Semantic Learning Deficits in School Age Children With Developmental Language Disorder
Semantic Learning Deficits in School Age Children With Developmental Language Disorder
School age children with developmental language disorder (DLD) have known semantic learning deficits but what is less well understood is why semantic learning is difficult for these children. This project will combine behavioral and brain methods to investigate the cognitive and linguistic processes underlying semantic learning in children with DLD compared to typically developing peers. The outcomes will have implications for semantic learning intervention approaches in DLD.
This project will elucidate deficits in learning semantic information in developmental language disorder (DLD, formerly referred to as specific language impairment) by combining behavioral and neural measures to examine differences in the semantic learning process between school-age children with and without DLD. Vocabulary knowledge, particularly semantic knowledge, has a critical influence on reading comprehension and academic success. Despite the strong association between vocabulary knowledge and academic success, vocabulary is an under-recognized area of deficit in school-age children with DLD. Younger children with DLD have well-established deficits in vocabulary and word learning and weaknesses in semantic knowledge. Additionally, the rate of vocabulary growth in children with DLD decreases compared to typically developing peers around age 10 and semantic representations of known vocabulary items are sparse. Even with this knowledge, the field's ability to make progress toward improved semantic learning in school age DLD is hindered by the lack of basic information on the underlying nature of the semantic learning deficits in this population. This project establishes how and why semantic learning differs between school-age children with and without DLD, providing a much-needed theoretical foundation for clinical research.
Storkel, expanding on an adult word learning model by Leach and Samuel, provides a clearly testable account of word learning that has been used with children with DLD. This account involves three processes: 1) triggering, in which a new lexical encounter is compared to existing lexical representations, 2) configuration, which adds information to the expanding lexical representation, and 3) engagement, which examines how the new lexical representation behaves dynamically with existing representations. The configuration process is arguably the most critical for semantic development. Successful configuration requires the simultaneous engagement of cognitive and linguistic processes, such as attention, inhibition, working memory, and semantic and syntactic processing. While it is widely accepted that configuration is the most affected word learning process in DLD, what is unknown is what underlies deficits in configuration and whether these deficits vary across the DLD profile. These questions are further compounded by difficulty measuring configuration and associated processes, given that they are largely internal, and therefore invisible. Electroencephalography (EEG) addresses this invisibility problem by allowing for a real-time examination of unconscious levels of semantic learning and cognitive and linguistic processes. A combined EEG-behavioral methods approach can illustrate how children with DLD are approaching configuration in terms of the relative contribution of these processes. The central hypothesis of this research is that children with DLD engage cognitive and linguistic processes at different points during configuration compared to their typical peers, resulting in poorer semantic learning outcomes.
To test the central hypothesis, the investigators will record behavioral and EEG data from 10-12 year old children with DLD and typical-language peers as they complete a semantic learning task. This age aligns with the point where vocabulary growth rates in DLD further diverge from typical peers [6]. In the semantic learning task, children listen to sets of three sentences that all end with the same nonword: half of the sentence triplets support learning meaning of the nonword, half do not. The investigators will analyze EEG data for event-related potentials (ERPs) as well as changes in neural oscillations (time frequency analysis). The investigators will combine EEG and behavioral measures to examine the following aims:
Aim 1. To investigate the cognitive and linguistic processes underlying configuration in children with DLD and typical language (TL) peers. This aim will include data from the semantic learning task. Based on the assessment of behavioral outcomes, the investigators predict that the TL group will be more accurate in semantic learning than the DLD group. ERP analyses will focus on the N400 component, associated with semantic processing. Time frequency analysis will focus on changes in the theta (4-8 Hz) and alpha (8-12 Hz) frequency bands, typically associated with lexical retrieval and attention/inhibition, respectively. For both neural measures, the investigators predict engagement of the same components (N400, theta, alpha) across groups but different patterns of change in those components during configuration between groups.
Aim 2. To investigate individual differences in configuration in children with DLD and TL peers. This aim will include data from the semantic learning task and a behavioral assessment battery. Assessment of behavioral data will focus the types of errors children make during semantic learning. The investigators expect that children with DLD will provide incorrect meanings for the nonword that best fit with the first sentence in the triplet and that TL children will provide incorrect meanings that best fit with the last sentence. The investigators will also examine individual differences related to semantic learning outcomes and fine-grained differences in N400 learning effects across groups. Here, the investigators expect that individual differences in general language ability and semantic knowledge, measured via the behavioral assessment battery, will be most predictive of both behavioral semantic learning and N400 change during learning.
Inclusion Criteria:
Exclusion Criteria: