Language ENvironment analysis (LENA) system investigation of day long recordings in children: A literature review

From data to discovery: Technology propels speech-language research and theory-building in developmental science

Research on speech and language development has a long history, but in the past decade, it has been transformed by advances in recording technologies, analysis and classification tools, and AI-based language models. We conducted a systematic literature review to identify recently developed (semi-)automatic tools for studying speech-language development and learners' environments in infants and children under the age of 5 years. The Language ENvironment Analysis (LENA) system has been the most widely used tool, with more and more alternative free- and/or open-source tools emerging more recently. Most studies were conducted in naturalistic settings, mostly recording longer time periods (daylong recordings). In the context of vulnerable and clinical populations, most research so far has focused on children with hearing loss or autism. Our review revealed notable gaps in the literature regarding cultural, linguistic, geographic, clinical, and social diversity. Additionally, we identified limitations in current technology-particularly on the software side-that restrict researchers from fully leveraging real-world audio data. Achieving global applicability and accessibility in daylong recordings will require a comprehensive approach that combines technological innovation, methodological rigour, and ethical responsibility. Enhancing inclusivity in participant samples, simplifying tool access, addressing data privacy, and broadening clinical applications can pave the way for a more complete and equitable understanding of early speech and language development. Automatic tools that offer greater efficiency and lower cost have the potential to make science in this research area more geographically and culturally diverse, leading to more representative theories about language development.

Validation of the Language ENvironment Analysis in Swedish Children

PURPOSE

The Language ENvironment Analysis (LENA) is a technological tool designed for comprehensive recordings and automated analysis of young children's daily language and auditory environments. LENA recordings play a crucial role in both clinical interventions and research, offering insights into the amount of spoken language children are exposed to in their homes, including adult word count (AWC) and child vocalization count (CVC). Although LENA was initially developed for American English, it has been validated in various other languages. This study aimed to evaluate the reliability of LENA in a Swedish context by comparing its automatic estimates with human transcriptions.

METHOD

Thirty-six children aged 11-29 months participated in the validation study. A total of 540 min (9 hr) of recordings were transcribed by one experienced speech-language pathologist and two special educators. Interrater agreement over 150 min was notably higher for CVC (intraclass correlation coefficient [ICC] = .97, 95% confidence interval [CI] [.91, .99], r = .96, p < .001) compared to AWC, which yielded an ICC of .82 (95% CI [.15, .96], r = .80, p = .006).

RESULTS

Both estimated CVC and AWC (n = 36) were significantly correlated with human transcriptions (r = .79 and .82, respectively; p < .001).

CONCLUSIONS

The findings align with prior validation studies, indicating that LENA is suitable for use in a Swedish context, particularly for families with children aged 11-29 months. Nonetheless, additional validation studies are necessary, particularly focusing on younger infants, to enhance our understanding of the tool's reliability in preverbal children and their caregivers.

Constraints to Child Language Development in Peri-Urban and Rural Areas: A Mixed-Methods Analysis From Southwestern China

PURPOSE

This mixed-methods study examined how differences in parental time, knowledge, and economic constraints, as well as community socioeconomic contexts, may contribute to differences in home language environment and child language ability outcomes between peri-urban and rural households in China.

METHOD

We conducted an explanatory sequential mixed-methods analysis using data from 158 children aged 18-24 months among peri-urban and rural households with low socioeconomic status (SES) in southwestern China. Audio recordings were collected from each household and analyzed using the Language ENvironment Analysis system. The Mandarin version of the MacArthur-Bates Communicative Development Inventories was administered to each child's primary caregiver. We also conducted qualitative interviews with primary caregivers in 31 peri-urban and 32 rural households. Interviews were recorded, transcribed, and coded.

RESULTS

The quantitative results reveal that children in peri-urban households heard less adult speech and had lower language ability than children in rural households. Directed content analysis of interviews found that peri-urban caregivers faced more severe time constraints and less favorable community socioeconomic contexts than rural primary caregivers. Taken together, these findings suggest that differences in time constraints and community socioeconomic contexts between the two populations are the most likely factors contributing to the inferior language environment and language ability among children in peri-urban households.

CONCLUSION

The mixed-methods study indicated that parental time constraints and community socioeconomic contexts should be considered alongside SES for a comprehensive understanding of factors influencing parental investment in the home language environment in China.

Shaping linguistic input in parent-infant interactions: The influence of the Infant's temperament

Parent-infant interactions highlight the role of parental input, considering both the quality, infant-directed speech, and quantity of interactions, adult words and communicative turns, in these interactions. However, communication is bidirectional, yet little is known about the infant's role in these interactions. This study (n = 35 4-month-old infants) explores how infant-directed speech, the number of adult words and turn-taking (both measured by the LENA system) are correlated with infants' temperament. Our findings reveal that, while mothers use the typical characteristics of infant-directed speech, they are not correlated with the infant's temperament. However, we observe more adult-infant turn-taking in both introverted infants (with lower Surgency scores) and infants with lower attention regulation (with lower Regulatory/Orienting scores). The number of adult words was not correlated with infants' temperament. We suggest that infants with an introverted temperament prefer quieter exchanges that may lead to more turns and that infants with lower attention regulation might create more opportunities for interactions due to their lower level of self-regulation. These findings suggest that infants' temperament is associated with how adults talk with infants (communicative turns) rather than how adults talk to infants (infant-directed speech, number of adult words). Our results underscore the infant's role in parent-infant communication.

Infants' Knowledge of Individual Words: Investigating Links Between Parent Report and Looking Time

Assessing early vocabulary development commonly involves parent report methods and behavioral tasks like looking-while-listening. While both yield reliable aggregate scores, findings are mixed regarding their reliability in measuring infants' knowledge of individual words. Using archival data from 126 monolingual and bilingual 14-31-month-olds, we further examined links across these methods at the word level, while controlling for potentially confounding child-level factors. When data were averaged at the child level, performance on the looking-while-listening task correlated well with parent-reported word production of the same words, as expected. However, mixed-effects model comparisons suggested that at the word level, looking-while-listening performance was significantly predicted by age and total productive vocabulary, but not by parent-reported knowledge of a word once these factors were controlled for. These findings invite careful consideration regarding the adequacy of these two popular methods for capturing children's idiosyncratic knowledge of individual words.

