second quantity degree (Q2) . For the first deviant (henceforth durQ1) we changed the quantity degree of the standard to be in Q1 [sɑtɑ] (syllable duration ratio change), which also resulted in the change in meaning ( ‘hundred, nom. sg’ ). For the second deviant (henceforth pitchdurQ3) we changed the quantity degree of the standard to be in Q3 [sɑːːtɑ] (both syllable duration ratio and pitch contour change), which also resulted in the change in meaning ( ‘to get, inf.’ ). The stimuli were naturally spoken by a female native Estonian speaker (the first author of the paper) in a soundproof room. Target words were segmented from carrier phrases using PRAAT (version 6.1; Boersma & Weenink, 2021). These words were embedded in the middle of the phrase (e.g., ütle … kõvasti ‘say … loudly’; ‘say a hundred loudly’) to maintain natural intonation patterns. In the context of a larger project, participants were presented with three different stimulus blocks (total duration 35 min). Here, we report the results from one of these blocks (duration 11 min), in which a total of 1,100 stimuli with alternating inter-stimulus interval (400, 425, or 450 ms; Figure 2A) were delivered through headphones using MATLAB (The MathWorks, 2015; version 8.5.0). We used the multi-feature paradigm with 60% of standards and 40% of deviants in which the stimuli were presented such that the number of standards between the deviants was either one or three (see Figure 2A) to investigate habituation differences between children with DLD and their TD peers (results presented elsewhere). The multi-feature paradigm has been successfully used in several studies with children of different ages and with different clinical populations (Kuuluvainen et al., 2016; Linnavalli et al., 2018; Niemitalo-Haapola et al., 2015; Partanen et al., 2013a,b; Schaadt et al., 2023; Torppa et al., 2022; Virtala et al., 2022), showing similar MMN/MMR amplitudes compared to a typical oddball paradigm (Lovio et al., 2009; Pakarinen et al., 2009; Partanen et al., 2013a). Importantly, we used less deviants (see also Fisher et al., 2011; Thiede et al., 2019; Virtala et al., 2022) compared to the typically used 5 deviants to reduce exhaustion of the auditory system, which might be particularly the case for clinical populations (Kujala et al., 2006; Partanen et al., 2013a). 2.3.2 EEG recording The EEG was recorded from 32 active electrodes (ActiveTwo system, BioSemi B.V., Amsterdam, Netherlands) placed at standard positions according to the 10-20 system, with the signal referenced to the CMS electrode and common mode rejection handled by the DRL electrode during data acquisition. Electrooculograms were recorded from two electrodes at the outer canthi of both eyes, one electrode on the orbital ridge under the right eye near the lower eye socket and one electrode located near the supraorbital ridge above the right eye, slightly lateral to the midline of the forehead. The EEG data were digitized at 512 Hz frequency with a bandpass filter of 0.16–100 Hz. 2.4 Behavioral tasks For the current study, we used the same method and stimuli as described in Themas et al. (2025a) to retest behavioral discrimination after one year of maturation. For clarity, a brief description is provided below. 2.4.1 Quantity discrimination task To behaviorally assess the children’s quantity discrimination abilities, the roving standard paradigm was used (Fong et al., 2020). In this paradigm, a stimulus was repeated 3-5 times to establish it as the perceptual baseline. Then a new stimulus with a different quantity degree was introduced and similarly repeated until it became a new baseline. The task was to actively listen and press a button as soon as a new stimulus was heard in the sequence of stimuli. Additional details regarding the task stimuli and procedure are available in Appendix A. 2.4.2 Lexical decision task Half of the stimuli in the lexical decision task were meaningful words considered to be familiar to 4-year-olds. Another half of the stimuli were pseudowords with no meaning, created by changing the quantity degree of the meaningful word. To enhance accessibility for participants with DLD, each auditory stimulus was paired with an image of an object. For example, participants saw a picture of a rabbit while hearing the meaningful word jänes ‘rabbit’, but they also saw the same picture when hearing a pseudoword jäänes derived from the meaningful word by lengthening the first vowel . The task was to press a button when the stimulus was a meaningful word and thus matched the image. Additional details regarding the task stimuli and procedure are available in Appendix B. 2.6 Procedure The procedure of the sampling phase and the first assessment (T1) is published in Themas et al. (2025a). The procedure of the second assessment (T2) was identical to the first assessment, published in Themas et al. (2025a). Thus, we provide a brief description of the process in the following (Figure 1). At the sampling phase, psychometric testing was carried out at the Children’s Clinic of Tartu University Hospital for the DLD group by an in-house speech and language therapist and clinical psychologist. A comprehensive neurological and physical examination was performed. The final diagnosis was established by an interdisciplinary clinical team consisting of a child neurologist, clinical psychologist, and speech and language therapist using the International Classification of Diseases (version 10; World Health Organization, 1992) criteria. The psychometric testing of the TD group was carried out in the experimental psychology laboratory, by the first author of the current paper and a psychologist with at least a master’s level of education. Testing was conducted over two sessions, each lasting slightly more than one hour, for a total duration of approximately 2.5 hours. After the sampling phase, the experimental testing (T1) was conducted approximately two weeks later. The participants were called back for the second assessment of neurophysiological and behavioral discrimination one year after T1 (DLD group after M = 12.3 months, SD = 0.5; TD group after M = 12.2 months, SD = 0.7). The reassessment session lasted approximately 1h and 40 min. The procedure was introduced to the children and their parents again. During the neurophysiological session, participants were seated comfortably in front of a monitor. They watched a silent cartoon of their choice during the EEG cap fitting and experimental tasks. The stimuli were presented using a comfortable volume level between 55-65 dB with JBL Tune290 earphones (Los Angeles, California, USA). After the neurophysiological experiment, the cap was removed, and the participants proceeded to the behavioral tasks. They were seated in front of a computer next to the experimenter, who helped the children through the teaching phases of the behavioral tasks and remained with the child during the testing phases. The stimuli were presented using a comfortable volume level between 55-65 dB with JBL Tune290 earphones. Throughout the whole procedure, a parent was always present in the laboratory. 2.7 Data analysis The raw EEG data that support the findings of this study are openly available in the University of Tartu Library DataDOI repository at https://doi.org/10.23673/re-525, referenced as Themas (2025). The code and data for modeling are available in the Zenodo repository at https://doi.org/10.5281/zenodo.15799644, referenced as Themas and Nickel (2025). 2.7.1. Descriptive statistics of the sample The Wilcoxon rank-sum test was used to compare the two participant groups with respect to age, standardized language test scores, and WPPSI‑IV UK non-verbal IQ scores. For variables like sex, handedness and the level of maternal education (basic education - completed 9 years of school; secondary education - completed 12 years of school; vocational education - providing specialized skills and knowledge for a particular trade or occupation; higher education - at least a bachelor’s degree) we used the Chi-squared test. 2.7.2 EEG data processing and analysis The Maryland analysis of developmental EEG (MADE; Debnath et al., 2020) was used to preprocess EEG data in MATLAB (The MathWorks, 2023; version 9.14) using EEGLAB (version 2023.1; Delorme & Makeig, 2004). The data was recorded with a sampling rate of 512 Hz and down sampled to 500 Hz following high (0.3 Hz, stopband of 0.1 Hz) and low (50 Hz, transition band of 10 Hz) pass filtering. Global artifact rejection was performed using the FASTER extension in EEGLAB (Nolan et al., 2010). Any participant whose dataset contained more than 10% of bad channels was removed from the study. Muscle and eye-movement artifacts were rejected using independent component analysis (ICA) and the components which corresponded to artifacts were selected using the adjusted-ADJUST algorithm (Leach et al., 2020). The data was divided into epochs spanning -100–698 ms and, which were then baseline corrected using the 100 ms pre-stimulus interval. Residual artifacts were identified by scanning electrodes for epochs in which voltage values surpassed ±125 mV. The missing channels were interpolated. Lastly, the data was re-referenced using an average reference. For a more detailed description, please see Appendix C. No participant had to be excluded due to severe artifacts. Consistent with T1 (Themas et al., 2025a), no significant group differences were found in the number of usable epochs (see Table 2) at the second assessment. Grand averages of the ERPs to the standard and deviant stimuli as well as the difference waveform were plotted using EEGLAB (version 2023.1; Delorme & Makeig, 2004; see Figure 3 and 4). To identify time windows and electrodes where ERPs to the standard and deviant stimuli significantly differed, that is the MMR, cluster-based permutation tests (MATLAB, Fieldtrip toolbox; Oostenveld et al., 2011) were calculated for each time point between stimulus onset and epoch end (698 ms). Only the largest positive and negative clusters representing the deviant-standard difference were considered, as this approach controls for the family-wise error rate and minimizes the likelihood of false discoveries. For a more detailed description of the process along with the original topographical plots of standard–deviant clusters, see Appendix D. The method and the code are based on Meyer et al. (2021). 