Abstract
Yanga, H., & Gray, S. (2017). Executive function in preschoolers with primary language impairment. Journal of Speech, Language, and Hearing Research, 60, 379–392.
The term specific language impairment (SLI) has often been used in the research literature to refer to a language delay or disorder in the mastery of language skills that exists in children who have no hearing loss or other developmental delays. The term SLI has become controversial and is used less frequently because research is demonstrating that language impairment is not so specific. These children frequently have difficulties in a number of cognitive functions, including executive functions (EFs). There has not been agreement on the term to use in place SLI. British researchers prefer the term developmental language disorder (DLD); American researchers are more likely to use the term primary language impairment (PLI).
The purpose of this study was to determine whether preschool-age children with PLI showed deficits in EF. A number of studies have shown that school-age children with PLI have EF deficits. Many off those studies used linguistically based EF tasks. EF tasks are known to be affected by the ability to use language for labeling or for internalized private speech. Therefore, such tasks may not be fair tests of EF in children with PLI.
In this study, the authors compared the performance of typically developing (TD) children and children with PLI on linguistic or visual EF tasks involving inhibition, updating, or mental-set shifting. They asked,
Are there significant between-groups differences for preschoolers with TD and PLI as measured by accuracy or response time (RT) on linguistically and visually based inhibition tasks?
Are there significant between-groups differences for preschoolers with TD and PLI as measured by accuracy on linguistically and visually based updating and mental-set shifting tasks?
The Study
Two groups of 4- and 5-year-olds completed the study: 22 with PLI and 30 with TD. Parents completed a questionnaire regarding mother’s education level and answered questions regarding their child’s attention and impulse control from the Strengths and Weaknesses of ADHD Symptoms and Normal Behavior Scale (Swanson et al., 2005). Because there was a significant difference between the two groups on this measure and because poor attentional control can affect performance on EF tasks, the raw score on this questionnaire was included as a covariate in the data analyses. All children met the following inclusionary criteria: (a) normal hearing sensitivity in each ear, (b) no history of neuropsychiatric disorders (e.g., attention-deficit disorder/attention-deficit/hyperactivity disorder, emotional disability, motor disorder) by parent report, and (c) SS of 75 or higher on the nonverbal matrices subtest of the Kaufman Brief Intelligence Test–Second Edition (K-BIT-2; Kaufman & Kaufman, 2004). Children in the TD group scored 95 or higher on the Structured Photographic Expressive Language Test–Preschool Second Edition (SPELT-P2; Dawson, Eyer, & Fonkalsrud, 2005); children with PLI were required to score 87 or lower. This cutoff score was based on the analysis from Greenslade, Plante, and Vance (2009) that yielded 90.6% sensitivity and 100% specificity for the SPELT-P2 in identifying preschoolers with language impairment.
Procedures
Six EF tasks and two baseline RT tasks were administered. The RT tasks were to document children’s motor and voice RTs when there was minimal language or cognitive demand because studies have shown that children with PLI are slower than their peers with TD on motor responses (e.g., Miller, Kail, Leonard, & Tomblin, 2001).
The three EFs of inhibition, updating, and mental-set shifting were assessed using two computerized tasks each, one linguistically based and one visually based. Each task started with recorded instructions followed by a training phase that taught children how to play the game. Children received feedback regarding the accuracy of their responses during training.
Baseline Motor RT Task. Children were to press a key as soon as they saw a soccer ball on the screen. There were 10 trials with an image and 10 trials without an image. The program recorded the accuracy and the RT in milliseconds from the onset of the picture and the key-press response.
Baseline Voice RT Task. This task assessed children’s voice onset times when they said the word cake in response to seeing the image of a cake appear on the screen.
Linguistic Inhibition Tasks. One horizontal row of five animals was presented in the center of the screen at a time. The middle animal was the target, and the surrounding animals were the flankers. There were two types of trials: congruent, in which all five animals were the same (either all cats or dogs), and incongruent, in which the animals surrounding the middle animal were different (i.e., a dog was surrounded by four cats or vice versa). Children were asked to say cat or dog to match the middle animal. There were two dependent variables: (a) conflict effect for accuracy, which refers to the difference in percentage correct between congruent and incongruent conditions. This indices the inhibitory effort that children have to make during the incongruent trials, and (b) conflict effect for RT, which is the mean RT difference for accurate incongruent trials minus the mean RT for correct congruent trials.
The visual inhibition task. A horizontal row of five fish was presented in the center of the screen. The middle fish was the target, and the surrounding fish were the flankers. In the congruent condition, all fish pointed in the same direction, and in the incongruent condition, the fish surrounding the middle fish pointed in the opposite direction of the target fish. Children were told to feed the middle fish by pressing the key that matched the direction it was pointing. The dependent variables were the same as the linguistic inhibition task.
Linguistic Updating task. The task used eight images (car, duck, horse, apple, chair, shoe, flower, and hat) presented one at a time in the center of the screen. Children were asked to help match toys by saying same when the toy was the same as the one on the previous screen (same trial) and saying different if it was different from the one on the previous screen (different trial). The dependent variable was accuracy, calculated as the percentage of correct responses.
Visual updating task. A cartoon butterfly was presented on the screen in one of eight locations. Children were asked to say stayed if the butterfly stayed at the same location as the previous screen or say moved if the butterfly moved to a different location from the previous screen.