Establishing the reliability of metrics extracted from long-form recordings using LENA and the ACLEW pipeline

Long-form audio recordings are increasingly used to study individual variation, group differences, and many other topics in theoretical and applied fields of developmental science, particularly for the description of children's language input (typically speech from adults) and children's language output (ranging from babble to sentences). The proprietary LENA software has been available for over a decade, and with it, users have come to rely on derived metrics like adult word count (AWC) and child vocalization counts (CVC), which have also more recently been derived using an open-source alternative, the ACLEW pipeline. Yet, there is relatively little work assessing the reliability of long-form metrics in terms of the stability of individual differences across time. Filling this gap, we analyzed eight spoken-language datasets: four from North American English-learning infants, and one each from British English-, French-, American English-/Spanish-, and Quechua-/Spanish-learning infants. The audio data were analyzed using two types of processing software: LENA and the ACLEW open-source pipeline. When all corpora were included, we found relatively low to moderate reliability (across multiple recordings, intraclass correlation coefficient attributed to the child identity [Child ICC], was < 50% for most metrics). There were few differences between the two pipelines. Exploratory analyses suggested some differences as a function of child age and corpora. These findings suggest that, while reliability is likely sufficient for various group-level analyses, caution is needed when using either LENA or ACLEW tools to study individual variation. We also encourage improvement of extant tools, specifically targeting accurate measurement of individual variation.

Child-directed speech in a large sample of U.S. mothers with low income

Research on early language input and socioeconomic status typically relies on correlations in small convenience samples. Using data from Baby's First Years, this paper assesses the causal impact of monthly, unconditional cash transfers on child-directed speech and child vocalizations among a large, racially diverse sample of low-income U.S. mothers and their 1-year-olds (N = 563; 48% girls; 2019-2020). The monthly, unconditional cash transfers did not impact mothers' child-directed speech during a 10-min at-home play session (effect sizes range from -.08 to .02), though there was wide variability within this sample. Future work will assess the impact of the continued cash transfer on children's language input and development over time.

A Feasibility Study of a Music Enrichment Program on Relative Reinforcing Value of Food and Home Environmental Enrichment among Families of Low Socioeconomic Status

BACKGROUND

Emerging evidence suggests that low socioeconomic status (SES) home environments may play a role by promoting excess energy intake through a lack of access to non-food reinforcers. Because of the deleterious effects of SES-related disparities on child health and development, feasible and culturally acceptable interventions are urgently needed. Community-based music enrichment programs may be an ideal intervention strategy.

METHODS

In collaboration with a local non-profit organization and music studio, we conducted a pilot randomized controlled trial to assess the effects of a music enrichment program versus a play date control in a group of 9-24-month-old healthy infants (N = 16). The study was conducted in accordance with the Declaration of Helsinki and approved by the Institutional Review Board of Children's Mercy Hospital Kansas City. This study is registered with clinicaltrials.gov (NCT05868811).

RESULTS

Overall, we found some intervention effects on the relative reinforcing value of food (RRVfood) and the home environmental enrichment measures (i.e., increased music use at home and the home language environment). Our intervention demonstrated large effects on the increased use of music at home. We did not find significant group differences in the RRVfood and home language environment, but some of the effect sizes were medium-to-large. Results also suggest that our intervention is feasible and acceptable. Parent feedback indicated that the intervention was well-liked and that the steps we took to help reduce barriers worked.

CONCLUSIONS

Music enrichment programs may be a high-impact, low-cost strategy to address socioeconomic disparities.

Estimating infants' language exposure: A comparison of random and volume sampling from daylong recordings collected in a bilingual community

In North America, the characteristics of a child's language environment predict language outcomes. For example, differences in bilingual language exposure, exposure to electronic media, and exposure to child-directed speech (CDS) relate to children's language growth. Recently, these predictors have been studied through the use of daylong recordings, followed by manual annotation of audio samples selected from these recordings. Using a dataset of daylong recordings collected from bilingually raised infants in the United States as an example, we ask whether two of the most commonly used sampling methods, random sampling and sampling based on high adult speech, differ from each other with regard to estimating the frequencies of specific language behaviors. Daylong recordings from 37 Spanish-English speaking families with infants between 4 and 22 months of age were analyzed. From each child's recording, samples were extracted in two ways (at random/based on high adult speech) and then annotated for Language (Spanish/English/Mixed), CDS, Electronic Media, Social Context, Turn-Taking, and Infant Babbling. Correlation and agreement analyses were performed, in addition to paired sample t-tests, to assess how the choice of one or the other sampling method may affect the estimates. For most behaviors studied, correlation and agreement between the two sampling methods was high (Pearson r values between 0.79 and 0.99 for 16 of 17 measures; Intraclass Correlation Coefficient values between 0.78 and 0.99 for 13 of 17 measures). However, interesting between-sample differences also emerged: the degree of language mixing, the amount of CDS, and the number of conversational turns were all significantly higher when sampling was performed based on high adult speech compared to random sampling. By contrast, the presence of electronic media and one-on-one social contexts was higher when sampling was performed at random. We discuss advantages of choosing one sampling technique over the other, depending on the research question and variables at hand.

The Predictability of Naturalistic Evaluation of All-Day Recordings for Speech and Language Development

OBJECTIVES

The study aimed to investigate the predictive potential of language environment and vocal development status measures obtained through integrated analysis of Language ENvironment Analysis (LENA) recordings during the prelinguistic stage for subsequent speech and language development in Korean-acquiring children. Specifically, this study explored whether measures from both LENA-automated analysis and human coding at 6-8 months and 12-14 months of age predict vocabulary and phonological development at 18-20 months.

METHOD

One-day home recordings from 20 children were collected using a LENA recorder at 6-8 months, 12-14 months, and 18-20 months. Both LENA-automated measures and measures from human coding were obtained from recordings at 6-8 months and 12-14 months. The number of different words, consonant inventory, and utterance structure inventory were identified from recordings of 18-20 months. Correlation and multiple regression analyses were performed to investigate whether measures related to early language environment and child vocalization at 6-8 months and 12-14 months were predictive of vocabulary and phonological measures at 18-20 months.