2.7.3 Behavioral data analysis For each behavioral task, we calculated the sensitivity index ( d′ ) for each participant. This index reflects performance accuracy, as shown in the equation below. Z (Hit Rate) represents how many standard deviations the observed hit rate was from the mean of the standard normal distribution and the Z (False Alarm Rate) indicates how many standard deviations the observed false alarm rate was from the mean of the standard normal distribution. d’ = Z(Hit Rate) - Z(False Alarm Rate) To assess whether there were significant differences between the measurements at T1 and T2, we conducted Wilcoxon signed-rank tests for paired samples. Group differences between children with and without DLD regarding d’ scores were calculated using Wilcoxon rank-sum test. Lastly, we wanted to describe how participants differentiated between different quantity contrasts and compared the following pairs: first quantity degree (Q1) versus second quantity degree (Q2; syllable duration difference), first quantity degree (Q1) versus third quantity degree (Q3; a large syllable duration difference and a pitch contour difference), and second quantity degree (Q2) versus third quantity degree (Q3; a small syllable duration difference and a pitch contour difference). In the quantity discrimination task, an error occurred when the participant failed to press the button in response to a change in quantity. In the lexical decision task, an error was defined as a button press where the participant incorrectly identified a word as correct, despite the word being pronounced in the wrong quantity degree. The errors in discriminating between specific quantity contrasts were counted, summed across participants, and group differences were calculated with the Wilcoxon rank-sum test. 2.7.4 Hypotheses testing and exploratory analysis We used linear mixed models (LMM) to analyze longitudinal changes in MMR amplitude, group differences of maturational change, and the relationship between behavioral and neurophysiological measures. The statistical modeling strategy followed Murphy et al. (2022) and was implemented in R (version 4.3.2; R Core Team) using the lme4 package (version 1.1-35.3; Bates et al., 2015) and lmerTest (version 3.1-3; Kuznetsova et al., 2017), with visualizations created using ggplot2 (version 3.4.4; Wickham, 2016). In all of the LMMs the MMR difference wave amplitudes (deviant – standard) served as a dependent variable. The time windows showing significant deviant-standard differences, as determined by cluster-based permutation tests, were divided into two segments: from the start to the midpoint, and from the midpoint to the end. Consequently, two mean amplitude values of the MMR were extracted from both T1 and T2 data, resulting in a total of four mean MMR values for each participant per condition. A random intercept was included to account for the correlation between repeated measures, and a random slope for time was added to capture anticipated variations in slopes over time among participants. To analyze changes in MMR amplitude over the course of one year across all children, as well as group differences in these changes, we fitted a separate LMM for each deviant condition (durQ1 and pitchdurQ3). The predicting variables were measurement time (with levels: T1, T2) and group (levels: DLD and TD). Participant’s age at T1 was added as an additional predicting variable to account for general maturation and variable time window (with levels: TW1, TW2) was also added to account for the data structure in the dependent variable. Lastly, we tested whether the variable of the overall language test score improved the models. To examine the association between behavioral task results and MMR amplitude, we created four separate models in which the predictors were: the scores from the quantity discrimination task at T1 and T2, and the scores of the lexical decision task at T1 and T2. The interaction of time and time window was additionally added to the model to account for the data structure of the dependent variable. Finally, group status (levels: DLD and TD) was added as an additional predictor to inspect whether the relationship between behavioral and neurophysiological discrimination manifests differently in the two experimental groups. All six models (durQ1 condition; pitchdurQ3 condition; behavioral discrimination task -neurophysiological relationship T1 and T2; lexical decision task -neurophysiological relationship T1 and T2) were fitted with maximum likelihood estimation and degrees of freedom were approximated using the Satterthwaite method. Models were created incrementally by adding predictors one-by-one. The optimal model was chosen by comparing the model residual variances by using analyses of variances (ANOVA) and by evaluating the model fit with Akaike information criterion (AIC). Since we did not perform a priori power analysis for the study, we calculated effect sizes (standardized beta coefficients, β ) with confidence intervals (Wald confidence intervals derived from the refitted standardized model) and examined marginal and conditional pseudo R-squared ( R² ) values to improve the interpretability of the models. For model diagnostics, we checked for potential multicollinearity, non-linearity between dependent and independent variables, heteroscedasticity in the model’s residual variance and normal distribution of the model’s residuals, random effects and random intercepts (see Appendix E). To additionally explore the relationship between behavioral task results and MMR amplitude, we performed a correlation analysis using Kendall’s tau. We used the mean MMR amplitude values from T1 and T2 derived from the relevant time windows and electrodes as defined by the cluster-based permutation tests. We also used the d’ scores of the two behavioral tasks at T1 and T2. Correlation analyses were conducted between the T1 MMR amplitude and the T1 behavioral results, and the same procedure was applied to the T2 data. 3 Results Here we present the results of T2, along with the comparison of the results between T1 and T2. The original data of the sampling phase and the T1 results are published in Themas et al. (2025a). 3.1 Sample statistics The description of age, sex, handedness, maternal education level, language test scores and nonverbal IQ scores for the two groups are presented in Table 3. After excluding three participants from the original DLD sample (Themas et al., 2025a), children with DLD did not differ significantly from their TD peers in age, sex, or handedness. As in the T1 sample, they did differ in maternal education level, standardized language test scores, and WPPSI-IV UK nonverbal IQ scores (see Table 3). 3.2 Results of the EEG data analysis Grand averages of the ERPs to the standard and deviant stimuli (durQ1) as well as of the difference waveform (deviant durQ1 – standard) at T1 and T2 separately for children with and without DLD can be found in Figure 3 and Figure 4. As reported in Themas et al. (2025a) the Cluster-based permutation tests (Figure 5; original plots in Appendix D) done at T1 revealed a more positive amplitude in response to the deviant stimulus compared to the standard stimulus between 240 and 340 ms after stimulus onset at frontal, fronto-central and central electrodes (AF3, F3, FC5, F7, Fz, FC1, C3, FC6, F8). The mean MMR values (deviant–standard) and their standard deviations for this significant cluster are presented in Table 4. At T2 the Cluster-based permutation tests (Figure 5; original plots in Appendix D) revealed a more negative amplitude in response to the deviant stimulus compared to the standard stimulus between 390 and 490 ms after stimulus onset at frontal, fronto-central, central, centro-parietal and parietal electrodes (F3, F4, FC1, FC2, C4, Cz, CP2, Pz). The above-mentioned time-window and electrodes were further analyzed with LMMs. For further significant clusters, not related to the MMR, see Appendix D. The mean MMR values (deviant–standard) and their standard deviations for this significant cluster are presented in Table 4. Grand averages of the ERPs to the standard and deviant stimuli (pitchdurQ3) at T1 and T2 separately for children with and without DLD can be found in Figure 3. As reported in Themas et al. (2025a) the Cluster-based permutation tests (Figure 5; original figures in Appendix D) revealed no time windows or electrodes where the waveforms of the standard and deviant stimuli were significantly different at T1, which was also the case after one year in the current study. Therefore, we did not further analyze the pitchdurQ3 condition in the present study. 3.3. Results of the behavioral data analysis In both of the behavioral discrimination tasks, participants performed significantly better at T2 compared to T1, with higher d’ scores in both the quantity discrimination ( V = 163, p = 2.25 × 10⁻⁵) and lexical decision tasks ( V = 257, p = 0.0012; see Table 5), and shorter reaction times in both tasks (quantity discrimination: V = 673, p = 8.07 × 10⁻⁵; lexical decision: V = 801, p = 0.011). Group means differed significantly in d’ scores of the quantity discrimination task and lexical decision task, but no differences in means were found in RTs (see Table 5). We further examined the discrimination of different quantity degrees. Table 5 shows the sum of errors made while discriminating specific quantity contrasts and group differences in means. 3.4. Results of the hypotheses testing and exploratory analysis 3.4.1 Maturation of duration discrimination and group differences LMMs were computed, yielding a total of 184 observations across T1 and T2. Two mean MMR amplitude values were extracted from each dataset (T1 and T2), resulting in four mean values per participant for each condition. At T1, the two values were derived from the 240–340 ms time window and from electrodes where deviant–standard differences were significant according to cluster-based permutation tests. The same procedure was applied to T2, using the 390–490 ms time window and electrodes identified as significant by the cluster-based permutation tests. As the four participants who had missing values either at T1 or T2 were all children with DLD, data had to be assumed to not be Missing at Random (NMAR). Thus, we decided against using Multivariate Imputation by Chained Equations (Buuren & Groothuis-Oudshoorn, 2011) and excluded these participants from further analyses. Below we report the results of the LMM which are relevant to our research questions. Other results from the model are described in Appendix F. The optimal model (see Table 6) revealed a main effect of time indicating significantly more negative amplitudes at T2 compared to T1 in the first time window in the DLD group. The CIs were relatively wide showing moderate precision of the effect, but the effect size was large ( β = -0.54). The time:group interaction showed a stronger negative change in the TD group from T1 to T2 compared to the DLD group (see Figure 6). Based on the CIs this effect was not as precise as the others in the model, but it was large ( β = -0.70). The correlation between the random intercept and random slope was large ( r = -0.928), indicating that participants with more positive amplitudes at T1 (positive intercepts) showed a more negative change over time (negative slopes), compared to those with lower amplitudes at T1. The fixed effects alone accounted for approximately 42% of the variance (marginal R² = 0.42), while the full model, including both fixed and random effects, explained approximately 89% (conditional R² = 0.89). 