Rates of return to human capital investment. With kind permission from Prof. James J. Heckman (Heckman 2008)
12.1.2 Requirements
Instruments for assessing language skills may be either language screening, which determines a categorial aspect (ill vs. healthy or normal developmental language status vs. abnormal developmental language status), or language tests, which assess a dimensional aspect (degree of language proficiency). It is a prerequisite for early identification of DDSL that there are instruments available that permit a reliable conclusion concerning a child’s language level (evidence-based diagnostics). The quality of a screening or a diagnostic instrument has to be shown by its fulfilment of the usual test-constructive quality criteria.
These primary quality criteria of objectivity (standardised procedure across examiners, absence of examiner bias), reliability (grade of accuracy of a measurement with respect to repeated measurements) and validity (grade of accuracy of the test result with respect to the parameter that it is supposed to measure) are essential preconditions for determining reliably a child’s level of language development. Most important during the construction or evaluation of a screening or diagnostic test is the calculation of its concurrent validity by determining its concordance with another method (reference test, gold standard, external criterion). Normative data that are regularly updated are obligatory requirements for tests assessing a child’s language skills for comparison with the range of variation among its peers. As secondary quality criteria, economic aspects and feasibility are especially important. A summary of psychometric criteria and design features for the assessment of instruments for early identification of linguistic skills in the German language is described in Rosenfeld and Kiese-Himmel (2011) and in Kiese-Himmel and Rosenfeld (2012). Section 11.3 provides tables of English and German language screening methods and tests, together with the quality criteria they fulfil, as well as an instruction on how to interpret the results of language testing and screening.
Because the application of each screening/test instrument for early language evaluation also results in a categorical decision (at risk of being language impaired—typically developing—borderline or unclear, control needed), cut-off criteria for the test scores need to be determined that enable an examiner to make this decision. Furthermore, information about the predictive validity, which demonstrates the extent to which a test score predicts scores from some criterion measurement that is collected at a later time, is desirable.
The most important criteria of diagnostic accuracy are sensitivity, specificity and positive or negative predictive values (Klee 2008). These criteria are not fixed parameters but depend on the population investigated and case definitions both for the analysed and reference tests. By convention and for practical reasons, sensitivity and specificity values of about 80% for each have been proposed as acceptable for screening for DDSL (Plante and Vance 1994). Higher sensitivity values are desirable but are hardly obtainable owing to the high variability of language development, the lack of objective tests and construction difficulties of language screening. For the evaluation of diagnostic accuracy of language tests, a generally accepted definition of SDDSL that determines the cut-off discrepancy between a child’s achieved language status and the average language status of typically developed peers is lacking to date. Therefore, an SDDSL definition proposed for the ICD-11, the upcoming International Classification of Diseases by the World Health Organization, contains the following sentence:
The individual’s ability to understand, produce or use language is markedly below what would be expected given the individual’s age and level of intellectual functioning (by convention and according to prevalence rates about 1.5 to 2 standard deviations below the age-related normative values in standardised and normed speech-language tests).
Neumann (2017)
12.1.3 Risks
In addition to the expected benefit of early identification of language impairment, its potential risks have also to be considered. Possible downsides include the concern of parents, the wasted resources incurred by erroneous classification, stigmatisation by the social environment and a disordered familial interaction.
Currently the instruments used for early identification of language impairment often show relatively high rates of false-positive and false-negative results. These cause children to be classified as potentially language impaired although they are not (false positive). Through this, on the one hand, parents may be made to feel insecure, and on the other, financial and time resources could be wasted on an unnecessary intervention. Even more crucial, parents of children falsely classified as negative are misleadingly appeased and would have probably sought intervention for their child much earlier without screening. Hence, early screening for language risks appears reasonable provided that it fulfils high-quality standards and that appropriate diagnostic and intervention methods are available so that the benefit outweighs potential risks.
12.1.4 Methods
Generally, two basic approaches can be distinguished for the early identification of language abnormalities or DDSL: (1) direct evaluation of a child’s linguistic skills during examination and (2) indirect evaluation of language development by means of parent questionnaires.
For direct evaluation of a child, methods such as the analysis of spontaneous speech, informal (non-standardised) techniques and standardised, normed language tests are used (see Sect. 11.3). It is often discussed that non-structured, informal and non-standardised methods approximate more a non-artificial communicative situation and thus describe a child’s linguistic skills best. On the other hand, this approach is limited by the uniqueness of the investigation, dependence on the specific test situation and by the subjectivity of the examiner. Quality criteria such as objectiveness and reliability are usually poor, and standardised procedures are difficult to apply. On the contrary, standardised language tests are quite often time-consuming and costly in resource use and normally require specific linguistically or psychologically trained personnel.
Parent questionnaires, in particular those that focus on the assessment of the vocabulary of a child, have been proven as both reliable and economic diagnostic instruments for the early identification of impaired language development. Parents, as ‘experts’ who know their child in his or her daily environment, in various situations and over a long period, estimate the linguistic level of their child by standardised questionnaires that do not require special knowledge. Short instruments with a careful choice of items in particular reduce the dread of filling in lengthy questionnaires and provide comparable reliability to longer ones. Examples for suitable questionnaires are the SBE-2-KT (German: von Suchodoletz et al. 2009), the SBE-3-KT (German: von Suchodoletz and Sachse 2009), the PLS-5 (Zimmerman et al. 2011a) and the PLS-5 ST (Zimmerman et al. 2011b). The SBE-2-KT has been translated into 33 languages and is freely accessible on the Internet. For more examples see Sect. 11.3.
12.1.5 Implications
Both the diagnostic accuracy and the prognostic power (prognostic validity) are decisive for early identification methods of developmental language abnormalities or disorders. Only if these are given are early identification procedures useful.
Until the end of the first year of life, early identification of developmental language disorders is quite unreliable and causes more uncertainty for parents that would trigger intervention. Hence, this approach cannot be recommended. For the end of second year of life, there are valid and reliable instruments available, especially parent questionnaires that have been proven to be of value in practice. Even if their prognostic validity is limited in identifying children who will be language impaired at a later age, they detect quite reliably children at risk of developing later developmental disorders of speech and language. As a consequence, these children can be offered an early intervention. For language screening from the end of the third year of life on, there are a few valid instruments available that offer the option of a timely intervention, as shown in Sect. 11.3.
12.2 Stimulation of Child Language Development
12.2.1 Introduction
Children acquire language inherently by a strong genetically determined motivation, given that their biological conditions and social environment are not extremely adverse (child illness or neglect, absence of social bonds). The extent and quality to which they develop language, however, depend both on environmental input—from parent-child interaction, early literacy, a language-stimulating familial and social surrounding—and genome-environment interplay. The parental input style is correlated with the child language development in several respects, which seem to reflect the influence of the input style on a child’s language proficiency. However, as, for example, shown in a recent large twin study, there are also child-to-parent effects that make an interpretation based on parent-to-child effects alone inadequate. Moreover, both parental language input factors and child language have been shown to be moderately heritable. The fact that parent and child share genes produces in itself a correlation between parent-language input style and child-language acquisition. Thus, social influences on a child’s language acquisition may not be overestimated, considering the effects of shared gene variants on parent and child, the partially inherited parental input style and child-to-parent effects. Nevertheless, parental language stimulation is important, and interventions can be effective. This section deals with evidence-based, culturally appropriate interventions that support child language development, parenting skills and language-related child-parent (caregiver) interaction.
12.2.2 Biological and Environmental Stimulation of Language Development
Language acquisition of children is driven by genome-environment interplay. Genetic factors are the motor of language development. Children have a strong inherent motivation to develop oral language, most probably caused by their genetic endowment and an evolutionary ‘necessity’ for functioning in a human social context (Pinker 2002; Berwick et al. 2013). Apart from a frequently uttered public opinion, they need not be taught explicitly in language but develop it inherently by themselves, given that some physiological preconditions are fulfilled, such as the absence of severely impaired hearing or cognitive abilities, and that they grow up in a social environment that is neither neglecting nor language-suppressing (e.g. lacking social bonds). However, the extent and level to which children develop language depend, among other causes, on the input they receive, in particular at early ages when brain development is very dynamic. During early child development, the home environment, in particular the parent-child interactions, establishes a major component of the environmental factors. Parents play an effective role with regard to the rate and quality of a child’s language acquisition. A language-stimulating environment entails talking empathetically to a child, sufficiently frequently and with an appropriate quality, responding to the child’s utterances, repeating and expanding them and making talk time as enjoyable as possible. Later on, around the age of 3 years in community settings, peers increasingly overtake parts of the language-directing influences, together with caregivers and other adults or adolescents. Language proficiency, as well as school readiness and educational success, increases with the amount of early literacy a child experiences, i.e. the extent to which dealing with written and spoken language, such as the presence and use of scripts, books or narratives, plays a role in the family or social environment of a child (e.g. Navsaria and Sanders 2015).