RESULTS

The results showed that the two main LENA-automated measures, conversational turn count (CTC) and child vocalization count, were positively correlated with all vocabulary and phonological measures at 18-20 months. Multiple regression analysis revealed that CTC during the prelinguistic stages was the most significant predictor of a number of different words, consonant inventory, and utterance structure inventory at 18-20 months. Also, adult word count in LENA-automated measures, child-directed speech ratio, and canonical babbling ratio measured by human coding significantly predicted some vocabulary and phonological measures at 18-20 months.

CONCLUSION

This study highlights the multifaceted nature of language acquisition and collectively emphasizes the value of considering both quantitative and qualitative aspects of language input to understand early language development in children.

The Effect of LENA (Language ENvironment Analysis) for Children with Hearing Loss in Denmark including a Pilot Validation for the Danish Language

Background/Objectives: This study aimed to investigate whether day-long recordings with Language Environment Analysis (LENA) can be utilized in a hospital-based Auditory Verbal Therapy (AVT) program in Denmark for children with hearing loss and to conduct a pilot validation in the Danish language. Methods and materials: A license for the LENA system (LENA SP) was purchased, and trials were offered to three families enrolled in the AVT program. Each family made two day-long recordings with 3-4 months in between and received feedback during the therapy sessions. From 18 × 10-min clips randomly pulled out of the recordings, a comparison of adult word counts (AWC) between the LENA algorithm counts and the counts made by two human transcribers was made and used for the pilot validation. Results: LENA proved to be valuable as a guiding tool for Danish parents. Pilot validation showed good correlations and an acceptable limit of agreement (LoA). Conclusions: LENA holds the potential for Danish validation and use in AVT/clinical practice. When used in clinical practice, parents must be informed of the biases and limitations, and possible ethical issues must be considered. Because of the GDPR rules, there is a need to discuss the possibility of implementing this tool clinically in Denmark and the EU.

Automatic assessment of infant carrying and holding using at-home wearable recordings

Assessing infant carrying and holding (C/H), or physical infant-caregiver interaction, is important for a wide range of contexts in development research. An automated detection and quantification of infant C/H is particularly needed in long term at-home studies where development of infants' neurobehavior is measured using wearable devices. Here, we first developed a phenomenological categorization for physical infant-caregiver interactions to support five different definitions of C/H behaviors. Then, we trained and assessed deep learning-based classifiers for their automatic detection from multi-sensor wearable recordings that were originally used for mobile assessment of infants' motor development. Our results show that an automated C/H detection is feasible at few-second temporal accuracy. With the best C/H definition, the automated detector shows 96% accuracy and 0.56 kappa, which is slightly less than the video-based inter-rater agreement between trained human experts (98% accuracy, 0.77 kappa). The classifier performance varies with C/H definition reflecting the extent to which infants' movements are present in each C/H variant. A systematic benchmarking experiment shows that the widely used actigraphy-based method ignores the normally occurring C/H behaviors. Finally, we show proof-of-concept for the utility of the novel classifier in studying C/H behavior across infant development. Particularly, we show that matching the C/H detections to individuals' gross motor ability discloses novel insights to infant-parent interaction.

The feasibility of measuring fidelity of implementation in parent-child interaction therapy: A clinician and parent fidelity study

PURPOSE

Measuring fidelity of implementation in parent-child interaction therapy (PCIT) involves assessing the training delivered by clinicians and how parents implement the techniques with their children. The aim of this study was to determine the feasibility of measuring fidelity of implementation for a PCIT intervention designed for young children with Down syndrome.

METHOD

We applied a framework to measure dosage, adherence, quality, and participant responsiveness using a mixed methods approach with observational and interview data.

RESULT

Our results showed that clinicians delivered 94% of the planned dosage; they adhered to the goals of program and reached the quality criterion in 4/6 rated sessions. Parents described their ability to engage with the program and perceived that it changed how they interacted and communicated with their children. Parents were unable to collect dosage data, but did adhere to 7/9 of the targeted techniques and met the quality criterion on 6/9 of these. It was also possible to measure the children's responsiveness scores when interacting with parents during the intervention.

CONCLUSION

This study revealed the opportunities and challenges that occur when measuring fidelity of implementation. There is a need to refine definitions of fidelity measures and to develop appropriate measurement tools so that a more consistent and useful framework can be used by speech-language pathologists (SLPs) to measure fidelity.

Fathers' unmodulated prosody in child-directed speech

Mothers and fathers modify prosodic characteristics of child-directed speech relative to adult-directed speech. Evidence suggests that mothers and fathers may differ in how they use child-directed speech as communicative partners. Thus, fathers create communicative challenges during father-child interaction that facilitate the child's adaptation to a wider potential range of interlocutors. In this paper, speech production differences are examined between mothers and fathers in child-directed speech to toddlers as compared to adult-directed speech. Using a longitudinal, large-scale design of audio recordings in naturalistic environments and automatic speech processing techniques, it was found that mothers, but not fathers, increased their fundamental frequency when addressing their toddlers. The results suggest that fathers do not modulate the prosody of their speech in the same way as mothers when communicating with their toddlers. Findings have implications for emotional and communicative practices of fathers compared to mothers and the differential role each plays in child development.

Factors of parental investment in the home language environment in peri-urban China: A mixed methods study

The home language environment is a critical point of investment in early language skills. However, few studies have quantitatively measured the home language environment of low-socioeconomic-status households in non-western settings. This mixed methods study describes the home language environment and early child language skills among households in a low-socioeconomic-status, peri-urban district of Chengdu, China, and identifies factors influencing parental investment in the home language environment. Audio recordings were collected from 81 peri-urban households with children ages 18-24 months and analysed using the Language Environment Analysis (LENATM) system. The Mandarin version of the MacArthur-Bates Communicative Development Inventory was administered to each child's primary caregiver. The quantitative results revealed large variation in home language environments and child language skills among the sample, with relatively low average scores when compared to other Chinese samples. Qualitative interviews with a subset of 31 caregivers revealed that many caregivers face constraints on their knowledge of interactive parenting, compounded, in some households, by time constraints due to work or household responsibilities. The findings indicate a need for increased sources of credible parenting information for peri-urban caregivers of young children to promote investment in the home language environment.