3.4.2. Behavioral-neurophysiological relationship The two linear-mixed models fitted with the behavioral discrimination task scores at T1 and T2 as the predictor (alongside the interaction of time and time window and group status) for the change in the MMR amplitude from T1 to T2 were not significant. Additionally, no significant result was found with the score of the lexical decision task at T1 as a predictor. However, a significant model fit was found when the behavioral lexical decision task score at T2 was used as a predictor (Table 7) with a small effect size ( β = -0.13) and relatively narrow CIs [-0.25, -0.01]. The fixed effect alone accounted for approximately 38% of the variance (marginal R² = 0.386). The model passed the diagnostic criteria. No group effect was found. The correlation analyses revealed no significant correlations either between the mean MMR amplitude and behavioral task scores at T1 nor at T2. 4 Discussion The aim of the present study was to investigate the maturation of duration and pitch contour perception, the two main cues for distinguishing between quantity degrees, in Estonian children with DLD compared to their TD peers longitudinally across one year. Further, we were interested in whether behavioral discrimination abilities are reflected in the MMR amplitudes in both groups. In contrast to a positive MMR at T1 (Themas et al., 2025a), a negative MMR was observed in the durQ1 condition one year later at T2, with a significant change in amplitude across both groups of children. However, the change in amplitude was less pronounced in the DLD group compared to their TD peers. In the pitchdurQ3 condition, we did not find a reliable MMR neither at T1 (Themas et al., 2025a) nor at T2. Finally, we found a significant relationship between the change of the MMR in the durQ1 condition from T1 to T2 and the d’ scores of the lexical decision task. 4.1 Maturation of the MMR In line with the hypothesis of a significant change in amplitude from T1 to T2 in the durQ1 condition in both groups, we found a significant shift from a positive to a negative MMR amplitude. These findings are in line with evidence that duration change detection is still maturing in children at preschool age. For example, Linnavalli et al. (2018) also found a significant increase in negativity in response to a shortening of a vowel in a consonant-vowel syllable in 5-6-year-olds that were followed longitudinally. Based on the assumption that the negative amplitude represents a more mature brain response to auditory changes (see e.g., Themas et al., 2023), we concluded that, in the present sample, the maturation period concerning duration change detection occurred between the ages of 5 to 7. Interestingly, LMM predictions in our previous study (Themas et al., 2025a) suggested differently, as we found decreased MMR amplitudes (i.e., more negative, mature MMRs) in older children with DLD, but increased MMR amplitudes (i.e., more positive, less mature MMRs) in older TD children (Themas et al., 2025a). This discrepancy to the findings of this longitudinal study might be explained by observations suggesting the maturational trajectory of the MMR to not be linear (Werwach et al., 2022). It could be that the group of TD children have gone through a change towards positivity that was measured at T1 and then started to increase again resulting in a negative MMR at T2. Even if the latter indicates that during childhood one year between repeated measures might be too long of a period to capture nuances in the maturational trajectory of the MMR, the longitudinal approach used in the current study is superior and more robust compared to single measurements, allowing us to draw more general conclusions, namely that the amplitude of the MMR became more mature in both groups. When looking more closely at the latency of the MMRs to the durQ1 condition, we found an increase in latency from T1 to T2 (T1 positivity: 240-340 ms vs. T2 negativity: 390-490 ms) in both groups, which is in contrast to studies showing an MMR latency decrease with age (Choudhury and Benasich, 2011; Morr et al., 2002; Shafer et al., 2000; Shafer et al., 2010; Themas et al., 2023). However, some other studies, though with younger children, also found increases in MMR (and other ERP) latencies with age (Alatorre-Cruz et al., 2023; Čeponienė et al., 2005; Choudhury & Benasich, 2011; Jing & Benasich, 2006; Virtala et al., 2022), possibly related to more sophisticated and at the same time more time-consuming processing of the acoustic stimulus features in older children (Alatorre-Cruz et al., 2023; Čeponienė et al., 2005; Siegler, 2004). In line, it has been suggested that the different MMR polarities reflect distinct underlying neural processes (Rivera-Gaxiola et al., 2005), potentially linked to bottom-up (positive MMR) and top-down (negative MMR) processing pathways, which are thought to involve separate neural generators (see Schaadt et al., 2015). Thus, the negative MMR at T2 may reflect more sophisticated top-down processes that are, however, less mature and more time-consuming in comparison to adults. Again, to prove this assumption, an additional measurement would be needed to show that processes become more efficient, reflected in shorter MMR latencies, when children get older. An alternative explanation for the observed latency differences could be that the positive MMR observed at T1 was not detectable at T2 due to it crossing zero caused by an ongoing polarity shift (Marklund et al., 2019). Thus, instead the negative MMR at T2 may represent a Late Discriminative Negativity (LDN; Cheour et al., 2001) that decreased in latency from T1 (590-690 ms after stimulus onset, see Appendix D) to T2 (390-490 after stimulus onset, see Appendix D). The results on the latency decrease of LDN are contradictory so far. Some report reduction of latency with age (Bishop et al., 2011; V. Mueller et al., 2008), some report no latency change (David et al., 2020; Hommet et al., 2009) and some report increase in latency with age (Virtala et al., 2022). We argue that our findings align with David et al. (2020) and Hommet et al. (2009), because we also observed a small but significant deviant-standard difference cluster at T2, occurring at the same time-window (590-690 after stimulus onset) and over similar region as the LDN at T1 (Appendix D). Thus, the negativity observed between 390-490 ms at T2 is unlikely to reflect an LDN, but rather an MMR. Instead, the LDN at T2 appeared later like at T1 and was smaller in amplitude compared to T1, possibly due to the participants being older at T2 (Bishop et al., 2011; Gumenyuk et al., 2004; Hommet et al., 2009; Linnavalli et al., 2018; Liu et al., 2014; Morales et al., 2022; Torppa et al. 2022). In contrast to our findings on auditory duration discrimination (durQ1) the cluster-based permutation tests showed no significant deviant-standard differences in the pitchdurQ3 condition. We would have expected the MMR to mature from T1, where we also did not find any significant MMRs in this condition, to T2, as, for example, Wu et al., 2025 found an adult like MMN in similarly aged children in response to pitch contour changes. However, in contrast to Wu et al. (2025)’s stimuli, our deviant differed by two, not only one feature from the standard. Even though it has been argued that auditory features are processed by independent neural mechanisms, such that the brain’s response to a deviant differing by two features from the standard should approximately be equal to the sum of each individual response (Paavilainen et al., 2001, also see Ylinen et al., 2005), for the maturing brain it might be difficult to discriminate two auditory features together. Another factor that should have facilitated the perception of pitch contour and duration though is that these features were presented within a meaningful and highly frequent Estonian word. Studies have shown stronger and more mature MMN in adults (Zora et al., 2015) and children (Campos et al., 2023) when phonological changes (e.g., in prosody) are perceived in meaningful words compared to pseudowords. Additionally, word frequency in a language also plays a role, such that high-frequency words are processed more efficiently by both adults (Alexandrov et al., 2011; Aleksandrov et al., 2017) and children (Strotseva-Feinschmidt et al., 2015). Taken these results into account, the pitchdurQ3 condition deviant should have elicited an MMN as it occurs more frequently (563.49 per million tokens; Institute of the Estonian Language, 2023) in Estonian than the durQ1 deviant (29.18 per million tokens; Institute of the Estonian Language, 2023). As, however, only the durQ1 deviant elicited an MMR, the discrimination of the two prosodic cues—pitch contour and duration— in the pitchdurQ3 condition appears to have posed a challenge for participants independent of word frequency. Interestingly, there is one study also showing no significant MMNs in Estonian adults, when discriminating similar pitch contour differences presented in comparable words (Kask et al., 2021). Finally, it should also be considered that the underlying effect could be small and difficult to detect given our relatively small sample size and the inherent noisiness and individual variability of child EEG data. Further investigation is warranted to determine the factors underlying the potential absence of the MMR in response to pitch contour and duration in the current study. 