Beyond Words: The Impact of Cognitive Load on Children’s Sentence Comprehension
By Beula Magimairaj, Ph.D.
Children’s struggles with understanding spoken language can be influenced by a variety of language and cognitive factors. For example, when we ask children to point to one of four pictures that matches a sentence they hear, we are not just testing their language skills. We are also testing their cognitive abilities, such as how well they focus on and distinguish the characteristics of the target picture vs. the foils, in addition to remembering the details of the spoken sentence.
Clinicians working with children who have comprehension difficulties must be prepared to look at both language skills and cognitive strengths or weaknesses to better identify the root causes of comprehension issues and to design effective interventions.
Working Memory and Long-Term Memory
I briefly outline here a study we conducted to examine the contribution of working memory (WM) and long-term memory (LTM) to children’s understanding of spoken sentences. WM helps us actively remember and manipulate information, while the knowledge we acquire over time is stored as LTM.
Our goal was to see how WM and LTM support comprehension when children are given two sets of sentences that have the same syntactic (grammatical) structure but different demands on memory and attention resulting from the nature of the picture foils. This paradigm tested whether children struggled with the complexity of processing interfering contrastive adjectival information (in foils) with grammaticality held constant vs when the adjectival information was superfluous (i.e., not obligatory for picture selection).
The paradigm was developed in a 2013 study by Laurence B. Leonard and colleagues, which examined sentence comprehension in preschoolers with and without developmental language disorder (DLD). Their results suggested that compared with age-matched typically developing controls, children with DLD struggled more in the high-demand condition with contrastive adjectives relative to sentences with superfluous adjectives, when sentence grammar was held constant.
In the current study, we expanded this approach to include 122 elementary school-age children with a range of cognitive abilities and clinically typical hearing. In addition, we examined the relationship of WM and LTM to sentence comprehension across two cognitive load levels.
We also created longer, age-appropriate sentences suitable for elementary school-age children to be used for the picture pointing task and measured both comprehension accuracy and speed (response time). The two sets of sentences used had the same syntactic structure but differed in how much obligatory detail children needed to process in the pictures to select the target picture.
Figure 1. Example of a sentence from the low cognitive demand condition (superfluous premodifying adjectives). Target (spoken sentence): “The small gray horse bumps the tall pink deer.”
In the low demand condition (Figure 1 above), two of the picture foils contained an unmentioned subject and an unmentioned object, respectively. The adjectives in the spoken sentence were not essential for picture selection (superfluous information) because all horses and deer pictured were the same.
In the high demand condition (Figure 2 below), the adjectives were contrastive in the pictures and therefore obligatory for accurate picture selection.
Figure 2. Example of a sentence from the high cognitive demand condition (contrastive premodifying adjectives). Target (spoken sentence): “The little brown cat pokes the happy black goat.”
Questions Examined
We explored three main questions:
1) Impact of cognitive load on accuracy and speed.
Would sentences with contrastive details in foils (high demand) make it harder for children to respond accurately and quickly compared with the sentence condition with superfluous information?
2) Types of errors.
Would children more often choose a picture with incorrect attributes in the high demand condition vs. making more “reversal” errors (switch up character who did the action with the one receiving it) in the low demand condition?
3) Role of memory skills in comprehension.
Would children’s WM or LTM play a significant role in understanding sentences, especially when extra attention was needed to process contrasting details?
We found that children were slower and less accurate with high-demand sentences, as these required more mental effort and attention to detail. As published in the Journal of Speech, Language, and Hearing Research in October 2024, we showed that WM played a key role in processing sentences with higher cognitive load. When contrastive detail in pictures needed to be processed, WM helped children focus and keep track of details, enabling them to select the correct picture.
Interestingly, LTM didn’t significantly influence performance in this task, likely because the sentences were simple. However, WM’s significant contribution highlights the importance of managing and processing information in real time, which is critical in complex language tasks.
These findings support the idea that comprehension challenges can stem from cognitive limitations besides language structure. For educators and clinicians, this suggests that sentence comprehension measures can provide insights into children’s cognitive strengths and areas that need support.
By understanding how details like adjectives affect comprehension, teachers and therapists can tailor materials to each child’s cognitive capacity, promoting better learning and language skills. Data supports the newly developed measure’s potential for assessing cognitive skills that are integral to sentence comprehension.
Similarly, keeping cognitive demands constant and varying grammatical complexity between sentence sets can reveal children’s syntactic processing ability. Such measures can be useful to identify and contrast the potential source(s) or predictors of listening comprehension difficulties in school-age children who are suspected to have auditory processing disorder.
A 2015 Emerging Research Grants scientist generously funded by Royal Arch Research Assistance, Beula Magimairaj, Ph.D., CCC-SLP, is a research scientist at Utah State University’s Early Childhood Education and Research Center.