Investigating Children's Interactions in Preschool Classrooms: An Overview of Research Using Automated Sensing Technologies

New technologies that combine digital sensors with automated processing algorithms are now being deployed to study preschool classrooms. This article provides an overview of these new sensing technologies, focusing on automated speaker classification, the analysis of children's and teachers' speech, and the detection and analysis of their movements over the course of the school day. Findings from recent studies utilizing these technologies are presented to illustrate the contribution of these sensing technologies to our understanding of classroom processes that predict children's language and social development. In particular, the potential to collect extended real-time data on the speech and movement of all children and teachers in a classroom provides a broader window on the variability of individual children's interactions with peers and teachers and their integration into classroom social networks. The article describes current challenges related to the use of sensing technologies in preschool settings, as well as advances that may overcome these challenges and allow for more in-depth investigations of children's early classroom experiences.

Early childhood education language environments: considerations for research and practice

The importance of developing early language and literacy skills is acknowledged by the United Nations Educational, Scientific, and Cultural Organization (UNESCO) as a global human rights issue. Indeed, research suggests that language abilities are foundational for a host of cognitive, behavioral, and social-emotional outcomes. Therefore, it is critical to provide experiences that foster language acquisition across early learning settings. Central to these efforts is incorporating assessments of language environments into research and practice to drive quality improvement. Yet, several barriers may be preventing language environment assessments from becoming widely integrated into early education. In this brief, we review evidence on the types of experiences that promote language development, describe characteristics of language environment assessments, and outline practical and philosophical considerations to assist with decision-making. Further, we offer recommendations for future research that may contribute knowledge regarding strategies to assess and support language development. In addressing both areas, we highlight the potential for early childhood language environments to advance equity.

Cultural differences in auditory ecology

Demographic differences in acoustic environments are usually studied using geographic area monitoring. This approach, however, may miss valuable information differentiating cultures. This motivated the current study, which used wearable sound recorders to measure noise levels and speech-to-noise ratios (SNRs) in the immediate acoustic environment of Latinx and European-American college students. Latinx experienced higher noise levels (64.8 dBC) and lower SNRs (3.7 dB) compared to European-Americans (noise levels, 63 dB; SNRs, 5.4 dB). This work provides a framework for a larger study on the impact of culture on auditory ecology.

The Language Environment at Home of Children With (a Suspicion of) a Developmental Language Disorder and Relations With Standardized Language Measures

PURPOSE

This study compares the home language environments of children with (a suspicion of) developmental language disorder (DLD) with that of children with typical development (TD). It does so by adopting new technology that automatically provides metrics about children's language environment (Language ENvironment Analysis [LENA]). In addition, relationships between LENA metrics and standardized language tests are explored in the DLD group.

METHOD

Ninety-nine 2- to 4-year-old toddlers participated: 59 with (a suspicion of) DLD and 40 with TD. LENA metrics on adult word count, conversational turn count, and child vocalization count were obtained. For all children, data on parental education and multilingualism were available. In the DLD group, data were collected on receptive and expressive vocabulary and grammar, and on nonverbal intelligence, using standardized tests.

RESULTS

We found lower adult word count, conversational turn count, and child vocalization count in the DLD group, independent of multilingualism but not of parental education. In the DLD group, receptive vocabulary was related to conversational turn count and child vocalization count, but not to adult word count. Expressive vocabulary, receptive grammar, and expressive grammar were not related to LENA metrics.

CONCLUSIONS

Toddlers with (a suspicion of) DLD vocalize less at home than children with TD. They also hear fewer adult words and experience fewer conversational turns. Children with DLD's language outcomes are to a limited extent related to language environment at home. Conversational turns and child vocalizations are in this respect more important than adult words, in line with findings for TD populations.

The validity of LENA technology for assessing the linguistic environment and interactions of infants learning Hebrew and Arabic

The present study assessed LENA's suitability as a tool for monitoring future language interventions by evaluating its reliability, construct validity, and criterion validity in infants learning Hebrew and Arabic, across low and high levels of maternal education. Participants were 32 infants aged 3 to 11 months (16 in each language) and their mothers, whose socioeconomic status (SES) was determined based on their years of education (H-high or L-low ME-maternal education). The results showed (1) good reliability for the LENA's automatic count on adult word count (AWC), conversational turns (CTC), and infant vocalizations (CVC), based on the positive associations and fair to excellent agreement between the manual and automatic counts; (2) good construct validity based on significantly higher counts for HME vs. LME and positive associations between LENA's automatic vocal assessment (AVA) and developmental questionnaire (DA) and age; and (3) good concurrent criterion validity based on the positive associations between the LENA counts for CTC, CVC, AVA, and DA and the scores on the preverbal parent questionnaire (PRISE). The present study supports the use of LENA in early intervention programs for infants whose families speak Hebrew or Arabic. The LENA could be used to monitor the efficacy of these programs as well as to provide feedback to parents on the amount of language experience their infants are getting and their progress in vocal production. The results also indicate a potential utility of LENA in assessing linguistic environments and interactions in Hebrew- and Arabic-speaking infants with developmental disorders, such as hearing impairment and cerebral palsy.

The home language environment and early language ability in rural Southwestern China

Using premier Language Environment Analysis technology to measure and analyze the home language environment, this observational study aims to describe the home language environment and child language ability, drawing on empirical data from 77 households with children aged 18–24 months from rural China. The results show large variation in measures of the home language environment and early language ability, similar to other rural Chinese samples. Results also demonstrate significant correlations between child age and the home language environment, maternal employment and the home language environment, father’s educational attainment and the home language environment, adult–child conversations and early language ability, and child vocalizations and early language ability.

The home language environment in rural China: variations across family characteristics

BACKGROUND

A rich language environment is an important element of a nurturing home environment. Despite their proven importance, vocabulary and conversation have been shown to vary widely across households-even within the same socio-economic class. One significant gap in the existing literature is its nearly exclusive geographic focus on Western and developed settings, with little attention given to poorer communities in lower/middle income countries. The purpose of this study was to empirically illustrate the characteristics of the home language environment in the low SES, non-Western cultural setting of rural China.