4.2 Group differences in the MMRs As we did not find any MMRs in the pitchdurQ3 condition, group differences could only be analyzed concerning the durQ1 condition. Even though both groups showed a significant shift from positivity to negativity in response to duration changes, we found differences between children with and without DLD, which is line with our hypothesis that group differences remain stable from T1 (see Themas et al., 2025a) to T2 in the durQ1 condition. Specifically, we found a smaller reduction of MMR amplitude from T1 to T2 and less negative amplitudes at T2 in children with DLD compared to their TD peers. While Uwer et al. (2002) did not find any differences in non-linguistic tone duration change discrimination between children with and without DLD that were, this result is not directly comparable to the current research as several studies have shown differences in the MMRs to non-linguistic and linguistic stimuli (Chen et al., 2022; Kuuluvainen et al., 2016; Paquette et al., 2015). However studies with school children with DD showed attenuated negative MMR amplitudes in children with DD compared to their TD peers (e.g., Männel et al., 2017; Schaadt & Männel, 2019). Therefore, we can conclude that at least some of the children in the DLD group might have difficulties perceiving duration changes as indexed by the less negative and therefore less mature MMRs. 4.3 Relationship between neurophysiological and behavioral discrimination Similar to the maturation of the MMR to the durQ1 condition, behavioral results showed increased accuracy and decreased reaction times at T2 compared to T1 when quantity degrees had to be discriminated. Further and in line with the neurophysiological data showing MMRs in the durQ1, but not in the pitchdurQ3 condition, the behavioral results also suggest duration changes to be easier to detect (discrimination of Q1 vs. Q2) compared to duration and pitch changes (discrimination of Q2 vs Q3). Finally, we found children with DLD to perform less accurately in the behavioral tasks compared to their TD peers, which was also represented in the neurophysiological data as the results show a less negative MMR and therefore less mature MMR in the durQ1 condition in the DLD group compared to their TD peers. To further tackle the relationship between neurophysiological and behavioral auditory discrimination, we also analyzed the associations between the change in the MMR amplitude from T1 to T2 and the d’ scores of the behavioral tasks at T1 and T2 at the individual level. We found the scores of the quantity discrimination tasks at T1 and T2 and the score of lexical decision task at T1 to not predict the change in the MMR amplitude. This is in line with several studies that also have not found correspondence between behavioral-neurophysiological discrimination (Bishop et al., 2010; Halliday et al., 2014; McArthur & Bishop, 2005; Nan et al., 2018; Pihko et al., 2006; Uwer et al., 2002). This may reflect differences in task demands between neurophysiological and behavioral discrimination. As such, behavioral discrimination measures additional abilities beyond auditory discrimination, and the ability to behaviorally discriminate auditory features is influenced by these extra demands — especially for children and more so for children with DLD. One may therefore assume that in adult studies the relationship between the MMN amplitude and behavioral auditory discrimination is more consistently found as behavioral task demands do not influence adults as much as children and therefore allow for a more precise estimation of behavioral auditory discrimination, resulting in a clearer relationship between MMN and behavioral performance. This seems to be the case as adult studies generally report a significant relationship (Kujala et al., 2007; Kujala & Näätänen, 2010; Näätänen et al., 2014; Pakarinen et al., 2007; Zhao, 2022). This may explain the significant result in the current research, where the LMM with the score of the lexical decision task at T2 as a predictor, showed that the higher the d′ score (indicating better performance) was, the more negative the MMR amplitude became. At second measurement the children were older, and other task demands besides auditory discrimination may have not influenced them as much as at T1 resulting in a more precise measure of behavioral discrimination. The significant result may also be attributed to methodological aspects. Compared to the quantity discrimination task, the instructions of the lexical decision task were generally easier to follow for the participants as the task was less abstract matching pictures and words pronounced with the correct or incorrect quantity degree. Additionally, the lexical decision task was more motivating to participants to partake, as it included pictures and the duration was shorter compared to the quantity discrimination task. The latter may have yielded a better estimation of children’s discrimination skills and consequently the behavioral-neurophysiological relationship was clearer. Additionally, this result is consistent with studies showing a significant relationship between MMN/MMR amplitude and behavioral performance in tasks requiring higher-order processing, such as phoneme processing and phonological awareness (Linnavalli et al., 2017; Maurer et al., 2003b). Similarly, our lexical decision task engages higher-order processes, as participants must compare incoming stimuli to stored word forms rather than rely solely on auditory discrimination. This suggests that the sensitivity of pre-attentive discrimination denoted by the MMR is important for later stages of speech processing. 4.4 Limitations As already noted, the conclusions of the present study are based on a relatively small sample size, reflecting difficulties in recruiting participants with DLD who met strict diagnostic criteria in a country with a small and geographically dispersed population, such as Estonia, especially during the COVID‑19 pandemic. This limited sample may reduce the generalizability of the findings and impact their statistical reliability. Additionally, the small sample size may have hindered the detection of small effects in statistical models. For example, in the duration change condition, the LMM did not show a significant main effect of group, but the confidence intervals suggested that a true effect size is likely small, and a large sample may be needed to detect such subtle group differences. Related to this, we did not perform a priori power analysis, even though we acknowledge its importance for avoiding Type II error. However, due to the high specificity of the stimuli used in the present study and as the properties of the MMR are highly dependent on the nature of the stimuli (e.g., Themas et al., 2023), extracting an effect size from studies using different stimuli or study population, would have led to an erroneous estimation of the effect size. Given that our study was the first EEG study conducted with Estonian children, performing a power analysis would have required speculation, potentially leading to inaccurate conclusions about the required sample size. 5 Conclusion In conclusion, our findings suggest that MMRs in response to duration changes — within the context of Estonian three-way quantity degrees — are still maturing at preschool age, with slower maturation in DLD participants, suggesting a delay in the maturation of auditory processing. No reliable MMN responses were observed in the combined pitch contour and duration change processing either at T1 or at T2, which may indicate that distinction in question still poses a challenge at late preschool and early school years. Although this represents a novel and important finding, as the perception of the combination of these features has not been reported in the major languages studied to date, further research is warranted to verify these results. Additionally, a significant association was identified between changes in MMR amplitude and performance on the behavioral lexical decision task, which shows that neural sensitivity appears to support behavioral discrimination of words in correct and wrong quantity degrees and pre-attentive discrimination plays a role in higher level speech processing. Even though our findings should be interpreted with caution due to the limited sample size and the inclusion of only two longitudinal measurement points, our study offers valuable insights into the maturation of auditory change detection mechanisms and contributes to a deeper understanding of how neurophysiological and behavioral measures interact in both typically developing and language-impaired Estonian children. Conflict of interest statement The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Data sharing statement The raw EEG data that supports the findings of this study is available on request from the corresponding author in University of Tartu Library DataDOI repository at https://doi.org/10.23673/re-525. The data is not publicly available due to privacy or ethical restrictions. The aggregated EEG data and codes that support the findings of this study are openly available in Zenodo repository at https://doi.org/10.5281/zenodo.15799644. Declaration of Generative AI and AI-assisted technologies in the writing process During the preparation of this work the author(s) used ChatGPT version 4 in order to correct the use of English language. After using this tool, the author(s) reviewed and edited the content as needed and take(s) full responsibility for the content of the publication. Additionally, ChatGPT was used to find mistakes in code and generate some parts of code which was used for data analysis. Before use the generated code was reviewed by the author(s). List of abbreviations AIC – Akaike information criterion ANOVA – analyses of variances β - standardized beta regression coefficient CI – confidence intervals d’ – sensitivity index DD – developmental dyslexia df – degrees of freedom DLD – developmental language disorder durQ1 – deviant stimulus in the first quantity degree, differing from the standard in syllable duration EEG – electroencephalography ERPs – event-related potentials imp. – imperative inf. – infinitive form ISI – interstimulus interval IQ – intelligence quotient LDN – Late Discriminative Negativity LMM – linear mixed model M – mean MADE – the Maryland analysis of developmental EEG max – maximum value min – minimum value MMN – mismatch negativity MMR – mismatch response N - the number of observations NMAR – not missing at random nom. – nominative case p – probability value pitchdurQ3 – deviant stimulus in the third quantity degree, differing from the standard in syllable duration and pitch contour Q1 – first quantity degree Q2 – second quantity degree Q3 – third quantity degree r – Pearson correlation coefficient RT – reaction time R² – R squared SD – standard deviation sg – singular sg2 – second person singular t – a ratio of the estimated fixed effect to its standard error T1 – first assessment T2 – second assessment after one