Much research has been performed on the characteristics of speech that is directed to young children or produced in interaction with them during early stages of language acquisition (e.g. Hoff 2006; Rowe 2012). To what extent this kind of child-directed speech predicts a child’s language development is of great interest. In particular, the quantity of input is related to the rate of language development but with wide variations with respect to the vocabulary of children. Additionally, qualitative factors have been shown to predict child language proficiency, usually with weak-to-moderate effect sizes (Dale et al. 2015). Qualitative factors concern all subdomains of language, i.e. phonology, lexicon, syntax, semantics and pragmatics, and comprise, among others, vocabulary diversity, age-appropriate mean length of utterance, repetition, exaggerated prosody, promotion of joint attention, proportion of conversation-eliciting speech as opposed to behaviour regulation, semantic contingency, decontextualized language use such as narratives and ‘grammatical tutorials’ such as sentence recasts and expansions (Dale et al. 2015; Hoff 2006; Rowe 2012).
The influence of specific aspects of child-directed speech differs with age and the developmental status of a child. During early child development, promotion of joint attention and use of exaggerated intonation play an important role, followed by vocabulary diversity and use of sentence recasts at later ages. The quantity of parental input has been shown to be most important during the second year of life, while diversity of parental vocabulary is more important in the third year, and the use of decontextualized language such as narratives and explanations is most important in the fourth year (Dale et al. 2015; Rowe 2012).
The majority of child-directed speech domains that facilitate a child’s language development are intercorrelated. Moreover, they correlate with the parents’ socio-economic status—corresponding to their educational level—that in itself encompasses a variety of parental behaviour. In addition, a variety of more proximate environmental factors contribute significantly to a child’s language acquisition and are reflected in specific features of child-directed speech (e.g. Rowe 2012). They include, among others, prenatal care and nutrition, exposure to toxins and disease, caregiver education and mental health, type and quality of childcare and multilingual versus monolingual context (Dale et al. 2015).
A recent large twin study including more than 8000 twin pairs showed that both language stimulation by the parents and subdomains of the child language itself were moderately heritable (Dale et al. 2015). Moreover, a child-to-parent effect also influences the way parents talk to their child. Hence, a substantial portion of the parental language input to child language represents the effect of shared genes on both parent and child.
Sometimes parents need to be made aware that it is important for their child to be exposed to a stimulating language environment in order to enhance his or her language skills and motivation to speak. This becomes particularly important in the case of developmental language delays (late talkers). For example, short and highly structured parent-based language intervention group programmes have been reported in a randomised controlled trial to be effective in overcoming specific expressive language delays (Buschmann et al. 2009).
In any population, a considerable number of children are at risk of experiencing developmental language abnormalities. Among these abnormalities, developmental disorders of speech and language (DDSL; synonym: language impairment, LI) have to be discerned from sociogenically caused language abnormalities, due to a socially weak environment with poor language stimulation or when multilingual children face difficulties in acquiring the common language. All these groups of children would benefit from more and high-quality language input. However, for developmental speech-language disorders, a sole improvement of input is not sufficient; speech and language treatment together with an optimised language input is required (de Langen-Müller et al. 2012).
12.2.3 Risk Factors for Delayed Language Development
Family history of language impairment
Male sex
Perinatal risk factors, such as prematurity (Nelson et al. 2006)
Other risk factors are reported with less consistency, such as childhood illnesses (Brookhouser et al. 1979; Singer et al. 2001), being born late in the family birth order (Tomblin et al. 1991), family size (Choudhury and Benasich 2006), older parents (Choudhury and Benasich 2006) or a younger mother (Tomblin et al. 1997) at birth, parental psychiatric disorders (Weindrich et al. 2000), low socio-economic status of the family or being of a minority race (Singer et al. 2001).