METHODS

Using Language Environment Analysis (LENA) automated language-analysis system, this study measured the home language environment of 38 children aged 20-27 months in Northwest rural China. Our primary measures of the home language environment were Adult Word Count (AWC), Conversational Turn Count (CTC) and Child Vocalization Count (CVC). Multivariate linear regression models were used to examine the association between home language environment and family/child characteristics, and language skills (Measured by MacArthur-Bates Communicative Developmental Inventory score).

RESULTS

In this paper, by comparison, we found that the home language environment of our rural sample fell far behind that of urban households. We also identify significant, positive correlations between language skills and both AWC and CTC. Our analysis finds no significant correlations between home language environment and family/child characteristics.

CONCLUSION

In this paper, we present the first ever findings using the LENA system to measure the home language environment of young children from poor rural communities in China. We found that the home language environment of lower-SES household was significantly worse than high-SES households, and demonstrated the importance of the home language environment to language skills, pointing to a need for more high-quality studies of the home language environment in rural China to better understand possible mechanisms behind low levels of parent-child language engagement and ways to improve the home language environment.

Talk of the Town mobile app platform: New method for engaging family in STEM learning and research in homes and communities

Children do not just learn in the classroom. They engage in "informal learning" every day just by spending time with their family and peers. However, while researchers know this occurs, less is known about the science of this learning-how this learning works. This is so because investigators lack access to those moments of informal learning. In this mini-review we present a technical solution: a mobile-based research platform called "Talk of the Town" that will provide a window into children's informal learning. The tool will be open to all researchers and educators and is flexibly adaptable to these needs. It allows access to data that have never been studied before, providing a means for developing and testing vast educational interventions, and providing access to much more diverse samples than are typically studied in laboratories, homes, and science museums. The review details the promise and challenges associated with these new methods of data collection and family engagement in STEM learning sciences.

Parental Language Input Predicts Neuroscillatory Patterns Associated with Language Development in Toddlers at Risk of Autism

In this study we investigated the impact of parental language input on language development and associated neuroscillatory patterns in toddlers at risk of Autism Spectrum Disorder (ASD). Forty-six mother-toddler dyads at either high (n = 22) or low (n = 24) familial risk of ASD completed a longitudinal, prospective study including free-play, resting electroencephalography, and standardized language assessments. Input quantity/quality at 18 months positively predicted expressive language at 24 months, and relationships were stronger for high-risk toddlers. Moderated mediations revealed that input-language relationships were explained by 24-month frontal and temporal gamma power (30-50 Hz) for high-risk toddlers who would later develop ASD. Results suggest that high-risk toddlers may be cognitively and neurally more sensitive to their language environments, which has implications for early intervention.

Multimodal Classification of Teaching Activities from University Lecture Recordings

The way of understanding online higher education has greatly changed due to the worldwide pandemic situation. Teaching is undertaken remotely, and the faculty incorporate lecture audio recordings as part of the teaching material. This new online teaching–learning setting has largely impacted university classes. While online teaching technology that enriches virtual classrooms has been abundant over the past two years, the same has not occurred in supporting students during online learning. To overcome this limitation, our aim is to work toward enabling students to easily access the piece of the lesson recording in which the teacher explains a theoretical concept, solves an exercise, or comments on organizational issues of the course. To that end, we present a multimodal classification algorithm that identifies the type of activity that is being carried out at any time of the lesson by using a transformer-based language model that exploits features from the audio file and from the automated lecture transcription. The experimental results will show that some academic activities are more easily identifiable with the audio signal while resorting to the text transcription is needed to identify others. All in all, our contribution aims to recognize the academic activities of a teacher during a lesson.

Vocal Turn-Taking in Families With Children With and Without Hearing Loss

OBJECTIVE

Vocal turn-taking is an important predictor of language development in children with and without hearing loss. Most studies have examined vocal turn-taking in mother-child dyads without considering the multitalker context in a child's life. The present study investigates the quantity of vocal turns between deaf and hard-of-hearing children and multiple members of their social environment.

DESIGN

Participants were 52 families with children who used hearing aids (HA, mean age 26.3 mo) or cochlear implants (CI, mean age 63.2 mo) and 27 families with normal-hearing (NH, mean age 26.6 mo) children. The Language ENvironment Analysis system estimated the number of conversational turns per hour (CTC/hr) between all family members (i.e., adult female, adult male, target child, and other child) during full-day recordings over a period of about 1 year.

RESULTS

The CTC/hr was lower between the target child and the adult female or adult male in the CI compared with the HA and NH groups. Initially, CTC/hr was higher between the target child and the adult female than between the adult male or the other child. As the child's age increased, turn-taking between the target child and the adult female increased in comparison to that between the target child and the adult male. Over time, turn-taking between the target child and the other child increased and exceeded turn-taking between the target child and the adult caregivers. The increase was observed earlier in families with siblings compared with those without.

CONCLUSIONS

The quantity of vocal turn-taking depends on the degree of child hearing loss and the relationship between the children and the members of their social environment. Longitudinally, the positive effect of an assistive device on the quantity of turns between the children and their family members was found. The effect was stronger in families with siblings.

Using big data from long-form recordings to study development and optimize societal impact

Big data are everywhere. In this chapter, we focus on one source: long-form, child-centered recordings collected using wearable technologies. Because these recordings are simultaneously unobtrusive and encompassing, they may be a breakthrough technology for clinicians and researchers from several diverse fields. We demonstrate this possibility by outlining three applications for the recordings-clinical treatment, large-scale interventions, and language documentation-where we see the greatest potential. We argue that incorporating these recordings into basic and applied research will result in more equitable treatment of patients, more reliable measurements of the effects of interventions on real-world behavior, and deeper scientific insights with less observational bias. We conclude by outlining a proposal for a semistructured online platform where vast numbers of long-form recordings could be hosted and more representative, less biased algorithms could be trained.