year TD – typically developing TW1 – first time window TW2 – second time window W – Wilcoxon test statistic WPPSI-IV UK – Wechsler Preschool and Primary Scale of Intelligence, fourth edition, United Kingdom version χ² - chi-squared statistic µV – microvolts References Alatorre-Cruz, G. C., Andres, A., Gu, Y., Downs, H., Hagood, D., Sorensen, S. T., Williams, D. K., & Larson-Prior, L. J. (2023). Impact of feeding habits on the development of language-specific processing of phonemes in brain: An event-related potentials study. Frontiers in Nutrition , 10 , 1032413. https://doi.org/10.3389/fnut.2023.1032413 Aleksandrov, A. A., Memetova, K. S., & Stankevich, L. N. (2017). Lexical context affects mismatch negativity caused by pseudowords. Human Physiology , 43 (4), 395–403. https://doi.org/10.1134/S036211971704003X Alexandrov, A. A., Boricheva, D. O., Pulvermüller, F., & Shtyrov, Y. (2011). Strength of Word-Specific Neural Memory Traces Assessed Electrophysiologically. PLoS ONE , 6 (8), e22999. https://doi.org/10.1371/journal.pone.0022999 Bates, D., Mächler, M., Bolker, B., & Walker, S. (2015). Fitting Linear Mixed-Effects Models Using lme4. Journal of Statistical Software , 67 (1). https://doi.org/10.18637/jss.v067.i01 Bishop, D. V. M., Hardiman, M. J., & Barry, J. G. (2011). Is auditory discrimination mature by middle childhood? A study using time-frequency analysis of mismatch responses from 7 years to adulthood. Developmental Science , 14 (2), 402–416. https://doi.org/10.1111/j.1467-7687.2010.00990.x Bishop, D. V. M., Hardiman, M. J., & Barry, R. J. (2010). Lower-Frequency Event-Related Desynchronization: A Signature of Late Mismatch Responses to Sounds, Which Is Reduced or Absent in Children with Specific Language Impairment. Journal of Neuroscience , 17 , 15578–15584. https://doi.org/10.1523/JNEUROSCI.2217-10.2010 Bishop, D. V. M., Hardiman, M., Uwer, R., & Von Suchodoletz, W. (2007). Atypical long-latency auditory event-related potentials in a subset of children with specific language impairment. Developmental Science , 10 (5), 576–587. https://doi.org/10.1111/j.1467-7687.2007.00620.x Bishop, D. V. M., Snowling, M. J., Thompson, P. A., Greenhalgh, T., & and the CATALISE-2 consortium. (2017). Phase 2 of CATALISE: A multinational and multidisciplinary Delphi consensus study of problems with language development: Terminology. Journal of Child Psychology and Psychiatry , 58 (10), 1068–1080. https://doi.org/10.1111/jcpp.12721 Boersma, P., & Weernink, D. (2021). Praat: Doing phonetics by computer (Version 6.1) [Computer software]. Buuren, S. V., & Groothuis-Oudshoorn, K. (2011). mice: Multivariate Imputation by Chained Equations in R. Journal of Statistical Software , 45 (3). https://doi.org/10.18637/jss.v045.i03 Calet, N., Martín-Peregrina, M. Á., Jiménez-Fernández, G., & Martínez-Castilla, P. (2021). Prosodic skills of Spanish-speaking children with developmental language disorder. Int J Lang Commun Disord , 56 , 784–796. https://doi.org/DOI: 10.1111/1460-6984.12627 Campos, A., Tuomainen, J., & Tuomainen, O. (2023). Mismatch Responses to Speech Contrasts in Preschoolers with and without Developmental Language Disorder. Brain Sciences , 14 (1), 42. https://doi.org/10.3390/brainsci14010042 Cantiani, C., Riva, V., Piazza, C., Bettoni, R., Molteni, M., Choudhury, N., Marino, C., & Benasich, A. A. (2016). Auditory discrimination predicts linguistic outcome in Italian infants with and without familial risk for language learning impairment. Developmental Cognitive Neuroscience , 20 , 23–34. https://doi.org/10.1016/j.dcn.2016.03.002 Čeponienė, R., Alku, P., Westerfield, M., Torki, M., & Townsend, J. (2005). ERPs differentiate syllable and nonphonetic sound processing in children and adults. Psychophysiology , 42 (4), 391–406. https://doi.org/10.1111/j.1469-8986.2005.00305.x Chen, A., Peter, V., & Burnham, D. (2022). Development of neural discrimination of pitch across speech and music in the first year of life, a mismatch response study. Language, Cognition and Neuroscience , 37 (9), 1153–1168. https://doi.org/10.1080/23273798.2022.2051571 Chen, Y., Tsao, F.-M., & Liu, H.-M. (2016). Developmental changes in brain response to speech perception in late-talking children: A longitudinal MMR study. Developmental Cognitive Neuroscience , 19 , 190–199. https://doi.org/10.1016/j.dcn.2016.03.007 Cheng, Y.-Y., & Lee, C.-Y. (2018). The Development of Mismatch Responses to Mandarin Lexical Tone in 12- to 24-Month-Old Infants. Frontiers in Psychology , 9 , 448. https://doi.org/10.3389/fpsyg.2018.00448 Cheng, Y.-Y., Wu, H.-C., Shih, H.-Y., Yeh, P.-W., Yen, H.-L., & Lee, C.-Y. (2021). Deficits in Processing of Lexical Tones in Mandarin-Speaking Children With Developmental Language Disorder: Electrophysiological Evidence. Journal of Speech, Language, and Hearing Research , 64 (4), 1176–1188. https://doi.org/10.1044/2021_JSLHR-19-00392 Cheng, Y.-Y., Wu, H.-C., Tzeng, Y.-L., Yang, M.-T., Zhao, L.-L., & Lee, C.-Y. (2015). Feature-specific transition from positive mismatch response to mismatch negativity in early infancy: Mismatch responses to vowels and initial consonants. International Journal of Psychophysiology , 96 (2), 84–94. https://doi.org/10.1016/j.ijpsycho.2015.03.007 Cheng, Y.-Y., Wu, H.-C., Tzeng, Y.-L., Yang, M.-T., Zhao, L.-L., & Lee, C.-Y. (2013). The Development of Mismatch Responses to Mandarin Lexical Tones in Early Infancy. Developmental Neuropsychology , 38 (5), 281–300. https://doi.org/10.1080/87565641.2013.799672 Cheour, M., Korpilahti, P., Martynova, O., & Lang, A.-H. (2001). Mismatch Negativity and Late Discriminative Negativity in Investigating Speech Perception and Learning in Children and Infants. Audiology and Neuro-Otology , 6 (1), 2–11. https://doi.org/10.1159/000046804 Chobert, J., Francois, C., Velay, J.-L., & Besson, M. (2014). Twelve Months of Active Musical Training in 8- to 10-Year-Old Children Enhances the Preattentive Processing of Syllabic Duration and Voice Onset Time. Cerebral Cortex , 24 (4), 956–967. https://doi.org/10.1093/cercor/bhs377 Choudhury, N., & Benasich, A. A. (2011). Maturation of auditory evoked potentials from 6 to 48 months: Prediction to 3 and 4 year language and cognitive abilities. Clinical Neurophysiology , 122 (2), 320–338. https://doi.org/10.1016/j.clinph.2010.05.035 Corriveau, K., Pasquini, E., & Goswami, U. (2007). Basic Auditory Processing Skills and Specific Language Impairment: A New Look at an Old Hypothesis. Journal of Speech, Language, and Hearing Research , 50 (3), 647–666. https://doi.org/10.1044/1092-4388(2007/046) Cumming, R., Wilson, A., & Goswami, U. (2015). Basic auditory processing and sensitivity to prosodic structure in children with specific language impairments: A new look at a perceptual hypothesis. Frontiers in Psychology , 6 . https://doi.org/10.3389/fpsyg.2015.00972 David, C., Roux, S., Bonnet‐Brilhault, F., Ferré, S., & Gomot, M. (2020). Brain responses to change in phonological structures of varying complexity in children and adults. Psychophysiology , 57 (9). https://doi.org/10.1111/psyp.13621 Debnath, R., Buzzell, G. A., Morales, S., Bowers, M. E., Leach, S. C., & Fox, N. A. (2020). The Maryland analysis of developmental EEG (MADE) pipeline. Psychophysiology , 57 (6). https://doi.org/10.1111/psyp.13580 Delorme, A., & Makeig, S. (2004). EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis. Journal of Neuroscience Methods , 134 (1), 9–21. https://doi.org/10.1016/j.jneumeth.2003.10.009 Fisher, D. J., Grant, B., Smith, D. M., & Knott, V. J. (2011). Effects of deviant probability on the ‘optimal’ multi-feature mismatch negativity (MMN) paradigm. International Journal of Psychophysiology , 79 (2), 311–315. https://doi.org/10.1016/j.ijpsycho.2010.11.006 Fong, C. Y., Law, W. H. C., Uka, T., & Koike, S. (2020). Auditory Mismatch Negativity Under Predictive Coding Framework and Its Role in Psychotic Disorders. Frontiers in Psychiatry , 11 , 557932. https://doi.org/10.3389/fpsyt.2020.557932 Friederici, A. D., Friedrich, M., & Weber, C. (2002). Neural manifestation of cognitive and precognitive mismatch detection in early infancy: Neuroreport , 13 (10), 1251–1254. https://doi.org/10.1097/00001756-200207190-00006 Friedrich, M., & Friederici, A. D. (2006). Early N400 development and later language acquisition. Psychophysiology , 43 (1), 1–12. https://doi.org/10.1111/j.1469-8986.2006.00381.x Friedrich, M., Herold, B., & Friederici, A. D. (2009). ERP correlates of processing native and non-native language word stress in infants with different language outcomes. Cortex , 45 (5), 662–676. https://doi.org/10.1016/j.cortex.2008.06.014 Friedrich, M., Weber, C., & Friederici, A. D. (2004). Electrophysiological evidence for delayed mismatch response in infants at-risk for specific language impairment. Psychophysiology , 41 (5), 772–782. doi: 10.1111/j.1469-8986.2004.00202.x. Gillam, S. L., Holbrook, S., & Kamhi, A. G. (2021). Developmental Language Disorder. In J. S. Damico, N. Müller, & M. J. Ball (Eds.), The Handbook of Language and Speech Disorders (1st ed., pp. 171–191). Wiley. https://doi.org/10.1002/9781119606987.ch9 Gumenyuk, V., Korzyukov, O., Alho, K., Escera, C., & Näätänen, R. (2004). Effects of auditory distraction on electrophysiological brain activity and performance in children aged 8–13 years. Psychophysiology , 41 (1), 30–36. https://doi.org/10.1111/1469-8986.00123 Guttorm, T. K., Leppänen, P. H. T., Poikkeus, A.