The Impact of COVID-19 on Speech and Language Development of Children Born Between 2018 and 2020 in South Wales. Abstract This study investigates the impact of the COVID-19 pandemic on the speech and language development of children born between 2018 and 2020, a cohort uniquely shaped by social restrictions during formative early developmental years. Speech and language delays can have long-term implications for cognitive, emotional, and academic development. The widespread disruption of early childhood services and reduced social interaction during COVID-19 may have contributed to these challenges. By focusing on children in South Wales, the study addresses gaps in regional and longitudinal data, and evaluates the adequacy of current strategies, including those initiated by the Welsh Government. It draws on developmental theories and existing literature to frame the urgency of a tailored intervention and policy response. Literature Review According to Law et al. (2013), delayed speech and language refers to a situation in which a child’s development of spoken language is significantly behind what is typical for their age. These delays can affect what a child says (expressive language) and what a child understands (receptive language), or can sometimes affect both Language is a crucial component of early childhood development, impacting cognitive abilities, social interaction, and future educational outcomes (NHS, 2022). Children typically babble by 6 months, say their first words around 12 months, combine words at 18–24 months, and form simple sentences by age 3. By the time they reach age 5, many children are able to hold basic conversations and understand simple grammar. Deviations from this trajectory can indicate developmental concerns requiring early intervention according to Welsh Government’s Stages of Speech and Language Development, Talk With Me (2021) program. The COVID-19 pandemic created unprecedented challenges for early childhood development, particularly in the areas of communication and socialisation. Children born between 2018 and 2020 experienced crucial early years during periods of lockdown and social restriction. This literature review supports a study aiming to assess how the COVID-19 pandemic affected speech and language development in pandemic born children, as they now begin to enter formal education settings in 2024–2025. Before the pandemic, speech and language delay was understood as a multifactorial issue with known causes including genetic factors, hearing impairments, developmental disorders such as autism spectrum disorder (ASD), and environmental risks such as low socioeconomic status, parental education levels, and limited exposure to language-rich environments (Law et al., 2000; Bishop et al., 2017). Identification typically occurred through health visitor checks and developmental screening in the early years, with early intervention proven to yield better long-term outcomes (Norbury et al., 2016). Targeted support could help children access education more effectively and improve social engagement. Strategies included speech therapy, language-rich environments, and parent-led interventions (Morgan et al., 2017). B. Early Childhood Development and Communication Early language development is strongly influenced by interaction, play, and relationships (Topping et al., 2013). Social engagement with caregivers and peers supports vocabulary growth, sentence construction, and turn-taking in conversation (Snow, 2006). According to Sylva et al. (2010), quality preschool environments can help mitigate developmental delays, especially for children from disadvantaged backgrounds. Children’s speech and language development thrives on consistent social feedback and peer modelling, these opportunities were significantly reduced during COVID-19 restrictions. C. Covid-19 Pandemic and Its Broader Impact Bronfenbrenner’s Ecological Systems Theory (1979) offers a useful lens for understanding these pandemic impacts. Children’s development is shaped not only by direct relationships (microsystems) but also by the wider societal and policy context (macrosystem). Disruptions to healthcare, education, and parental work routines all interact to influence developmental outcomes. These included reduced access to childcare, fewer health visitor appointments, and limited opportunities for social play. Key services such as speech and language therapy were delayed or delivered remotely, which may not have been as effective for very young children (Royal College of Speech and Language Therapists [RCSLT], 2021). Increased screen time was another by-product of lockdown. Studies have shown that passive screen exposure, in the absence of interaction, can negatively impact speech and language development (Madigan et al., 2020). While some digital tools can support learning, they often lack the reciprocal exchange needed for language acquisition. D. Studies on Covid-19 and Language/Social Development Recent studies have drawn attention to the pandemic’s effects on children born during this period, often referred to as “pandemic babies.” For example, the Welsh Government Delphi study (2021) addresses the impact of COVID-19 on children under age 5 at the time and identified the need for early identification and intervention for speech and