Simulating Speech Error Patterns Across Languages and Different Datasets

Children's speech acquisition is influenced by universal and language-specific forces. Some speech error patterns (or phonological processes) in children's speech are observed in many languages, but the same error pattern may have different effects in different languages. We aimed to explore phonological effects of the same speech error patterns across different languages, target audiences and discourse modes, using a novel method for large-scale corpus investigation. As an additional aim, we investigated the face validity of five different phonological effect measures by relating them to subjective ratings of assumed effects on intelligibility, as provided by practicing speech-language pathologists. Six frequently attested speech error patterns were simulated in authentic corpus data: backing, fronting, stopping, /r/-weakening, cluster reduction and weak syllable deletion-each simulation resulting in a "misarticulated" version of the original corpus. Phonological effects were quantified using five separate metrics of phonological complexity and distance from expected target forms. Using Swedish child-speech data as a reference, phonological effects were compared between this reference and a) child speech in Norwegian and English, and b) data representing different modes of discourse (spoken/written) and target audiences (adults/children) in Swedish. Of the speech error patterns, backing-the one atypical pattern of those included-was found to cause the most detrimental effects, across languages as well as across modes and speaker ages. However, none of the measures reflects intuitive rankings as provided by clinicians regarding effects on intelligibility, thus corroborating earlier reports that phonological competence is not translatable into levels of intelligibility.

Quantifying Everyday Ecologies: Principles for Manual Annotation of Many Hours of Infants' Lives

Everyday experiences are the experiences available to shape developmental change. Remarkable advances in devices used to record infants' and toddlers' everyday experiences, as well as in repositories to aggregate and share such recordings across teams of theorists, have yielded a potential gold mine of insights to spur next-generation theories of experience-dependent change. Making full use of these advances, however, currently requires manual annotation. Manually annotating many hours of everyday life is a dedicated pursuit requiring significant time and resources, and in many domains is an endeavor currently lacking foundational facts to guide potentially consequential implementation decisions. These realities make manual annotation a frequent barrier to discoveries, as theorists instead opt for narrower scoped activities. Here, we provide theorists with a framework for manually annotating many hours of everyday life designed to reduce both theoretical and practical overwhelm. We share insights based on our team's recent adventures in the previously uncharted territory of everyday music. We identify principles, and share implementation examples and tools, to help theorists achieve scalable solutions to challenges that are especially fierce when annotating extended timescales. These principles for quantifying everyday ecologies will help theorists collectively maximize return on investment in databases of everyday recordings and will enable a broad community of scholars—across institutions, skillsets, experiences, and working environments—to make discoveries about the experiences upon which development may depend.

Dimensions of the language environment in infancy and symptoms of psychopathology in toddlerhood

The quality of the early environment influences the development of psychopathology. Children who are deprived of sufficient environmental enrichment in infancy may be at higher risk for developing symptoms of psychopathology in toddlerhood. In this study, we investigated the prospective association between naturalistic measures of adult language input obtained through passive monitoring of infants' daily environments and emerging psychopathology in toddlerhood. In a sample of 100 mothers and their infants recruited from the community (mean age [range] = 6.73 [5-9] months), we used the Language ENvironment Analysis (LENA) system to measure multiple dimensions of infants' language environments, including both the quantity and consistency of adult speech and conversational turns in infants' daily lives as well as the quantity of infant vocalizations. Subsequently, during toddlerhood (mean age [range] = 18.29 [17-21] months), mothers reported on their children's symptoms of psychopathology. Infants who experienced more consistent adult speech and conversational turns had lower symptoms of psychopathology in toddlerhood, independent of negative emotionality in infancy, maternal depressive symptoms, and laboratory-based measures of maternal sensitivity. These findings have implications for the measurement of environmental factors that may confer risk and resilience to emerging psychopathology.

Vocal development in a large-scale crosslinguistic corpus

This study evaluates whether early vocalizations develop in similar ways in children across diverse cultural contexts. We analyze data from daylong audio recordings of 49 children (1-36 months) from five different language/cultural backgrounds. Citizen scientists annotated these recordings to determine if child vocalizations contained canonical transitions or not (e.g., "ba" vs. "ee"). Results revealed that the proportion of clips reported to contain canonical transitions increased with age. Furthermore, this proportion exceeded 0.15 by around 7 months, replicating and extending previous findings on canonical vocalization development but using data from the natural environments of a culturally and linguistically diverse sample. This work explores how crowdsourcing can be used to annotate corpora, helping establish developmental milestones relevant to multiple languages and cultures. Lower inter-annotator reliability on the crowdsourcing platform, relative to more traditional in-lab expert annotators, means that a larger number of unique annotators and/or annotations are required, and that crowdsourcing may not be a suitable method for more fine-grained annotation decisions. Audio clips used for this project are compiled into a large-scale infant vocalization corpus that is available for other researchers to use in future work.

Describing Vocalizations in Young Children: A Big Data Approach Through Citizen Science Annotation

Purpose Recording young children's vocalizations through wearables is a promising method to assess language development. However, accurately and rapidly annotating these files remains challenging. Online crowdsourcing with the collaboration of citizen scientists could be a feasible solution. In this article, we assess the extent to which citizen scientists' annotations align with those gathered in the lab for recordings collected from young children. Method Segments identified by Language ENvironment Analysis as produced by the key child were extracted from one daylong recording for each of 20 participants: 10 low-risk control children and 10 children diagnosed with Angelman syndrome, a neurogenetic syndrome characterized by severe language impairments. Speech samples were annotated by trained annotators in the laboratory as well as by citizen scientists on Zooniverse. All annotators assigned one of five labels to each sample: Canonical, Noncanonical, Crying, Laughing, and Junk. This allowed the derivation of two child-level vocalization metrics: the Linguistic Proportion and the Canonical Proportion. Results At the segment level, Zooniverse classifications had moderate precision and recall. More importantly, the Linguistic Proportion and the Canonical Proportion derived from Zooniverse annotations were highly correlated with those derived from laboratory annotations. Conclusions Annotations obtained through a citizen science platform can help us overcome challenges posed by the process of annotating daylong speech recordings. Particularly when used in composites or derived metrics, such annotations can be used to investigate early markers of language delays.

Validation of the Language ENvironment Analysis (LENA) system for Dutch

The validity of the Language ENvironment Analysis (LENA) System was evaluated for Dutch. 216 5-min samples (six samples per age per child) were selected from daylong recordings at 5, 10 and 14 months of age of native Dutch-speaking younger siblings of children with autism spectrum disorder (N = 6) and of typically developing children (N = 6). Two native Dutch-speaking coders counted the amount of adult words (AWC), child vocalisations (CVC) and conversational turns (CT). Consequently, correlations between LENA and human estimates were explored. Correlations were high for AWC at all ages (r = .73 to .81). Regarding CVC, estimates were moderately correlated at 5 months (r = .57) but the correlation decreased at 10 (r = .37) and 14 months (r = .14). Correlations for CT were low at all ages (r = .19 to .28). Lastly, correlations were not influenced by the risk status of the children.