-M., Eklund, K. M., Lyytinen, P., & Lyytinen, H. (2005). Brain Event-Related Potentials (ERPs) Measured at Birth Predict Later Language Development in Children with and Without Familial Risk for Dyslexia. Cortex , 41 (3), 291–303. https://doi.org/10.1016/S0010-9452(08)70267-3 Halliday, L. F., Barry, J. G., Hardiman, M. J., & Bishop, D. V. (2014). Late, not early mismatch responses to changes in frequency are reduced or deviant in children with dyslexia: An event-related potential study. Journal of Neurodevelopmental Disorders , 6 (1), 21. https://doi.org/10.1186/1866-1955-6-21 Hasting, A. S., Winkler, I., & Kotz, S. A. (2008). Early differential processing of verbs and nouns in the human brain as indexed by event‐related brain potentials. European Journal of Neuroscience , 27 (6), 1561–1565. https://doi.org/10.1111/j.1460-9568.2008.06103.x He, C., Hotson, L., & Trainor, L. J. (2007). Mismatch Responses to Pitch Changes in Early Infancy. Journal of Cognitive Neuroscience , 19 (5), 878–892. https://doi.org/10.1162/jocn.2007.19.5.878 Hommet, C., Vidal, J., Roux, S., Blanc, R., Barthez, M. A., De Becque, B., Barthelemy, C., Bruneau, N., & Gomot, M. (2009). Topography of syllable change-detection electrophysiological indices in children and adults with reading disabilities. Neuropsychologia , 47 (3), 761–770. https://doi.org/10.1016/j.neuropsychologia.2008.12.010 Institute of the Estonian Language. (2023). Estonian National Corpus [Corpus]. Sketch Engine. Jing, H., & Benasich, A. A. (2006). Brain responses to tonal changes in the first two years of life. Brain and Development , 28 (4), 247–256. https://doi.org/10.1016/j.braindev.2005.09.002 Karlep, K. (2000). Writing disabilities of Estonian children. Trames. Journal of the Humanities and Social Sciences, 4 (1), 53. https://doi.org/10.3176/tr.2000.1.03 Kask, L., Põldver, N., Lippus, P., & Kreegipuu, K. (2021). Perceptual asymmetries and auditory processing of Estonian quantities. Frontiers in Human Neuroscience , 15 . https://doi.org/10.3389/fnhum.2021.612617 Kujala, T., Lovio, R., Lepistö, T., Laasonen, M., & Näätänen, R. (2006). Evaluation of multi-attribute auditory discrimination in dyslexia with the mismatch negativity. Clinical Neurophysiology , 117 (4), 885–893. https://doi.org/10.1016/j.clinph.2006.01.002 Kujala, T., & Leminen, M. (2017). Low-level neural auditory discrimination dysfunctions in specific language impairment—A review on mismatch negativity findings. Developmental Cognitive Neuroscience , 28 , 65–75. https://doi.org/10.1016/j.dcn.2017.10.005 Kujala, T., & Näätänen, R. (2010). The adaptive brain: A neurophysiological perspective. Progress in Neurobiology , 91 (1), 55–67. https://doi.org/10.1016/j.pneurobio.2010.01.006 Kujala, T., Tervaniemi, M., & Schröger, E. (2007). The mismatch negativity in cognitive and clinical neuroscience: Theoretical and methodological considerations. Biological Psychology , 74 (1), 1–19. https://doi.org/10.1016/j.biopsycho.2006.06.001 Kuuluvainen, S., Alku, P., Makkonen, T., Lipsanen, J., & Kujala, T. (2016). Cortical speech and non-speech discrimination in relation to cognitive measures in preschool children. European Journal of Neuroscience , 43 (6), 738–750. https://doi.org/10.1111/ejn.13141 Kuznetsova, A., Brockhoff, P. B., & Christensen, R. H. B. (2017). lmerTest Package: Tests in Linear Mixed Effects Models. Journal of Statistical Software , 82 (13). https://doi.org/10.18637/jss.v082.i13 Lee, C.-Y., Yen, H., Yeh, P., Lin, W.-H., Cheng, Y.-Y., Tzeng, Y.-L., & Wu, H.-C. (2012). Mismatch responses to lexical tone, initial consonant, and vowel in Mandarin-speaking preschoolers. Neuropsychologia , 50 , 3228-3239. https://doi.org/10.1016/j.neuropsychologia.2012.08.025 Leach, S. C., Morales, S., Bowers, M. E., Buzzell, G. A., Debnath, R., Beall, D., & Fox, N. A. (2020). Adjusting ADJUST: Optimizing the ADJUST algorithm for pediatric data using geodesic nets. Psychophysiology , 57 (8):e13566. https://doi.org/10.1111/psyp.13566 Leonard, L. B. (2014). Children with specific language impairment (2nd ed.). The MIT Press. https://doi.org/10.7551/mitpress/9152.001.0001 Leonard, L. B., & Kueser, J. B. (2019). Five overarching factors central to grammatical learning and treatment in children with developmental language disorder. International Journal of Language & Communication Disorders , 54 (3), 347–361. https://doi.org/10.1111/1460-6984.12456 Leppik, K., Lippus, P., & Asu, E. L. (2023). The perception and production of Estonian vowels and vocalic quantity contrasts by Spanish L1 learners. Ampersand , 11 , 100147. https://doi.org/10.1016/j.amper.2023.100147 Leppänen, P. H. T., Guttorm, T. K., Pihko, E., Takkinen, S., Eklund, K. M., & Lyytinen, H. (2004). Maturational effects on newborn ERPs measured in the mismatch negativity paradigm. Experimental Neurology , 190 , 91–101. https://doi.org/10.1016/j.expneurol.2004.06.002 Linnavalli, T., Putkinen, V., Huotilainen, M., & Tervaniemi, M. (2017). Phoneme processing skills are reflected in children’s MMN responses. Neuropsychologia , 101 , 76–84. https://doi.org/10.1016/j.neuropsychologia.2017.05.013 Linnavalli, T., Putkinen, V., Huotilainen, M., & Tervaniemi, M. (2018). Maturation of Speech-Sound ERPs in 5–6-Year-Old Children: A Longitudinal Study. Frontiers in Neuroscience , 12 , 814. https://doi.org/10.3389/fnins.2018.00814 Lippus, P., Asu, E. L., Teras, P., & Tuisk, T. (2013). Quantity-related variation of duration, pitch and vowel quality in spontaneous Estonian. Journal of Phonetics , 41 (1), 17–28. https://doi.org/10.1016/j.wocn.2012.09.005 Liu, H.-M., Chen, Y., & Tsao, F.-M. (2014). Developmental Changes in Mismatch Responses to Mandarin Consonants and Lexical Tones from Early to Middle Childhood. PLoS ONE , 9 (4), e95587. https://doi.org/10.1371/journal.pone.0095587 Lovio, R., Pakarinen, S., Huotilainen, M., Alku, P., Silvennoinen, S., Näätänen, R., & Kujala, T. (2009). Auditory discrimination profiles of speech sound changes in 6-year-old children as determined with the multi-feature MMN paradigm. Clinical Neurophysiology , 120 (5), 916–921. https://doi.org/10.1016/j.clinph.2009.03.010 Lyu, S., Põldver, N., Kask, L., Wang, L., & Kreegipuu, K. (2024). Native language background affects the perception of duration and pitch. Brain and Language , 256 , 105460. https://doi.org/10.1016/j.bandl.2024.105460 Männel, C., Schaadt, G., Illner, F. K., Van Der Meer, E., & Friederici, A. D. (2017). Phonological abilities in literacy-impaired children: Brain potentials reveal deficient phoneme discrimination, but intact prosodic processing. Developmental Cognitive Neuroscience , 23 , 14–25. https://doi.org/10.1016/j.dcn.2016.11.007 Marklund, E., Schwarz, I.-C., & Lacerda, F. (2019). Amount of speech exposure predicts vowel perception in four- to eight-month-olds. Developmental Cognitive Neuroscience , 36 , 100622. https://doi.org/10.1016/j.dcn.2019.100622 Maurer, U., Bucher, K., Brem, S., & Brandeis, D. (2003a). Development of the automatic mismatch response: From frontal positivity in kindergarten children to the mismatch negativity. Clinical Neurophysiology , 114 (5), 808–817. https://doi.org/10.1016/S1388-2457(03)00032-4 Maurer, U., Bucher, K., Brem, S., & Brandeis, D. (2003b). Altered responses to tone and phoneme mismatch in kindergartners at familial dyslexia risk: NeuroReport , 14 (17), 2245–2250. https://doi.org/10.1097/00001756-200312020-00022 McArthur, G. M., & Bishop, D. V. M. (2005). Speech and non-speech processing in people with specific language impairment: A behavioural and electrophysiological study. Brain and Language, 93 (3), 260-273. https://doi-org.ezproxy.utlib.ut.ee/10.1016/j.bandl.2005.01.002 Meyer, M., Lamers, D., Kayhan, E., Hunnius, S., & Oostenveld, R. (2021). Enhancing reproducibility in developmental EEG research: BIDS, cluster-based permutation tests, and effect sizes. Developmental Cognitive Neuroscience , 52 , 101036. https://doi.org/10.1016/j.dcn.2021.101036 Morales, S., Bowers, M. E., Leach, S. C., Buzzell, G. A., McSweeney, M., Yoder, L., Fifer, W., Elliott, A. J., & Fox, N. A. (2022). Development of auditory change-detection and attentional capture, and their relation to inhibitory control. Psychophysiology , 60 (4). https://doi.org/10.1111/psyp.14211 Morr, M. L., Shafer, V. L., Kreuzer, J. A., & Kurtzberg, D. (2002). Maturation of Mismatch Negativity in Typically Developing Infants and Preschool Children. Ear and Hearing , 23 (2), 118–136. https://doi.org/10.1097/00003446-200204000-00005 Mueller, V., Brehmer, Y., von Oertzen, T., Li, S.-C., & Lindenberger, U. (2008). Electrophysiological correlates of selective attention: A lifespan comparison. BMC Neuroscience , 9 (1), 18. https://doi.org/10.1186/1471-2202-9-18 Mueller, J. L., Friederici, A. D., & Männel, C. (2012). Auditory perception at the root of language learning. Proceedings of the National Academy of Sciences , 109 (39), 15953–15958. https://doi.org/10.1073/pnas.1204319109 Murphy, J. I., Weaver, N. E., & Hendricks, A. E. (2022). Accessible analysis of longitudinal data with linear mixed effects models. Disease Models & Mechanisms , 15 (5), dmm048025. https://doi.org/10.1242/dmm.048025 Näätänen, R., Gaillard, A. W. K., & Mantysalo, S. (1978). Early selective-amention effect on evoked potential reinterpreted. Acta Psychologica , 42 (4), 313–329. https://doi.org/10.1016/0001-6918(78)90006-9 Näätänen, R., Paavilainen, P., Rinne, T., & Alho, K. (2007). The mismatch negativity (MMN) in basic research of central auditory processing: A review. Clinical Neurophysiology , 118 (12), 2544–2590. https://doi.org/10.1016/j.clinph.2007.04.026 Näätänen, R., Pakarinen, S., Rinne, T., & Takegata, R. (2004). The mismatch negativity (MMN): Towards the optimal paradigm. Clinical Neurophysiology , 115 (1), 140–144. https://doi.org/10.1016/j.clinph.2003.04.001 Näätänen, R., Sussman, E. S., Salisbury, D., & Shafer, V. L. (2014). Mismatch Negativity (MMN) as an Index of Cognitive Dysfunction. Brain Topography , 27 (4). https://doi.org/DOI: 10.1007/s10548-014-0374-6 Nan, Y., Liu, L., Geiser, E., Shu, H., Gong, C. C., Dong, Q., Gabrieli, J. D. E., & Desimone, R. (2018). Piano training enhances the neural processing of pitch and improves speech perception in Mandarin-speaking children. Proceedings of the National Academy of Sciences , 115 (28), E6630–E6639. https://doi.org/10.1073/pnas.1808412115 Niemitalo-Haapola, E., Haapala, S., Jansson-Verkasalo, E., & Kujala, T. (2015). Background Noise Degrades Central Auditory Processing in Toddlers. Ear & Hearing , 36 (6), e342–e351. https://doi.org/10.1097/AUD.0000000000000192 Nolan, H., Whelan, R., & Reilly, R. B. (2010). FASTER: Fully automated statistical thresholding for EEG artifact rejection. Journal of Neuroscience Methods , 192 (1), 152–162. https://doi.org/10.1016/j. Jneumeth.2010.07.015 Oostenveld, R., Fries, P., Maris, E., & Schoffelen, J.-M. (2011). FieldTrip: Open Source Software for Advanced Analysis of MEG, EEG, and Invasive Electrophysiological Data. Computational Intelligence and Neuroscience , 2011 , 1–9. https://doi.org/10.1155/2011/156869 Paavilainen, P., Valppu, S., & Näätänen, R. (2001). The additivity of the auditory feature analysis in the human brain as indexed by the mismatch negativity: 1+1≈2 but 1+1+1<3. Neuroscience Letters , 301 (3), 179–182. https://doi.org/10.1016/S0304-3940(01)01635-4 Padrik, M., Hallap, M., & Mäll, R. (2013). 5-6 aastatse laste kõne test. Studium Publishers. Pakarinen, S., Lovio, R., Huotilainen, M., Alku, P., Näätänen, R., & Kujala, T. (2009). Fast multi-feature paradigm for recording several mismatch negativities (MMNs) to phonetic and acoustic changes in speech sounds. Biological Psychology , 82 (3), 219–226. https://doi.org/10.1016/j.biopsycho.2009.07.008 Pakarinen, S., Takegata, R., Rinne, T., Huotilainen, M., & Näätänen, R. (2007). Measurement of extensive auditory discrimination profiles using the mismatch negativity (MMN) of the auditory event-related potential (ERP). Clinical Neurophysiology , 118 (1), 177–185. https://doi.org/10.1016/j.clinph.2006.09.001 Paquette, N., Vannasing, P., Lefrançois, M., Lefebvre, F., Roy, M.