Reading comprehension intervention in populations with developmental language disorder: A scoping review · Lam JHY, Wang J, Pratt AS et al.

Introduction

Developmental language disorder (DLD), formerly characterized as specific language impairment, language learning disorder, or primary language impairment, is a high-incidence neurodevelopmental disorder associated with difficulties in producing and understanding language (Bishop et al., 2017). DLD affects 7–10 % of children worldwide (Norbury et al., 2016; Tomblin et al., 1997), and approximately 25–50 % of children with DLD experience co-occurring reading difficulties, particularly in reading comprehension (Adlof & Hogan, 2018). These challenges are likely attributable to impairments in semantics, morphosyntax, and pragmatics (Castle et al., 2018). Children’s reading comprehension abilities can affect their engagement in learning, academic performance, and future career development (e.g., Dockrell et al., 2011; Nag, 2014). The current scoping review examines features of reading comprehension interventions in children with DLD, including an evaluation of intervention outcomes.

Reading comprehension is an extension of language comprehension that represents an integral component of speech-language pathologists’ (SLPs’) expertise. Literacy, including both reading and writing, falls within the scope of practice for SLPs worldwide (American Speech-Language-Hearing Association [ASHA], 2016; Health & Care Professions Council, 2013; Speech Pathology Australia, 2020; The Hong Kong Institute of Speech Therapists, 2022). The role of SLPs in the domain of literacy involves prevention of written language difficulties, reading and writing assessment, and delivery of interventions targeting difficulties with written language (ASHA, 2001). SLPs often report low confidence and self-efficacy in providing literacy-related services (Yi & Erickson, 2024a, 2024c), which may be attributable to limited training in literacy during graduate education (Yi & Erickson, 2024b). This scoping review fills an important gap in knowledge by summarizing the available evidence on reading comprehension interventions for populations with DLD and identifying potential areas of future research. Therefore, the purpose of the current scoping review is twofold: (1) to summarize the research about reading comprehension interventions in populations with DLD using the Active View of Reading as a framework for organizing the existing literature (AVR; Duke & Cartwright, 2021) and (2) to identify limitations and gaps in the available evidence base.

Theories of reading comprehension

Reading comprehension is a complex and multidimensional construct. There has been a preference towards parsimonious models of reading such as The Simple View of Reading (SVR; Gough & Tunmer, 1986; Hoover & Gough, 1990) which describes variability in reading comprehension across two component skills: decoding (or the act of mapping sounds onto symbols) and language comprehension (or the act of attaching meaning to the decoded words). According to this model, children with DLD, who by definition struggle with oral language, are also likely to struggle in reading comprehension, given their difficulties with language comprehension. However, the high levels of covariability between decoding and language comprehension suggest that other skills may exert influence on the reading process, above and beyond the two components of the SVR (Cartwright et al., 2020; Cutting & Scarborough, 2006; Foorman et al., 2020). Subsequent models of reading comprehension have attempted to account for these additional factors, including internal processes such as self-regulation (e.g., Kim, 2017) and external knowledge, such as familiarity with the text structure and/or genre (e.g., Hogan et al., 2011; Kim, 2017).

The Active View of Reading (AVR; Duke & Cartwright, 2021), as shown in Fig. 1, is an expanded model of reading comprehension that builds upon the SVR. The AVR includes four components that explore the different component skills that affect reading comprehension: active self-regulation, word recognition, bridging processes, and language comprehension. In contrast to other models of reading comprehension, the AVR emphasizes that all of its components are instructionally malleable, thus making it ideal for guiding the development and implementation of reading comprehension interventions.

Given that the AVR represents a relatively recent expansion of the SVR, research exploring its application to diverse populations is limited. With respect to readers of diverse orthographies, we could not identify any empirical studies validating its use across orthographic types. However, there is a substantial body of research showing that the SVR is valid and applicable for beginning readers of transparent (non-English) alphabetic orthographies (see Florit & Cain, 2011, for a meta-analyses; and more recently, Silinskas et al., 2024, for a discussion of the SVR in Lithuanian) as well as opaque logographic orthographies (e.g., deep version of Arabic orthography, Asadi & Ibrahim, 2018; Chinese, Joshi, 2018). At both ends of the orthographic spectrum, significant variance in reading comprehension was explained by decoding and language comprehension, though the relative influence of the two components was different for readers of different orthographic types.