language delays in young children across Wales. A current study is also being conducted by the University of London, Born in Covid Year - Core Lockdown Effects (BICYCLE). This study explores not only how COVID-19 impacted children’s speech but also their cognitive development, assessing whether targeted strategies are needed for those born in 2020. McHale et al. (2022) report that children who missed out on nursery due to closures demonstrated lower levels of expressive language. Health visitors in Wales reported rising rates of referrals to speech and language therapists since 2021, though access remains limited due to backlogs (Public Health Wales, 2023). In contrast to this statement, Bajjada et al. (2024) found that due to online support available for families through videos, it was reported that there had not been an increase to speech and language referrals. The Welsh Government has acknowledged this issue through the “Talk With Me” delivery plan, which aims to improve early language skills through better identification and support mechanisms. The plan includes increased training for early years practitioners, better parental engagement, and integrated approaches across education and healthcare sectors (Welsh Government, 2022). E. Reception/Nursery Entry Concerns (2024–2025) Educators and early years professionals have raised concerns about a “developmental gap” among children starting school post-pandemic. An Ofsted report (2023) highlighted that many children are entering Reception with poor vocabulary, limited turn-taking ability, and difficulty with listening and attention. Practitioners have observed that children are often more anxious in group settings, with reduced ability to regulate emotions, both of which can affect communication and learning. In Wales, the Welsh Government has taken further action by developing specific communication support strategies in partnership with Public Health Wales, aiming to reduce inequality and provide earlier, universal screening. Identifying Gaps in the Literature Despite emerging evidence of post-pandemic developmental challenges, several gaps remain in the literature: • Few studies focus specifically on children born between 2018 and 2020, a group uniquely affected by the pandemic during their early developmental windows. • There is limited longitudinal data tracking these children beyond the age of 2 or 3. • UK-based research that evaluates early education settings and regional policy responses is underdeveloped, especially in devolved contexts like Wales. This research seeks to fill this gap by focusing on the speech and language development of this cohort of children in South Wales as they enter early education in 2024–2025. It aims to assess the extent of developmental delays, gather insights from both parents and professionals, and evaluate whether current support structures meet the needs of this cohort Theoretical Frameworks Several developmental theories inform this review. Vygotsky’s Social Development Theory (1978) argues that language arises from social interaction, a view especially relevant when considering the isolation of lockdowns. Bronfenbrenner’s Ecological Systems Theory (1979) provides a multi-layered understanding of how disruptions to family, community, and public services affect development. Additionally, classic language acquisition theories frame the discussion: Skinner emphasised reinforcement and imitation; Chomsky proposed innate grammar structures; and Piaget saw language as part of overall cognitive development. These frameworks help contextualise how environmental disruption might influence developmental outcomes. Conclusion The evidence clearly shows that early speech and language development is influenced by a complex interplay of biological, social, and environmental factors. COVID-19 created widespread disruptions to many of these influences — including healthcare, childcare, socialisation, and education. Emerging research has highlighted increased delays in communication and social skills among children born during or just before the pandemic. However, key gaps remain, especially regarding longitudinal impacts, region-specific data, and children born between 2018 and 2020. As these children now begin school, there is an urgent need for research that evaluates how they have been affected, what interventions are working, and where further support is required. This study is timely and necessary, offering new insights that can inform policy, practice, and targeted early years interventions in the post-pandemic era. Methodology This study adopts a mixed methods approach, combining both quantitative and qualitative research methods to gain complementary and contrasting perspectives on the impact of COVID-19 on the speech and language development of children born between 2018 and 2020 in South Wales. Quantitative research typically involves numerical data and statistical analysis to identify patterns and generalise findings which will be conducted through online surveys, while qualitative research focuses on understanding experiences and perceptions through non-numerical data such as interviews (Cohen et al., 2018). This study employs a mixed methods approach, integrating both to capture complementary perspectives (Hammond and Wellington, 2020). This integration allows for a more comprehensive and nuanced understanding of the research topic, capturing both statistical trends and lived experiences. To ensure a comprehensive understanding of the impact of COVID-19 on children's speech and language delays, this study incorporates cross-sectional surveys, semi-structured interviews with both parents and professionals, with a further study of children in their natural environments, such as schools and nurseries. This multi-method approach aims to better capture the challenges and behaviours of children, understanding developmental delays within their real-life settings. Sampling Strategy To achieve focused and meaningful analysis, the study employs probability sampling within the South Wales region, rather than sampling the entire population of Wales. Sampling across the whole country would introduce wide demographic variability, including differences in socioeconomic status, access to early years services, and urban and rural divides, all of which could confound comparisons and obscure specific effects of COVID-19 on children's speech and language development. Concentrating on South Wales enables greater demographic consistency, facilitating more accurate comparisons and clearer identification of trends and contributing factors within the study’s timeframe and scope. This targeted approach enhances the feasibility of data collection, allowing for in-depth interviews later in the study. Moreover, using probability sampling within this defined area helps ensure the sample remains representative of the local population, increasing the applicability of findings within South Wales and providing valuable insights to inform local policy and early intervention strategies. Further research could later extend these methods to other areas of Wales, allowing comparisons between rural and urban contexts. Participant Recruitment The initial recruitment phase uses a probability sampling approach focusing on families in the Cardiff area of South Wales. Recruitment targets parents engaging with parenting groups on social media via an online survey platform, who have children born between 2018 and 2020. The study excludes all other participants outside this birth window as it is targeting the children entering reception in 2025. Participants are also recruited through local schools, nurseries, and healthcare providers. Additionally, parents, caregivers, and early years professional, including speech and language therapists and teachers are invited to take part. Cardiff was chosen due to its significantly larger population—more than double the size of Swansea, Newport, or Carmarthenshire, according to 2022 mid-year population estimates (StatsWales, 2022). This larger population density is expected to yield a stronger response rate, increasing the likelihood of successful follow-up for later stages such as observations and interviews. Data Collection Procedures The initial online survey will be conducted from December to January. This timeframe is selected because children will have completed their first school or nursery term, allowing practitioners sufficient time to observe and assess developmental progress. Within the first six weeks of starting nursery or primary school in Wales, children undergo a statutory baseline assessment as required by the Curriculum and Assessment (Wales) Act 2021. This assessment evaluates learning areas including language, literacy, and communication skills to establish each child’s developmental stage (Welsh Government, 2021). Similarly, in England, the Reception Baseline Assessment (RBA) is conducted within the first six weeks of reception, assessing early mathematics, literacy, communication, and language skills (Department for Education, 2024). Both assessments provide starting points to measure children’s progress throughout school life. During the Autumn term, any concerns identified through these assessments are usually communicated to families, ensuring parents or caregivers are informed if their child has emerging needs, even if previously unaware. The survey aims to collect data on: • The number of families formally informed of speech and language delays. • Those who believe their child has such delays but have not received feedback from early years settings or schools. • Child’s age and ethnicity • Checklist of challenges the child faces • Whether the child has been referred for additional support • Parent/carer giver work status during lockdown periods. • Did their child attend an early years setting during 2018 to 2020. • Open-ended comments section This survey combines quantitative data, numerical and checklist information as well as qualitative data through open responses, allowing more detailed insight into each child’s situation. Benefits and Limitations of the Survey Method The survey method offers several advantages as it can target a large number of respondents quickly and at low cost. Response deadlines are built into the software limiting unwanted data beyond the collection period and automated data export into spreadsheets reduces manual entry errors, increasing data accuracy (Bhandari, 2022). The targeted geographical scope within Cardiff allows for focused demographic analysis, facilitating comparison of quantitative data such as the prevalence of speech and language delays against local authority referral figures. Qualitative responses enrich the data by capturing information not pre-considered in checklists, helping to generate key concepts, as suggested by Cohen et al. (2018). However, limitations include reliance on social media algorithms for survey visibility, which may skew participant demographics. There is a professional targeted follow up survey with early years settings, schools and health care workers as well as the schools sharing the original survey within their school community, this will give greater credibility for participation. There is a risk of ineligible respondents completing the survey, structural bias based on the researcher’s experiences influencing survey design, and partial or incomplete responses due to lack of understanding or motivation. The accessibility of the survey may also lead to fictitious responses, potentially skewing data. Notably, Likert-scale methods are not used here, as they limit findings to attitudes rather than exploring the underlying reasons behind speech and language delays linked to COVID-19 lockdown periods. Follow-up Semi-Structured Interviews Following survey analysis, a sample of ten families will be selected for semi-structured interviews. This sample size is limited by time constraints and the subsequent workload of qualitative analysis. Families volunteer for follow-up, reducing gatekeeper influence; however, family dynamics may still influence participation. Selection is purposive, focusing on families with similar characteristics and challenges identified from survey responses, to compare home and work experiences during COVID-19 and their relation to children’s speech and language outcomes. The interviews explore factors such as parental working patterns (including working from home), hours and times of day, other household duties, the number of adults and children in the home, and parents’ perceptions of why their child has speech and language delays upon entering nursery or school. Questions also cover any early interventions in place or expected within six months, and parental hopes for these interventions. Semi-structured interviews are preferred over structured formats because they allow flexibility and richer data collection. According to Lambert (2019), this method facilitates rapport building, encourages open discussion, and enables researchers to stay focused while exploring new relevant themes that arise. Bell (2014) also argues that structured interviews can miss opportunities to record rich, nuanced data, underscoring the choice of semi-structured methods here. In comparison Morgan (2012) advocates for using focus groups over one to one interview in situations where participants may find individual face to face interactions intimidating. The choice of one to one interviews was made due to the increased coordination required to organise a focus group that suits everyone’s availability, the potential costs of securing a suitable venue, and the risk that participants may feel judged by others, potentially affecting their honesty when sharing personal experiences. Additionally, one to one interview better protect participants’ identities, which is particularly important in professional settings where individuals may wish to report poor practices, such as whistleblowing. Audio recordings (with permission) are transcribed, permitting detailed qualitative analysis. Limitations include the risk that participants may withhold information if not fully prepared or may respond with socially desirable answers, especially if involved with outside agencies such as social services. The researcher remains aware of the Hawthorne effect, where participants may alter their behaviour during interviews or being observed. The participants are more likely to adhere and be compliant, especially if they are being coerced to attend the research study by a third party or is known to the researcher. This has potential to skewer the findings, producing biased and inaccurate results declares Elston, D (2021). The halo effect is a form of unconscious bias that can influence researchers’ perceptions, particularly during the selection of participants following an initial online survey. To mitigate this bias, participants will be selected based on relevant shared characteristics but will not be personally known to the researcher. While this approach does not entirely eliminate the potential for the halo effect, the researcher will remain critically aware of how participants’ personalities or physical appearances might unconsciously shape assumptions about their honesty or credibility. This consideration is underscored by the findings of Sanrey et al. (2020), whose studies identified the presence of the halo effect in teachers’ judgments across various evaluative measures, which emphasises the importance of researcher self-reflection and bias awareness in qualitative research. Data Triangulation While mixed methods involve using quantitative and qualitive methods within this study, triangulation specifically refers to using multiple methods to validate findings, to enhance reliability (Cohen et al., 2018), this study uses mixed methods primarily to explore different facets of the issue rather than to cross-validate results. The purpose is to research the home