-S., McKerral, M., Lepore, F., Lassonde, M., & Gallagher, A. (2013). Neurophysiological Correlates of Auditory and Language Development: A Mismatch Negativity Study. Developmental Neuropsychology , 38 (6), 386–401. https://doi.org/10.1080/87565641.2013.805218 Partanen, E., Pakarinen, S., Kujala, T., & Huotilainen, M. (2013a). Infants’ brain responses for speech sound changes in fast multifeature MMN paradigm. Clinical Neurophysiology , 124 (8), 1578–1585. https://doi.org/10.1016/j.clinph.2013.02.014 Partanen, E., Torppa, R., Pykäläinen, J., Kujala, T., & Huotilainen, M. (2013b). Children’s brain responses to sound changes in pseudo words in a multifeature paradigm. Clinical Neurophysiology , 124 (6), 1132–1138. https://doi.org/10.1016/j.clinph.2012.12.005 Pihko, E., Mickos, A., Kujala, T., Pihlgren, A., Westman, M., Alku, P., Byring, R., & Korkman, M. (2006). Group Intervention Changes Brain Activity in Bilingual Language-Impaired Children. Cerebral Cortex , 17 (4), 849–858. https://doi.org/10.1093/cercor/bhk037 Pettigrew, C., Murdoch, B., Chenery, H., & Kei, J. (2004). Review The relationship between the mismatch negativity (MMN) and psycholinguistic models of spoken word processing. Aphasiology , 18 (1), 3–28. https://doi.org/10.1080/02687030344000463 Polver, S., Háden, G. P., Bulf, H., Winkler, I., & Tóth, B. (2023). Early maturation of sound duration processing in the infant’s brain. Scientific Reports , 13 (1), 10287. https://doi.org/10.1038/s41598-023-36794-x Pulvermüller, F., Kujala, T., Shtyrov, Y., Simola, J., Tiitinen, H., Alku, P., Alho, K., Martinkauppi, S., Ilmoniemi, R. J., & Näätänen, R. (2001). Memory Traces for Words as Revealed by the Mismatch Negativity. NeuroImage , 14 (3), 607–616. https://doi.org/10.1006/nimg.2001.0864 R Core Team. (2023). R: A language and environment for statistical computing (Version 4.3.2) [Software]. R Foundation for Statistical Computing. https://www.R-project.org/ Richards, S., & Goswami, U. (2019). Impaired Recognition of Metrical and Syntactic Boundaries in Children with Developmental Language Disorders. Brain Sciences , 9 (2), 33. https://doi.org/10.3390/brainsci9020033 Rivera‐Gaxiola, M., Silva‐Pereyra, J., & Kuhl, P. K. (2005). Brain potentials to native and non‐native speech contrasts in 7‐ and 11‐month‐old American infants. Developmental Science , 8 (2), 162–172. https://doi.org/10.1111/j.1467-7687.2005.00403.x Schaadt, G., & Männel, C. (2019). Phonemes, words, and phrases: Tracking phonological processing in pre-schoolers developing dyslexia. Clinical Neurophysiology , 130 (8), 1329–1341. https://doi.org/10.1016/j.clinph.2019.05.018 Schaadt, G., Männel, C., van der Meer, E., Pannekamp, A., Oberecker, R., & Friederici, A. D. (2015). Present and past: Can writing abilities in school children be associated with their auditory discrimination capacities in infancy? Research in Developmental Disabilities , 47 , 318–333. https://doi.org/10.1016/j.ridd.2015.10.002 Schaadt, G., Werwach, A., Obrig, H., Friederici, A. D., & Männel, C. (2023). Maturation of consonant perception, but not vowel perception, predicts lexical skills at 12 months. Child Development , 94 (3). https://doi.org/10.1111/cdev.13892 Shafer, V. L., Morr, M. L., Kreuzer, J. A., & Kurtzberg, D. (2000). Maturation of Mismatch Negativity in School-Aged Children. Ear & Hearing , 21 , 242–251. 10.1097/00003446-200006000-00008 Shafer, V. L., Yu, Y. H., & Datta, H. (2010). Maturation of Speech Discrimination in 4- to 7-Yr-Old Children as Indexed by Event-Related Potential Mismatch Responses. Ear and Hearing , 31 (6), 735–745. https://doi.org/10.1097/AUD.0b013e3181e5d1a7 Siegler, R. S. (2004). U-Shaped Interest in U-Shaped Development-and What It Means. Journal of Cognition and Development , 5 (1), 1–10. https://doi.org/10.1207/s15327647jcd0501_1 Squires, N. K., Squires, K. C., & Hillyard, S. A. (1975). Two varieties of long-latency positive waves evoked by unpredictable auditory stimuli in man. Electroencephalography and Clinical Neurophysiology , 38 (4), 387–401. https://doi.org/10.1016/0013-4694(75)90263-1 Strotseva-Feinschmidt, A., Cunitz, K., Friederici, A. D., & Gunter, T. C. (2015). Auditory Discrimination Between Function Words in Children and Adults: A Mismatch Negativity Study. Frontiers in Psychology , 6 . https://doi.org/10.3389/fpsyg.2015.01930 Sundström, S., Lyxell, B., & Samuelsson, C. (2019). Prosodic aspects of repetition in Swedish-speaking children with developmental language disorder. International Journal of Speech-Language Pathology , 21 (6), 623–634. https://doi.org/10.1080/17549507.2018.1508500 The MathWorks Inc. (2015). MATLAB (version: 8.5.0, 2015a) [Software]. Natick, Massachusetts: The MathWorks Inc. https://www.mathworks.com The MathWorks, Inc. (2023). MATLAB (Version 9.14.0, R2023a) [Software]. Natick, MA: The MathWorks, Inc. https://www.mathworks.com Themas, L. (2025). Auditory mismatch response EEG data for Estonian children with DLD [Data set]. University of Tartu Library DataDOI. https://doi.org/10.23673/re-525 Themas, L. & Nickel, E. S. (2025). DLD-Estonian-Quantity/second_round_analysis: Code and Data for Longitudinal EEG Study on Auditory Mismatch Response in Estonian Children with DLD (v1.1). Zenodo. https://doi.org/10.5281/zenodo.15799644 Themas, L., Lippus, P., Padrik, M., Kask, L., & Kreegipuu, K. (2023). Maturation of the mismatch response in pre-school children: Systematic literature review and meta-analysis. Neuroscience & Biobehavioral Reviews , 153 , 105366. https://doi.org/10.1016/j.neubiorev.2023.105366 Themas, L., Nickel, E. S., Lippus, P., Padrik, M., & Kreegipuu, K. (2025a). Exploring pitch and length perception: An EEG study on quantity discrimination in preschoolers with developmental language disorder. Acta Psychologica , 258 , 105157. https://doi.org/10.1016/j.actpsy.2025.105157 Themas, L., Schaadt, G., Lippus, P., Padrik, M., & Kreegipuu, K. (2025b). Maturation of Auditory Pitch and Duration Perception in Children with DLD: Insights from a Longitudinal Study [Preregistration]. OSF. https://doi.org/10.17605/OSF.IO/PCW98 Thiede, A., Virtala, P., Ala-Kurikka, I., Partanen, E., Huotilainen, M., Mikkola, K., Leppänen, P. H. T., & Kujala, T. (2019). An extensive pattern of atypical neural speech-sound discrimination in newborns at risk of dyslexia. Clinical Neurophysiology , 130 (5), 634–646. https://doi.org/10.1016/j.clinph.2019.01.019 Torppa, R., Kuuluvainen, S., & Lipsanen, J. (2022). The development of cortical processing of speech differs between children with cochlear implants and normal hearing and changes with parental singing. Frontiers in Neuroscience , 16 . https://doi.org/10.3389/fnins.2022.976767 Uwer, R., Albrecht, R., & Suchodoletz, W. (2002). Automatic processing of tones and speech stimuli in children with specific language impairment. Developmental Medicine & Child Neurology , 44 (8), 527–532. https://doi.org/10.1111/j.1469-8749.2002.tb00324.x Vandewalle, E., Boets, B., Ghesquière, P., & Zink, I. (2012). Auditory processing and speech perception in children with specific language impairment: Relations with oral language and literacy skills. Research in Developmental Disabilities , 33 (2), 635–644. https://doi.org/10.1016/j.ridd.2011.11.005 Virtala, P., Putkinen, V., Kailaheimo-Lönnqvist, L., Thiede, A., Partanen, E., & Kujala, T. (2022). Infancy and early childhood maturation of neural auditory change detection and its associations to familial dyslexia risk. Clinical Neurophysiology , 137 , 159–176. https://doi.org/10.1016/j.clinph.2022.03.005 Weber, C., Hahne, A., Friedrich, M., & Friederici, A. D. (2005). Reduced stress pattern discrimination in 5-month-olds as a marker of risk for later language impairment: Neurophysiologial evidence. Cognitive Brain Research , 25 (1), 180–187. https://doi.org/10.1016/j.cogbrainres.2005.05.007 Wechsler, D. (2013). Wechsler Preschool & Primary Scale of Intelligence—Fourth UK Edition (WPPSI-IV UK). Werwach, A., Männel, C., Obrig, H., Friederici, A. D., & Schaadt, G. (2022). Longitudinal trajectories of electrophysiological mismatch responses in infant speech discrimination differ across speech features. Developmental Cognitive Neuroscience , 56 , 101127. https://doi.org/10.1016/j.dcn.2022.101127 Wickham, H. (2016). ggplot2: Elegant graphics for data analysis . Springer-Verlag. https://ggplot2.tidyverse.org World Health Organization. (1992). International classification of diseases and related health problems (10th rev.). World Health Organization. Wu, H., Zhang, Y., Liu, Y., Zhang, S., Zhang, L., Shu, H., & Zhang, Y. (2025). The Neural Development of Chinese Lexical Tone Perception: A Mismatch Negativity Study Across Childhood, Adolescence, and Adulthood. Brain Sciences , 15 (1), 93. https://doi.org/10.3390/brainsci15010093 Zhao, T. C. (2022). Neural–Behavioral Relation in Phonetic Discrimination Modulated by Language Background. Brain Sciences , 12 (4), 461. https://doi.org/10.3390/brainsci12040461 Zora, H., Heldner, M., & Schwarz, I.-C. (2016). Perceptual Correlates of Turkish Word Stress and Their Contribution to Automatic Lexical Access: Evidence from Early ERP Components . Frontiers in Neuroscience , 10 . https://doi.org/10.3389/fnins.2016.00007 Zora, H., Schwarz, I.