The validity of the Language Environment Analysis system in two neonatal intensive care units

AIM

To evaluate the validity of the Language Environment Analysis (LENA) system's automatic measures in two neonatal intensive care units supporting parent-infant closeness, and in two Finno-Ugric languages: Finnish and Estonian.

METHODS

The sound environment of 70 very preterm infants was recorded for 16 h in the neonatal intensive care units with the LENA system roughly at the gestational age of 32 (+2) weeks. Of these, the recordings of 14 infants (20%, two 5-min samples with a high percentage of speech, totally 140 min) were analysed in detail and in two different ways. Parental closeness diaries were used to document the presence of the parents. Agreements between LENA system and human coder estimates were analysed.

RESULTS

Findings showed a high variation in agreements. The highest agreements were found in female and adult word counts (r = 0.91 and 0.95). The agreements for child vocalisation count, conversational turns and silence were modest or low (r = -0.03 to 0.64).

CONCLUSION

Our study provides novel information on the validity of the LENA system in the neonatal intensive care unit. Findings show that the LENA system provides valid information on adult words, but LENA estimates for child vocalisations were less valid at this early age.

Accuracy of the Language Environment Analyses (LENATM) system for estimating child and adult speech in laboratory settings

Laboratory observations are a mainstay of language development research, but transcription is costly. We test whether speech recognition technology originally designed for day-long contexts can be usefully applied to this use-case. We compared automated adult word and child vocalization counts from Language Environment Analysis (LENATM) to those of transcribers in 20-minute play sessions with Spanish-speaking dyads (n = 104) at 1;7 and 2;2. For adult words, results indicated moderate associations but large absolute differences. Associations for child vocalizations were weaker with larger absolute discrepancies. LENA has moderate potential to ease the burden of transcription in some research and clinical applications.

Non-nutritive suck and voice onset time: Examining infant oromotor coordination

The variability of a child’s voice onset time (VOT) decreases during development as they learn to coordinate upper vocal tract and laryngeal articulatory gestures. Yet, little is known about the relationship between VOT and other early motor tasks. The aims of this study were to evaluate the relationship between infant vocalization and another early oromotor task, non-nutritive suck (NNS). Twenty-five full-term infants (11 male, 14 female) completed this study. NNS was measured with a customized pacifier at 3 months to evaluate this early reflex. Measures of mean VOT and variability of VOT (measured via coefficient of variation) were collected from 12-month-old infants using a Language Environmental Analysis device. Variability of VOTs at 12 months was significantly related to NNS measures at 3-months. Increased VOT variability was primarily driven by increased NNS intraburst frequency and increased NNS burst duration. There were no relationships between average VOT or range of VOT and NNS measures. Findings from this pilot study indicate a relationship between NNS measures of intraburst frequency and burst duration and VOT variability. Infants with increased NNS intraburst frequency and NNS burst duration had increased VOT variability, suggesting a relationship between the development of VOT and NNS in the first year of life. Future work is needed to continue to examine the relationship between these early oromotor actions and to evaluate how this may impact later speech development.

Variations in the Home Language Environment and Early Language Development in Rural China

The home language environment is critical to early language development and subsequent skills. However, few studies have quantitatively measured the home language environment in low-income, developing settings. This study explores variations in the home language environment and child language skills among households in poor rural villages in northwestern China. Audio recordings were collected for 38 children aged 20-28 months and analyzed using Language Environment Analysis (LENA) software; language skills were measured using the MacArthur-Bates Mandarin Communicative Developmental Inventories expressive vocabulary scale. The results revealed large variability in both child language skills and home language environment measures (adult words, conversational turns, and child vocalizations) with 5- to 6-fold differences between the highest and lowest scores. Despite variation, however, the average number of adult words and conversational turns were lower than found among urban Chinese children. Correlation analyses did not identify significant correlations between demographic characteristics and the home language environment. However, the results do indicate significant correlations between the home language environment and child language skills, with conversational turns showing the strongest correlation. The results point to a need for further research on language engagement and ways to increase parent-child interactions to improve early language development among young children in rural China.

Randomised control language intervention for infants of adolescent mothers

AIM

Create a Language ENvironment Analysis (LENA)-based intervention to increase adolescent and infant speech and improve 12-month language outcomes.

METHODS

Randomised control trial of adolescent (15-19 years) mother-infant pairs comparing language-motor (intervention) and motor (control) groups. Intervention included reviewing language-motor curriculums, formative feedback on 4 LENA recordings (baseline, post-curriculum, 4 and 12-months) and 16-weekly language-motor texts. Controls reviewed a motor curriculum, summative feedback of four recordings after study completion and 4-monthly motor texts. Primary outcome was 12-month MacArthur scores. Secondary outcomes were LENA counts and social impacts to language outcomes.

RESULTS

A total of 108 infants were randomised. Groups had similar baseline characteristics and LENA counts. Both groups had low maternal Peabody Picture Vocabulary age-equivalents (14.2 years). On post-curriculum recording, intervention infants had higher vocalisations (188 vs 109, P = .02) and conversations (49 vs 30, P = .005) than controls. Group 4-month and 12-month LENA counts and 12-month MacArthur scores were similar. In regression analyses, more people in the home and cohabiting with the infant's father were associated with higher MacArthur scores.

CONCLUSIONS

Linguistic feedback and a simple curriculum resulted in short-term increased vocalisations and conversational turns for infants of adolescent mothers that were not sustained over time. Household characteristics provided protective effects on outcomes.

Head and body structure infants' visual experiences during mobile, naturalistic play

Infants’ visual experiences are important for learning, and may depend on how information is structured in the visual field. This study examined how objects are distributed in 12-month-old infants’ field of view in a mobile play setting. Infants wore a mobile eye tracker that recorded their field of view and eye movements while they freely played with toys and a caregiver. We measured how centered and spread object locations were in infants’ field of view, and investigated how infant posture, object looking, and object distance affected the centering and spread. We found that far toys were less centered in infants’ field of view while infants were prone compared to when sitting or upright. Overall, toys became more centered in view and less spread in location when infants were looking at toys regardless of posture and toy distance. In sum, this study showed that infants’ visual experiences are shaped by the physical relation between infants’ bodies and the locations of objects in the world. However, infants are able to compensate for postural and environmental constraints by actively moving their head and eyes when choosing to look at an object.