environments during COVID-19 and their relation to speech and language delays, rather than to confirm findings across methods. However, triangulation could be employed in future by comparing observational data with survey responses as well as longitudinal data tracking at 12 month intervals, to deepen validity. Conclusion This methodology provides a solid foundation for exploring the impact COVID19 has had on speech and language development, ensuring actionable insights inform policy practice in early childhood education and healthcare with the opportunity to expand on the study in further years. Ethical Considerations This research adheres to a comprehensive set of ethical standards, incorporating institutional, professional, and sector-specific guidance. The principles outlined by the University of Wales Trinity Saint David (UWTSD, 2022), the British Psychological Society (BPS, 2021), and the British Educational Research Association (BERA, 2024) inform every stage of the study design and implementation. These ethical frameworks share a foundational commitment to participant welfare, the minimisation of harm, respect for autonomy, and the pursuit of integrity throughout the research process. Comparative Ethical Frameworks Although unified by shared ethical principles, these three frameworks diverge in their scope and focus. The UWTSD Research Ethics & Integrity Code of Practice provides an institutionally specific framework aligned with broader UK legal obligations, such as the General Data Protection Regulation (GDPR) and Data Protection Act 2018, focusing on compliance, data security, and safeguarding the institution's reputation and the rights of research participants (UWTSD, 2022). In contrast, the BPS (2021) guidelines reflect sector-wide psychological ethics, with an emphasis on risk-based judgments around competence, confidentiality, and consent, particularly in relation to children and vulnerable populations. The BERA (2024) guidelines are more nuanced for educational contexts, placing significant emphasis on the role of children as active agents in research and highlighting the importance of the UNCRC, particularly Articles 3 and 12, which affirm children's rights to be consulted and protected (UNCRC, 1989). BERA provides greater guidance on eliciting children's assent in developmentally appropriate ways and promotes participatory research methods. Unlike BPS, which traditionally places adults in decision-making roles, BERA's model views children not only as vulnerable but also as capable contributors. This distinction is critical in educational research, where children's voices must be authentically represented and ethically protected (BERA, 2024). These principles have been foundational in shaping the methodology for this study, particularly concerning participant recruitment, consent, and data collection involving minors. Informed Consent and Child Assent Voluntary, informed consent is a cornerstone of ethical research. All adult participants will be presented with a detailed Participant Information Sheet explaining the purpose, procedures, risks, and benefits of the study in plain, jargon-free language. Consent will be sought prior to participation via an opt-in checkbox for the online survey, ensuring participants are fully aware of their involvement and their right to withdraw without repercussion. For interviews, signed consent forms will be obtained and stored securely. In the case of participants aged under sixteen, consent will be obtained from a parent or legal guardian. However, following best practice (BERA, 2024; UWTSD, 2022), children will also be involved in an age-appropriate assent process. This recognises that ethical research with children must respect their evolving capacities and ensure their participation is both informed and voluntary. Where children are observed, behavioural cues such as withdrawal, distress, or avoidance will be interpreted as the absence of assent, and participation will be immediately halted. Children aged 16–18 who are judged to have sufficient understanding will be allowed to provide independent consent (BPS, 2021). Surveys will be accessible in participants’ home languages via built-in AI translation tools and Welsh language support through Helo Blod, ensuring linguistic inclusivity and compliance with the Welsh Language Standards. These measures aim to mitigate language as a barrier to informed decision-making. Right to Withdraw Participants will retain the right to withdraw from the research at any point without justification. This right will be highlighted during recruitment, at the beginning of data collection, and during any follow-up communication. If a participant withdraws, any identifiable data (e.g. interview recordings or survey responses) will be excluded from the dataset and securely destroyed unless the data has already been anonymised beyond re-identification. Under no circumstances will participants be pressured or persuaded to re-engage with the study, aligning with ethical principles of autonomy and respect (BPS, 2021; UWTSD, 2022). Risk Management Ethical research requires the proactive identification and mitigation of risk. This study anticipates a range of potential risks, particularly emotional, reputational, and physical. For parents or professionals participating