Abstract
The ability to make inferences is critical for oral and written discourse comprehension. Early oral inferential comprehension of narratives is a predictor of later reading comprehension (Silva & Cain, 2015), and inferring was found to be an important, unique predictor of reading comprehension in 8- to 11-year-old English-speaking children (Cain, Oakhill, & Bryant, 2004). Consequently, it is important to understand the skills contributing to inferring in young children with DLD in order to develop targeted interventions.
Children are able to form causal inferences (e.g., a character’s internal response) from 3 years of age, with more complex inferential comprehension skills (e.g., prediction) developing around 5 to 6 years (Filiatrault-Veilleux, Bouchard, Trudeau, & Desmaris, 2016). One of the main contexts in which young children engage in inferencing is during the shared reading of narratives. Inferential comprehension of narratives involves three types of inferences: causal, evaluative, and informational. Causal inferences, which relate to narrative structure and comprehension, include predictions, linking information within or across texts or with background knowledge, and inferring character feelings, attitudes, and motives. Judgments of morality or convention are required to form evaluative inferences, and informational inferences involve providing unstated information on narrative setting (character, time, place), elaborating on information using background knowledge, or defining words from information provided (Van Kleeck, 2008). Narratives present a naturalistic context in which to assess and support the development of inferential comprehension in children.
The purpose of this study was to identify the language and cognitive skills that significantly contributed to oral inferential comprehension in a group of young children with DLD through administration of a comprehensive range of word, sentence, and discourse-level language and cognitive measures.
The Study
Participants in the study were between were aged between 5 years 2 months and 6 years 2 months. The sample consisted of 60 males (78.9%) and 16 females (21.1%) with diagnosed DLD.
Measures
The following language and cognitive skills were assessed, with inferential comprehension of narrative as the primary outcome.
Inferential comprehension: The Squirrel Story Narrative Comprehension Assessment (NCA). The children watched and listened to the story on the iPad and were then asked five literal and 14 inferential questions. Responses were evaluated using a scoring scale (0, 1, or 2 points for each question). (Note: The pictures and questions can be viewed at this website: https://www.blacksheeppress.co.uk/product/squirrel-story-narrative-comprehension-assessment-nca/)
Teachers completed two checklists for each participant, assessing theory of mind development, and general communication and pragmatic skills: Theory of Mind Inventory (ToMI; Hutchins, Prelock, & Bonazinga, 2010); general language and pragmatics Children’s Communication Checklist–second edition (Bishop, 2003a).
Macrostructure/microstructure: The participants listened to the Squirrel Story again and then retold it while looking through the pictures. The story retells were scored for narrative macrostructure (story structure and story content ratings) and narrative microstructure (level of language used/syntax and vocabulary ratings).
Expressive and receptive vocabulary: Expressive Vocabulary Test–second edition (EVT-2; Williams, 2007); receptive single-word vocabulary Peabody Picture Vocabulary Test–Fourth Edition (PPVT-4; Dunn & Dunn, 2007).
Receptive and expressive grammar: Word Structure subtest of the Clinical Evaluation of Language Fundamentals, Preschool–second edition (CELF-P2; Semel et al., 2006); receptive grammar Test for Reception of Grammar–second edition (TROG-2; Bishop, 2003b).
Executive function: The bear/dragon and grass/snow tasks were administered (Carlson, 2005). For the bear/dragon task, participants watched an iPad video with puppets and was asked to follow simple instructions (e.g., “touch your knee”) given by a “nice” bear puppet and ignore instructions given by a “naughty” dragon puppet. For the grass/snow task, participants were asked to point to a white card when they heard the word “grass” and a green card when they heard the word “snow.”
Working memory and linguistic processing speed (rapid automatic naming).
Nonverbal IQ from the Wechsler Preschool and Primary Scale of Intelligence (Wechsler, 2002).
Results
Narrative macrostructure and microstructure, literal comprehension, vocabulary awareness, working memory/processing speed, and theory of mind were significant predictors of inferential comprehension.
However, only the narrative macrostructure and theory of mind remained significantly associated with inferential comprehension.
Implications
Other studies have shown stronger relationships between microstructure elements (vocabulary, syntax, working memory) and inferential abilities than this study. These contrasting findings may reflect that different profiles of skills are important for inferential comprehension in different populations (i.e., typically developing vs. DLD). Alternatively, the results may indicate that the skills contributing to inferential comprehension in children with DLD may change over time. In particular, this may reflect the gradual development of skills which tend to be particularly weak in children with DLD, such as vocabulary, grammar, and working memory. Children will clearly need to develop these microstructure skills to comprehend narratives, and they do contribute to some degree to inferential abilities, but they are clearly insufficient for inferential reasoning. Study results show that children needed to be able to construct a coherent situation model for the stories if they were to make appropriate inference. To construct situation models, children must have higher order discourse skills—narrative macrostructure and theory of mind skills.
The findings of this study provide speech and language pathologists with information for understanding of the skills that contribute significantly to inferential comprehension of narratives in young children with DLD. The findings should support the development of targeted, evidence-based interventions to support and improve oral inferential comprehension in children with DLD. The data indicate that literal comprehension is distinct from inferential comprehension. In both assessment and intervention, literal and inferential comprehension should be considered as related but independent skills. Narrative macrostructure and theory of mind skills were the greatest predictors of inferential comprehension and hence, they should be intervention targets. Development of microstructure skills could occur within activities aimed at developing macrostructure and theory of mind skills.