The Distributed L1 and L2 Language-Learning Environments of Dual Language Learners Across Home and School Settings

Purpose This study explored the language-learning environments of typically developing dual language learners (DLLs) who learned Cantonese (first language [L1]) at home and English (second language [L2]) in preschool settings through direct and indirect measures. Method Nine typically developing Cantonese-English DLLs participated in this study. Participants' daylong activities were audio-recorded using the digital language processor of the Language ENvironment Analysis system. A manual coding scheme was developed to examine the audio recordings with the focus of the amount of L1 and L2 used by participants, adults, and their peers across home and school settings. In addition, participants' language use was indirectly examined using parent questionnaires, teacher reports, and classroom observations. Results The results of the audio recordings showed that Cantonese was the primary language used at home, and both Cantonese and English were used in school settings, consistent with the parent and teacher reports. Correlation analyses revealed that the amount of L1 used by the participants was associated with the L1 used by their peers: and the amount of L2 used by participants was positively related to the L2 used by adults at home. Conclusions The findings illustrate how parent/teacher reports and daylong audio recordings could complement each other in the investigation of DLLs' language-learning environments.

A meta-analysis of the predictability of LENA™ automated measures for child language development

Early language environment plays a critical role in child language development. The Language ENvironment Analysis (LENA™) system allows researchers and clinicians to collect daylong recordings and obtain automated measures to characterize a child's language environment. This meta-analysis evaluates the predictability of LENA's automated measures for language skills in young children. We systematically searched reports for associations between LENA's automated measures, specifically, adult word count (AWC), conversational turn count (CTC), and child vocalization count (CVC), and language skills in children younger than 48 months. Using robust variance estimation, we calculated weighted mean effect sizes and conducted moderator analyses exploring the factors that might affect this relationship. The results revealed an overall medium effect size for the correlation between LENA's automated measures and language skills. This relationship was largely consistent regardless of child developmental status, publication status, language assessment modality and method, or the age at which the LENA recording was taken; however, the effect was weakly moderated by the gap between LENA recordings and language measures taken. Among the three measures, there were medium associations between CTC and CVC and language, whereas there was a small-to-medium association between AWC and language. These findings extend beyond validation work conducted by the LENA Research Foundation and suggest certain predictive strength of LENA's automated measures for child language. We discussed possible mechanisms underlying the observed associations, as well as the theoretical, methodological, and clinical implications of these findings.

Female caregivers talk more to 18-56-months-old children with and without hearing impairment than male caregivers measured with LENA™ - A cross-sectional pilot study

The objective of the current study was to investigate possible differences in word count use per day (number of adult words) by caregivers of different gender, in a sample of Norwegian children (N = 17) with hearing impairment (HI) (n = 8) and normal hearing (NH) (n = 9), aged 18-56 months. The current study had a cross-sectional, descriptive study design. One all-day recording with the LENA technology was conducted to measure adult word use in the home environment (Md length: 12.46 h, 9.13-16 h). Female caregivers used a significantly higher amount of words than male caregivers close to the children, regardless of their hearing status, HI: p = .01, NH: p = .01. All children were exposed to a higher number of adult words from female caregivers. There is a need to conduct more and further research about possible caregiver differences, and investigate not only the quantity of word use, but also the qualitative interaction patterns between caregivers of different gender and young children with HI, and in relation to early intervention actions.

A step-by-step guide to collecting and analyzing long-format speech environment (LFSE) recordings

Recent years have seen rapid technological development of devices that can record communicative behavior as participants go about daily life. This paper is intended as an end-to-end methodological guidebook for potential users of these technologies, including researchers who want to study children’s or adults’ communicative behavior in everyday contexts. We explain how long-format speech environment (LFSE) recordings provide a unique view on language use and how they can be used to complement other measures at the individual and group level. We aim to help potential users of these technologies make informed decisions regarding research design, hardware, software, and archiving. We also provide information regarding ethics and implementation, issues that are difficult to navigate for those new to this technology, and on which little or no resources are available. This guidebook offers a concise summary of information for new users and points to sources of more detailed information for more advanced users. Links to discussion groups and community-augmented databases are also provided to help readers stay up-to-date on the latest developments.

A Cross-cultural Mixed Methods Investigation of Language Socialization Practices

This embedded mixed methods study explores how cultural differences in language socialization practices influence parent-child verbal interactions. The Language ENvironment Analysis (LENA) System audio recorded families of children who are and are not deaf and hard of hearing in Canada and Vietnam. Software automatically calculated an average conversational turn count. Canadian families participated in more turns than Vietnamese families regardless of hearing status. Interviews with the children's caregivers provided context for these results. Within Vietnamese families, the language socialization practice "Intelligence" results in reduced opportunities for turn-taking, while the Canadian focus on creating personal "Identity" encouraged them. "Intelligence" encompasses Vietnamese participants' desire to ensure their children are learning and "Identity" expresses the Canadian participants' appeal to encourage individuality in their children. The findings suggest directions for the adaptation of intervention. It is the first known study to incorporate LENA results into a mixed methods design.

Socioeconomic Disparities in Language Input Are Associated With Children's Language-Related Brain Structure and Reading Skills

The mechanisms underlying socioeconomic disparities in children's reading skills are not well understood. This study examined associations among socioeconomic background, home linguistic input, brain structure, and reading skills in 5-to-9-year-old children (N = 94). Naturalistic home audio recordings and high-resolution, T1-weighted MRI scans were acquired. Children who experienced more adult-child conversational turns or adult words had greater left perisylvian cortical surface area. Language input mediated the association between parental education and left perisylvian cortical surface area. Language input was indirectly associated with children's reading skills via left perisylvian surface area. Left perisylvian surface area mediated the association between parental education and children's reading skills. Language experience may thus partially explain socioeconomic disparities in language-supporting brain structure and in turn reading skills.