in interviews, sensitive topics may elicit distress, anxiety, or embarrassment, especially if they relate to parenting practices or professional challenges. To mitigate such risks, interviews will be conducted in a sensitive, non-judgemental manner, with participants informed they may skip questions or terminate the session at any time. Debriefing sessions will include information about support services such as Childline, Families First, or professional counselling organisations. For children, emotional risks may arise from being observed or questioned. Observation protocols will be designed to minimise intrusion, and researchers will remain vigilant for non-verbal cues of discomfort or dissent. Assent procedures will be ongoing, with verbal or behavioural withdrawal always respected. Physical risks to participants and researchers will be addressed through a lone working policy. Interviews will take place in neutral, safe environments such as schools or community centres. If home visits are necessary, a lone working risk assessment will be completed, and a check-in system with the research supervisor will be activated. These procedures are compliant with the Health and Safety at Work etc. Act 1974. It is also important to recognise that researchers’ own prior experiences, social positioning, and identities influence how they interpret and engage with the data. As Roberts-Holmes (2014, p. 97) states, “assumptions, beliefs, and values derived from previous experiences, age, gender, sexuality, ethnicity, and class all help to form experiences of life.” These inherent subjectivities may shape the way participants' actions or narratives are perceived and reported. To mitigate this form of bias, reflexivity will be embedded throughout the research process. The researcher will maintain a reflective journal, critically engaging with their positionality and assumptions, in order to enhance transparency, credibility, and trustworthiness of the study’s findings. Institutional risks, such as reputational harm or non-compliance with legal obligations, will be mitigated through strict adherence to UWTSD’s Code of Practice and data governance policies. All tools and platforms used in the research will be pre-approved and GDPR-compliant. Data Protection, Confidentiality, and Anonymity To protect participants’ identities and maintain confidentiality, data will be handled in accordance with the UK GDPR and Data Protection Act 2018. Online survey responses will be collected using secure, encrypted platforms (e.g., JISC Online Surveys), and no unnecessary personally identifiable information (PII) such as names or IP addresses will be collected. For interviews, audio recordings will be encrypted and transferred to a password-protected server immediately after collection. Transcription will involve the removal of all identifiable data, and pseudonyms will be used in reporting. Consent forms will be stored separately from research data—physical copies in locked storage and digital versions in encrypted folders with restricted access. Only the researcher and supervisory team will have access to the raw data. In line with UWTSD’s Research Data Management Policy (2022), data will be retained only as long as necessary. Anonymised data may be archived for up to ten years to support future academic inquiry, in accordance with UKRI and funder expectations. The ICO Data Anonymisation Code of Practice (2012) affirms that where data is anonymised to a standard that prevents identification, it ceases to be classified as personal data, thus reducing compliance burdens and protecting participant privacy. Ethical Integrity and Accountability All procedures outlined have been designed to uphold the highest standards of ethical research. Ethical approval will be sought through UWTSD’s Ethics Committee, with all risk assessments and safeguarding protocols formally documented. Data collection will not commence until approval is granted. Additionally, the researcher will engage in reflective practice throughout the research lifecycle, recognising that ethical decision-making is dynamic and context dependent. Researcher wellbeing has also been considered, particularly due to the emotional labour involved in discussing sensitive issues with parents, children, or professionals. Access to UWTSD’s student counselling services will be available to the researcher as needed. Conclusion In summary, this study has been designed with robust ethical foundations that reflect both statutory obligations and professional best practice. By drawing upon the ethical codes of UWTSD, BPS, and BERA, the research ensures a comprehensive and sensitive approach to issues of informed consent, child participation, data protection, and risk management. Particular care has been taken to respect the rights, agency, and wellbeing of children and vulnerable groups, while also safeguarding the researcher and institutional integrity. Ethical approval, participant dignity, and data security are not treated as procedural hurdles but as essential pillars underpinning the legitimacy and rigour of the research. Ongoing reflection and responsiveness to ethical challenges will be maintained throughout the research process, ensuring that participants are treated with respect, compassion, and fairness at every stage.