There is emerging evidence validating the AVR for children with neurodevelopmental disorders, such as DLD. A previous meta-analysis showed that interventions targeting the four components of the AVR yielded medium to large effect sizes on reading comprehension in children with reading difficulties (Burns et al., 2023). However, this study did not disentangle the source of children’s difficulties in reading comprehension. Children with DLD represent a distinct subgroup of struggling readers with a unique profile of strengths and challenges, which may, in turn, require a different set of intervention targets. Most notably, children with DLD are at higher risk for reading comprehension difficulties even when decoding skills are adequate (Bishop et al., 2017; Catts et al., 2005). Compared to children with dyslexia or struggling readers, children with DLD often show broader impairments in vocabulary, morphosyntax, and discourse skills that underpin text comprehension (Nation et al., 2004; Snowling et al., 2016). A more recent systematic review demonstrated the applicability of the AVR in explaining reading comprehension ability in children with DLD (Lam et al., 2024). Consistent with the AVR, results of the review showed strong predictive relationships between language-based measures (vocabulary, listening comprehension, oral language comprehension, grammatical knowledge) and reading comprehension among children with DLD.

Sources of reading comprehension difficulties in children with DLD

The AVR can potentially guide reading comprehension interventions for children with DLD. The model suggests that active self-regulation is the basis for reading comprehension, as it influences all other components of reading comprehension. Active self-regulation refers to the cognitive skills that help readers focus on what they are reading, including executive function skills (attention, inhibition), meta-cognitive strategy use, and motivation and engagement (Duke & Cartwright, 2021). When applying these skills, children deliberately utilize strategies to construct meaning from the text, and they maintain motivation and engagement to stay on task. Reading comprehension interventions targeting self-regulation skills are important for reading comprehension due to the dynamic nature of how reading comprehension is achieved. Reading comprehension is a goal-oriented process where readers use executive function skills to make inferences, monitor comprehension, and engage in the reading process. A previous meta-analysis showed that interventions targeting active self-regulation can improve reading comprehension in average readers, as compared to interventions focusing solely on word recognition and language comprehension (Burns et al., 2023). Because it is well-documented that children with DLD have lower performance across various aspects of executive function skills, including interference control and working memory (e.g., Blom & Boerma, 2020), it is plausible that they employ lower self-regulation during reading comprehension. Thus, reading comprehension interventions that target self-regulation and strategy use may have the potential to improve reading comprehension skills for children with DLD.

The AVR’s word recognition and language comprehension components are consistent with the Simple View of Reading. Word recognition was originally described in the SVR model as decoding, which includes both decoding via one’s knowledge of sound-symbol correspondence and the automatization of word recognition (Gough & Tunmer, 1986; Hoover & Tunmer, 2020). Children with DLD often have difficulty in phonological processing (Catts et al., 2005; Snowling et al., 2019). These difficulties may impact serial recognition and the production of speech sounds, which can in turn hinder word recognition when reading. Thus, children with DLD may stand to benefit from phoneme- and word-level intervention to help improve their reading comprehension abilities.

Language comprehension refers to a set of skills ranging from language skills, such as syntactic skills to broader cultural and world knowledge (Duke & Cartwright, 2021). Language comprehension is a major area of weakness among children with DLD (Bishop et al., 2017). These may include difficulties in verbal reasoning, syntax, and theory of mind (Montgomery et al., 2018; Nilsson & de Lopez, 2016). Interventions targeting language comprehension in children with DLD, as shown in previous meta-analyses, are operationalized as discourse-level interventions using narrative, listening comprehension, and conversational skills (Cleave et al., 2015; Winters et al., 2022). These extended discourse interventions aim to enhance comprehension of sentences, inference skills, and knowledge of text structure. Interventions targeting language comprehension might also have a generalization effect on reading comprehension outcomes in children with DLD.