Linguistic Shifting Tasks. This measured children’s ability to shift between two sorting rules—the color of the background and the shape of the item (e.g., glasses presented on a red background; scissors on a blue background). Children were asked to sort each test stimulus, presented in the center of the screen, by color or shape. This task was linguistically based because the color of the backgrounds and shape of the items were easily labeled by children (e.g., this one is red so it goes with red). The task included one preshift block and one postshift block. Before each block, children were reminded of rules for the current sorting game and asked to demonstrate knowledge of both rules by correctly answering where the objects should be sorted (e.g., where do the blue ones go in the color game?). During the test phase, children heard auditory instructions that indicated the current sorting dimension (i.e., “This is a blue card. Where does it go?”), which reduced the working memory load of remembering the sorting rule.
Visual shifting task used the same paradigm as the linguistic shifting task; however, the target and stimuli did not have readily associable names. The two target stimuli were different polygon shapes presented on backgrounds of two different patterns (diagonal-lined or tessellated pattern). The response mode was a key press,
Results
Baseline RT. The PLI group RTs were significantly slower than the TD group RTs on the motor task, but there were no significant between-groups differences for RTs on the voice task.
Inhibition tasks. All children in the TD group and 95% of the PLI group passed training for the linguistic inhibition task. All children in both groups passed training for the visual inhibition task. Both groups made more errors and responded more slowly to incongruent than congruent stimuli on both inhibition tasks. The PLI and TD groups did not differ significantly on accuracy or RTs for either the linguistically based or visually based inhibition tasks.
Updating tasks. Overall, 100% of children in the TD group and 68.2% and 81.8% of the PLI group passed training on the linguistic and visual updating tasks, respectively. The proportion of the children who failed the training trials was significantly larger in the PLI than the TD group on both tasks. Thus, the updating tasks were more difficult for the PLI group than the TD group. The PLI group accuracy was significantly lower than the TD group accuracy on both updating tasks.
Shifting tasks. The training pass rates for the linguistic and visual shifting tasks were 100% for both groups. The groups did not differ in their understanding of the sorting rules. However, even when all the children demonstrated their understanding of the sorting rules before each block, many children were not able to sort the stimuli according to the correct rules during the preshift block, especially children in the PLI group. Being able to sort was defined as sorting stimuli correctly above the 50% chance level. Two children with TD (6.7%) and three children with PLI (13.6%) could not sort the stimuli during the linguistic preshift block. Two different children with TD (6.7%) and 10 children with PLI (45.5%) could not sort correctly during the visual preshift block. Thus, a relatively large proportion of the PLI group had difficulty with the visual shifting task, even after they demonstrated sufficient knowledge of the dimensions of the stimuli and the sorting rules.
Discussion
Children with PLI may catch up with their peers on inhibition but still struggle with later developing updating and shifting EFs well into the school years. Preschoolers with PLI showed difficulties on a broad range of EF tasks, even though participants did not have diagnosed attention-deficit/hyperactivity disorder. EF deficits were not limited to the linguistic tasks; the children with PLI exhibited similar difficulty on the visual EF tasks. Findings add to evidence showing that EF deficits in preschoolers with PLI are not limited to the language domain but are more general in nature. Study results showed that preschoolers with PLI exhibited general EF deficits in updating and shifting, but not inhibition, when compared with their peers of the same age with TD. The authors suggest that this could be because EF is on a slower developmental trajectory in children with PLI, and at some point, they will catch up to their peers who are on a faster developmental course. Inhibition may develop earlier than updating and mental-set shifting.
The updating task required that children retrieve the representations for the new item and compare the new representations with the previous one to determine whether they were the same or different. For the linguistic task, children had to replace the representations of the old item (car) with the new one (apple) in working memory so that they could compare it to the next item (e.g., horse). This task required continuous online updating of activated semantic and phonological representations, which clearly was challenging for the PLI group. Children with language impairment are known to have problems retrieving, maintaining, and comparing phonological and semantic representations within the linguistic domain of working memory (Vugs, Hendriks, Cuperus, & Verhoeven, 2014).
Even when the children with PLI understood the sorting task, they did not consistently sort even in the preshift condition (e.g., when they might be asked to sort all items by color). They had even greater difficulty when asked to shift sorting stimuli by one dimension to the other. To shift, for example, from sorting by color to sorting by object, children had to update and maintain the current sorting rule in working memory but also inhibit attention to the prior dimension. The authors of the study had reduced the working memory load in the study by reminding children of the current sorting dimension of the stimulus each trial, and they had decreased the inhibition load by separating the two dimensions into foreground and background instead of integrating them together. Despite that, the children with PLI had marked difficulty inhibiting their attention to one dimension to shift to the other dimension.
Because preschoolers with PLI may have deficits in several EFs, the authors of this study recommend that early intervention focus not only on language development but also on developing strategies that will help young children process and store linguistic and visuospatial information. EFs and working memory may be enhanced via interventions (Wener & Archibald, 2011) and curriculum (Diamond, Barnett, Thomas, & Munro, 2007), and working memory–based interventions may improve language, reading, and working memory skills in children with language impairment (Farquharson & Franzluebbers, 2014).