-C., & Heldner, M. (2015). Neural correlates of lexical stress: Mismatch negativity reflects fundamental frequency and intensity. NeuroReport , 26 (13), 791–796. https://doi.org/10.1097/WNR.0000000000000426 Zora, H., Wester, J., & Csépe, V. (2023). Predictions about prosody facilitate lexical access: Evidence from P50/N100 and MMN components. International Journal of Psychophysiology , 194 , 112262. https://doi.org/10.1016/j.ijpsycho.2023.112262 https://doi.org/10.1097/01.wnr.0000185959.11465.9bYlinen, S., Huotilainen, M., & Näätänen, R. (2005). Phoneme quality and quantity are processed independently in the human brain. NeuroReport , 16 (16), 1857–1860. Figures Figure 1. The stages of the longitudinal study. Sampling phase - participant selection; N – number of participants; T1 – experimental measurement at time one; T2 – experimental measurement at time two. Figure 2. Description of Stimuli and experimental procedure. A) structure of the stimulus presentation which ST indicating the standard stimulus and DEV indicating the decent stimulus. B) The wave, spectrogram, pitch contour (blue line on spectrogram) and transcription of the stimuli. Figure 3. ERPs to the standard and deviant stimuli at T1 and T2 separately for children with and without DLD. DurQ1 condition - Depicted are the mean grand averages and their corresponding confidence intervals across all electrodes showing significant effects in the cluster-based permutation test. The gray-shaded areas indicate the time windows of statistically significant differences between ERPs to the standard and deviant stimuli. PitchdurQ3 condition - Depicted are the mean grand averages and their corresponding confidence intervals across electrodes commonly employed in MMR analyses (F3, F4, FC, C3, C4, Cz) as no significant deviant-standard difference was found in cluster-based permutation testing. For visualization purposes, the waveforms were smoothed using a 15-point moving average (equivalent to 30 ms at 500 Hz sampling rate). Figure 4. Difference waves (deviant – standard) at T1 and T2 separately for children with and without DLD. Depicted are the grand averages of the differences waves across all electrodes showing significant effects in the cluster-based permutation test. The gray-shaded areas indicate the time window of statistically significant differences between ERPs to the standard and deviant stimuli in the durQ1 condition. Figure 5. Contrasts of the responses to the deviant and standard stimuli based on the cluster-based permutation tests across both groups (children with and without DLD). White asterisks on the topographic plots indicate significant differences between the response to the standard and deviant stimuli. Figure 6 . Changes in individual MMR values from T1 to T2 with mean estimates from the LMM. Dots - individual raw MMR amplitude values at T1 and at T2. Thin lines - connect each participant’s mean amplitude at T1 to their corresponding value at T2. Thick line – connects the mean LMM estimate of the MMR value at T1 to the corresponding estimate at T2. Tables Table 1. Characteristics of the standard and two deviant stimuli. standard ST Q2 460 270 200 durQ1 DEV 1 Q1 405 190 200 pitchdurQ3 DEV 2 Q3 458 316 150 Note. Stimuli name – names that the stimuli will be referred to; Quant. degree – quantity degree; Total duration – total duration of the stimulus in milliseconds; 1st syllable duration – the duration of the first syllable in the stimulus in milliseconds; Pitch fall – the approximate start of the fall of the pitch contour after stimulus onset in milliseconds. Table 2. Means, minimum values, maximum values and standard deviation of the usable epochs by group and sex at T1 and T2. T1 T2 T1 T2 T1 T2 T1 T2 DLD 578 578 11 125 734 718 176 146 TD 578 578 417 273 706 729 73 133 Male 610 613 11 349 734 729 137 118 Female 593 558 70 125 704 725 136 159 Tabel 3. Sample statistics. T1 N = 23 N = 25 Mode Mode (χ², df) p Sex male (N = 13) male (N = 14) (0, 1) 1 Mat. edu. sec. (N = 8) high. (N = 22) (18, 3) .00036 Hand. right (N = 22) right (N = 22) (7.2 × 10⁻³², 1) 1 Mean SD Mean SD (W) p Age 69 6.0 69 6.2 (306) .92 Lang. test 20 6.9 36 4.7 (535) 3.0 × 10⁻⁸ Nonverbal IQ 96 8.3 108 13 (421) .0019 T2 N = 22 N = 25 Mode Mode (χ², df) p Sex male (N = 12) male (N=14) (0, 1) 1 Mat. edu. sec. (N = 8) high. (N = 22) (16, 3) .00096 Hand. right (N = 21) right (N = 22) (0.15, 1) .070 Mean SD Mean SD (W) p Age 82 6.0 81 5.8 (453) .64 Note. Mat. edu. - maternal education level: sec. – secondary education, high. – higher education; Hand. – handedness; Age is reported in months; Lang. test at T1- the scaled overall score of the language test; Nonverbal IQ at T1 - the overall scaled score of the WPPSI-IV UK nonverbal IQ score. Table 4. Means and standard deviations of the amplitudes of the MMRs at T1 and T2. (r, df) p Mean SD Mean SD T1 T2 T1 T2 T1 T2 T1 T2 Q1 (-0.10, 44) .50 1.01 -0.83 1.94 0.96 1.34 -1.4 1.56 0.86 Q3 (-0.0073, 44) .96 -0.56 -1.4 2.28 2.1 -0.18 -1.0 1.48 1.7 Note. Q1 – MMR in the duration change condition (durQ1). Q3 – MMR in the duration and fundamental frequency contour change (pitchdurQ3). The values are presented in microvolts (µV). Pearsons’s product-moment correlation was calculated between the mean values of MMR-s at T1 and T2. Tabel 5. Means and group differences in d’ and RT values. Sum of errors in discriminating quantity contrasts and group differences at T1 and T2. Quant. diff. T1 T2 T1 T2 T1 (W) p T2 (W) p d’ 1.7 2.3 2.6 3.4 (427) .0012 (346) .00046 RT 1697 1136 1555 1229 (187) .93 (206) .14 Q1 vs Q2 100 45 45 22 (96) .00014 (169) .019 Q1 vs Q3 42 33 17 4 (182) .035 (72) 2.2 × 10⁻⁶ Q2 vs Q3 111 55 76 36 (163) .016 (192) .070 Lex. des. T1 T2 T1 T2 T1 (W) p T2 (W) p d’ 1.6 2.3 2.6 3.0 (403) .0067 (430) .00089 RT 1738 1582 1590 1492 (201) .12 (237) .43 Q1 vs Q2 42 22 10 8 (128) .00071 (160) .0061 Q1 vs Q3 10 7 2 3 (207) .034 (202) 0.18 Q2 vs Q3 51 45 26 16 (161) .012 (146) 0.0040 Note. Quant. diff. - quantity disxrimination task; Lex. des. - lexical decision task; Group diff. - group differences. Table 6. The results of the MMR linear mixed model of the duration deviant condition. Intercept 3.2 1.3 52 2.5 -0.16, 0.64 .016* TW 2 0.59 0.12 92 4.8 0.20, 0.47 5.3 × 10⁻⁶ *** Time 2 -0.95 0.42 50 -2.3 -1.0, -0.07 .028* Group TD 0.60 0.48 46 1.2 -0.20, 0.87 .22 Age 1 -0.04 0.02 46 -2.2 -0.25, -0.02 .030* TW 2: Time 2 -0.94 0.17 92 -5.4 -0.72, -0.34 4.7 × 10⁻⁷ *** Time 2: Group TD -1.2 0.56 46 -2.2 -1.3, -0.08 .031* Significance Codes: *** p < 0.001, ** p < 0.01, * p < 0.05, . p < 0.1, p ≥ 0.1 Note. TW – time-window 1 and 2; Time 1 and 2; Group – DLD (developmental language disorder) and TD (typically developing); Age 1 – age at T1 and reported in months. The intercept is TW1, Time 1 and Group DLD. Table 7. The results of the linear mixed model for the association between the MMR durQ1 Condition and behavioural lexical decision task d′ score. Intercept 1.44 0.40 77 3.57 0.14, 0.70 6.29×10 −4 *** Lex. des. -0.23 0.11 46 -2.13 -0.25, -0.01 .0385* Time 2 -1.63 0.31 54 -5.33 -1.25, -0.58 1.93×10 −6 *** TW2 0.59 0.12 92 4.84 0.20, 0.47 5.27×10 −6 *** TW 2 : Time2 -0.94 0.17 92 -5.42 -0.72, -0.34 4.73×10 −7 *** Significance Codes: *** p < 0.001, ** p < 0.01, * p < 0.05, . p < 0.1, p ≥ 0.1 Note. Lex. des. – lexical decision task; TW – time-window 1 and 2; Time 1 and 2. The intercept is TW 1 and Time 1. Supplementary Material File (figure legends.docx) Download 13.58 KB File (figure3.tif) Download 12.35 MB File (image3.tif) Download 12.35 MB File (table1.docx) Download 22.27 KB File (table2.docx) Download 15.82 KB File (table3.docx) Download 28.40 KB File (table4.docx) Download 13.79 KB File (table5.docx) Download 28.30 KB File (table6.docx) Download 23.28 KB File (table7.docx) Download 22.84 KB Information & Authors Information Version history V1 Version 1 13 February 2026 Copyright This work is licensed under a Non Exclusive No Reuse License. Keywords auditory information processing behavioral developmental disorder eeg erp mmn Authors Affiliations Liis Themas 0000-0002-7615-0895 [email protected] Tartu Ulikool Psuhholoogia instituut View all articles by this author Gesa Schaadt Freie Universitat Berlin Wissenschaftsbereich Psychologie View all articles by this author Pärtel Lippus Tartu Ulikool Eesti ja uldkeeleteaduse instituut View all articles by this author Marika Padrik University of Tartu View all articles by this author Liis Kask Tartu Ulikool Psuhholoogia instituut View all articles by this author Ulvi Vaher Tartu Ulikooli Kliinikum Lastekliinik View all articles by this author Mairi Männamaa Tartu Ulikooli Kliinikum Lastekliinik View all articles by this author Kairi Kreegipuu Tartu Ulikool Psuhholoogia instituut View all articles by this author Metrics & Citations Metrics Article Usage 213 views 96 downloads .FvxKWukQNSOunydq8rnd { width: 100px; } Citations Download citation Liis Themas, Gesa Schaadt, Pärtel Lippus, et al. Auditory Mismatch Response to Pitch and Duration Changes in Children with Developmental Language Disorder: A Longitudinal Approach. Authorea . 13 February 2026. DOI: https://doi.org/10.22541/au.177100799.93641922/v1 If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download. For more information or tips please see 'Downloading to a citation manager' in the Help menu . Format Please select one from the list RIS (ProCite, Reference Manager) EndNote BibTex Medlars RefWorks Direct import Tips for downloading citations document.getElementById('citMgrHelpLink').addEventListener('click', function() { popupHelp(this.href); return false; }); $(".js__slcInclude").on("change", function(e){ if ($(this).val() == 'refworks') $('#direct').prop("checked", false); $('#direct').prop("disabled", ($(this).val() == 'refworks')); }); View Options View options PDF View PDF Figures Tables Media Share Share Share article link Copy Link Copied! Copying failed. Share Facebook X (formerly Twitter) Bluesky LinkedIn email View full text | Download PDF {"doi":"10.22541/au.177100799.93641922/v1","type":"Article"} Now Reading: Share Figures Tables Close figure viewer Back to article Figure title goes here Change zoom level Go to figure location within the article Download figure Toggle share panel Toggle share panel Share Toggle information panel Toggle information panel Go to previous graphic Go to next graphic Go to previous table Go to next table All figures All tables View all material View all material xrefBack.goTo xrefBack.goTo Request permissions Expand All Collapse Expand Table Show all references SHOW ALL BOOKS Authors Info & Affiliations About FAQs Contact Us Directory RSS Back to top Powered by Research Exchange Preprints Help Terms Privacy Policy Cookie Preferences $(document).ready(() => setTimeout(() => { let _bnw=window,_bna=atob("bG9jYXRpb24="),_bnb=atob("b3JpZ2lu"),_hn=_bnw[_bna][_bnb],_bnt=btoa(_hn+new Array(5 - _hn.length % 4).join(" ")); $.get("/resource/lodash?t="+_bnt); },4000)); (function(){function c(){var b=a.contentDocument||a.contentWindow.document;if(b){var d=b.createElement('script');d.innerHTML="window.__CF$cv$params={r:'9fdfbc705d581b23',t:'MTc3OTE1ODQxNw=='};var a=document.createElement('script');a.src='/cdn-cgi/challenge-platform/scripts/jsd/main.js';document.getElementsByTagName('head')[0].appendChild(a);";b.getElementsByTagName('head')[0].appendChild(d)}}if(document.body){var a=document.createElement('iframe');a.height=1;a.width=1;a.style.position='absolute';a.style.top=0;a.style.left=0;a.style.border='none';a.style.visibility='hidden';document.body.appendChild(a);if('loading'!==document.readyState)c();else if(window.addEventListener)document.addEventListener('DOMContentLoaded',c);else{var e=document.onreadystatechange||function(){};document.onreadystatechange=function(b){e(b);'loading'!==document.readyState&&(document.onreadystatechange=e,c())}}}})();