The bridging processes component is the final component of the AVR model. It addresses the overlapping skills between word recognition and language comprehension that were not accounted for in the SVR framework (Duke & Cartwright, 2021). Specifically, vocabulary knowledge and morphological awareness are considered bridging processes in the AVR, because both word recognition and language comprehension skills are required. Children with DLD have significant impairments in semantics and morphosyntax (Kan & Windsor, 2010; Krok & Leonard, 2015). As such, interventions targeting vocabulary knowledge in children with DLD are relatively well-studied and have been conducted across a wide range of ages, spanning preschool to adolescence (Cirrin & Gillam, 2008; Lowe et al., 2018). Vocabulary interventions in children with DLD have employed a variety of strategies, including enhancing phonological awareness to improve the letter-sound correspondence (e.g., Gillam et al., 2008), leveraging morphological awareness to improve word definitions and word classes (e.g., Katz & Carlisle, 2009), and embedding vocabulary instruction within highly contextualized books and conversational speech (e.g., Joffe et al., 2019). The diversity of treatment approaches echoes the overlapping and interconnected nature of the bridging processes component. By strengthening semantic networks, these interventions enhance children’s ability to understand the meaning of the text during reading. Considered collectively, the prior literature on DLD suggests that children with language-learning disorders may experience difficulties across all four components of the AVR, which may impact their reading comprehension. However, to date, no study has systematically summarized the features of interventions targeting reading comprehension in this population. Thus, the current study aims to examine intervention studies conducted with children with DLD to evaluate the latest evidence for improving reading comprehension and to identify current gaps in the literature.

Considerations for examining reading comprehension intervention research

The variability among interventions documented in the existing literature presents significant challenges for health practitioners, educators, and policy makers in interpreting the evidence. To address these challenges, an international consortium of experts in speech and language pathology identified key barriers to evaluating intervention research within the field, focusing specifically on dosage, context, and outcomes (Frizelle et al., 2023).

A major challenge in evaluating the evidence for intervention research is dose form and dosage-related information. Dose form is defined as the experimental manipulation of the intervention technique, helping researchers to determine the optimal dosage and establish causal inference between the intervention and the outcome (Frizelle et al., 2023). Given that reading comprehension involves multiple subcomponents (Catts & Kamhi, 2017), the dose form and technique adopted in the intervention may consist of multiple skills addressing multiple aspects of reading comprehension. In this case, it is important for clinicians to identify the specific skills that the intervention is targeting. Furthermore, a quantitative description of dosage is crucial for implementation, given differing strengths of associations between reading comprehension and its component skills. This includes dose (i.e., the number of teaching episodes delivered within a session), dose frequency (i.e., number of doses divided by unit time), and total intervention duration (Warren et al., 2007).

Outcome measures can also impact effect sizes and the statistical significance of interventions. Reading comprehension is commonly measured using both standardized and non-standardized measures. Standardized reading measures are designed to describe a child’s overall reading ability and provide psychometric information, such as internal reliability, test-retest reliability, and validity, as described in test manuals. In contrast, non-standardized measurements are usually designed to measure targeted reading behaviors and are often more proximal to the intervention targets. A prior meta-analysis on reading comprehension interventions identified measurement type (i.e., standardized versus non-standardized tests) as a significant moderator of intervention outcomes, with non-standardized tests demonstrating larger intervention effect sizes than standardized tests (Filderman et al., 2022). In addition, standardized tests select features that differentiate between typical and atypical processing, which might not be comprehensive enough to capture all aspects for tracking intervention outcomes (McCauley & Swisher, 1984). Thus, it is important to review the content and scope of the assessment measuring reading outcomes to examine whether it can effectively capture change over time.

Intervention context is another critical aspect to consider. Home-based and school-based interventions differ in terms of types of available support, student motivation, and implementation challenges. According to self-determination theory (Krapp, 2005), the school environment has higher collaboration between professionals and greater peer support compared to the home environment. School is also a common intervention context for literacy support. On the other hand, intervention in the home context enables parental support throughout the intervention, and family has a strong impact on early child literacy development through parental involvement in reading and enhancing children’s reading motivation (Loera et al., 2011; McElvany & Artelt, 2009).

Research aims

Although there is a recent meta-analysis on reading comprehension interventions using the AVR (Burns et al., 2023), this has yet to be extended to include children with DLD. Because children with DLD have specific difficulties in the subcomponents of reading, such as language skills (language comprehension, bridging processes), word recognition, and active self-regulation, the target skills and the dosage of intervention for this population may differ from the general population. It is thus meaningful to examine the needs of this specific population in reading comprehension. The aims of the current study are: To synthesize the existing literature about reading comprehension interventions in populations with DLD using the AVR framework.To identify potential limitations of existing reading comprehension interventions in populations with DLD.To identify potential research gaps in reading comprehension interventions in populations with DLD.