Recurrent otitis media and related hearing loss has been suspected as a risk factor for developmental speech-language disorders (e.g. Roberts et al. 2002). However, the results of meta-analyses do not reveal a clear indication for such a risk (Casby 2001; Roberts et al. 2004). Nevertheless, there are indications for a negative impact of even mild conductive hearing loss on phonological, lexical, grammatical and pragmatic development, at least at younger ages (Schönweiler 2002). Typically, however, children seem to make up for these language affections at school age, and the home environment of a child seems to be a stronger predictor of language skills than recurrent otitis media (Roberts et al. 2002). Thus, the inclusion of more complex parameters, such as assessment of speech-in-noise and binaural hearing, and appropriate, ‘softer’ outcomes is recommended to be considered in further studies.
12.2.4 Other Potentially Adverse Factors That May Influence Child Language Development
Common beliefs of parents and the public concern factors of a contemporary lifestyle or behaviour as hindering the language development of children, such as background noise in homes, nurseries or other social contexts, long TV exposure, busy parents, reflecting workload and habits, little talking among family members, lack of shared familial meals or lack of parental interaction with their children, but only some of these beliefs are supported by high-quality studies. Other studies point to potential risk factors, but do not provide evidence for a causal role in language acquisition. For example, several electrophysiological experiments have demonstrated that background noise distracts children and may hamper their speech processing (e.g. White-Schwoch et al. 2015). However, the impact of such noise on language acquisition is supported only by some studies (e.g. Wright et al. 1997) not by others (e.g. Alston and James-Roberts 2005).
A considerable proportion of children have TV or video exposures for more than 2 h per day (Chandra et al. 2016; Zimmerman and Christakis 2005). It has been shown that extensive TV watching in childhood and adolescence is associated with adverse health indicators such as obesity, poor fitness, sleep disturbances, musculoskeletal disorders, smoking and raised cholesterol (Chandra et al. 2016; Zimmerman and Christakis 2005). Television viewing before the age of 3 years has also some moderate but significant negative effects on the cognitive development of children (Hancox et al. 2004), makes both children and parents less attentive and shortens toy play time of young children (Schmidt et al. 2008). Nevertheless, there is no strong evidence that extensive TV watching has a negative impact on language development. A few studies have found such a negative impact (Chonchaiya and Pruksananonda 2008, Tomopoulos et al. 2010; Zimmerman et al. 2007), but others have not (e.g. Ruangdaraganon et al. 2009). The results of Zimmerman et al. (2007) were not confirmed by a reanalysis (Ferguson and Donnellan 2014), and scientific debates on this topic continue. Positive or negative effects of television viewing also differ among distinct sociocultural populations and are content and screen time dependent. Indirect effects have to be expected, in that long screen times withhold children from more creative or energy expending activities (Hancox et al. 2004).
12.2.5 Language-Stimulating Family Interventions
Family intervention strategies usually follow one of two approaches that can be used by parents to stimulate their children’s language skills—a child-centred approach and a hybrid approach.
There is no conclusive evidence to guide selection of the most effective approach for children with varying types of delays and disabilities (McCauley and Fey 2006).
12.2.5.1 Child-Centred Approaches
Child-centred strategies are indirect approaches where the parent follows the child’s lead and adds language to each activity performed by the child. The adult chooses the materials but does not direct the activity. Rather he or she follows the child’s lead, providing follow-up on a child’s utterance with his or her own language that is appropriate to the context and the child’s level of understanding. No specifications are prescribed, as the focus is on general communication (Fey 1986; Wolery and Sainato 1996).
Naturalistic
Example: Wait for the child to show interest, get down to the child’s linguistic level and initiate communication.
Interaction-promoting strategies
Example: Encourage the child to take turns in a conversation. Ask questions and wait for a response.
Language-modelling strategies
Example: Label objects and actions, expand utterances and extend topics.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
