Abstract
Speech sound disorder (SSD) is a common communication difficulty affecting approximately 9% of children between ages 3 and 17 years (National Institute on Deafness and Other Communication Disorders, 2016). Children with SSD exhibit unintelligible speech characterized by omissions, substitutions, and distortions of speech sounds within words that cannot be explained by an anatomical or neurological anomaly (American Psychiatric Association, 2013). Many children with SSD experience lifelong difficulties, including bullying and diminished job opportunities (McCormack et al., 2009), as well as reduced language and literacy abilities (Hayiou-Thomas et al., 2017). SSD includes children with articulation errors (i.e., motor-based phonetic) and children with linguistic errors (i.e., rule-based and phonological). Dodd (2014) proposed five SSD subtypes based on the surface error patterns that predict different profiles of performance on speech processing tasks:
Articulation disorder: A mis-learned motor pattern with errors most common on /s/, /z/, /r/, and /θ/; production is distorted at the single-sound level and within all word positions in a similar way. Breakdown is at the motor execution level; cognitive-linguistic skills are intact.
Phonological delay: Children with phonological delay perform at the lower end of the typical range of speech and language tasks (Hayiou-Thomas et al., 2017). Typical errors are considered delayed when the error pattern is no longer present in at least 10% of children in the same 6-month age-band but is present in more than 10% of younger children. Delayed errors include voicing (from 3;0), stopping (from 3;6), and fronting (from 4;0).
Phonological disorder: Consistent use of unusual, nondevelopmental speech error patterns that are produced by fewer than 10% of children between 2;0 and 7;0 (e.g., initial consonant deletion, all final consonants are nasals, substitution of alveolars with velars). Atypical error patterns suggest phonological development has deviated from the typical trajectory rather than becoming “stuck” at an earlier developmental level. One possible explanation is that children with phonological disorder are unable to derive the rules governing each language’s phonological system (Dodd, 2014).
Inconsistent phonological disorder: Speech productions demonstrate greater than 40% variability based on normative data (Dodd et al., 2002). This variability is attributable to a weak ability to plan the sequence of phonemes rather than impaired knowledge of phonological contrasts and phonotactic constraints (Dodd, 2005, 2014).
Childhood apraxia of speech: This speech disorder involves multiple deficits across phonological and phonetic planning, and motor program implementation (Ozanne, 2005).
A variety of interventions are available for SSDs (Williams et al., 2020). The authors of this article suggest that different types of SSDs may respond best to different types of interventions. Understanding the basis for specific SSD symptom profiles could provide guidance for selection of intervention practices that would yield the greatest success by ensuring therapy is tailored to a specific SSD subgroup (Dodd et al., 2018). This study was designed to investigate factors that contributed to two types of SSDs. The researchers compared phonological working memory and cognitive flexibility performance in preschool children with phonological delay and phonological disorder. Two research questions were asked:
Study
Participants
Following assessment with the Diagnostic Evaluation of Articulation and Phonology (DEAP) (Dodd et al., 2002), children were classified as having either phonological delay or phonological disorder. Children who received a phonological delay diagnosis
presented with delayed speech error patterns (i.e., error patterns used by more than 10% of children in a younger age), and
produced less than four individual words containing atypical speech errors (i.e., errors seen in less than 10% of children at any age according to DEAP norms) on the DEAP phonology subtest.
Children who received a phonological disorder diagnosis
presented with at least one atypical speech error pattern at least 5 times in different lexical items when naming 50 single words on the DEAP phonology, or
when naming 50 words on the DEAP phonology subtest, the child made more than four atypical errors (i.e., 4+ atypical errors patterns made at least one time each)
Thirteen children with phonological delay (between ages 3;6 and 5;3) were matched on gender and age with children with phonological disorder. All children performed within normal limits on receptive language ability. There was no significant difference in mean percentage consonants correct, indicating that the two groups were similar in the severity of their SSD.
Assessments
Each child completed a general cognitive assessment (including visuospatial and auditory discrimination skills) and two core executive function components (cognitive flexibility and phonological working memory) to determine whether executive function performance differences might account for the presence/absence of atypical errors (i.e., the surface-level feature differentiating phonological delay and phonological disorder). The assessment yielded three experimental test scores, two for cognitive flexibility and one for phonological working memory.
Cognitive Flexibility
Woodcock–Johnson III Tests of Cognitive Abilities (WJ-III) Concept Formation (Subtest 5) (Woodcock et al., 2007): This is an inductive reasoning task measuring a child’s ability to derive rules for visually presented stimuli, that is, work out the rule for how items go together along color, shape, size, and number dimensions.
The Flexible Item Selection Task (FIST; Jacques & Zelazo, 2001) subtest measures rule abstraction and trial-by-trial cognitive shift in preschool-age children. The FIST battery consists of three tasks that are always presented in the same order:
○ Item Identification Task (colors, shapes, and sizes); ○ Favorite Items Task (selecting two unrelated items); ○ FIST, which was the raw score used to compare cognitive flexibility (i.e., rule abstraction and cognitive shift). For Task 3 (FIST), the child is shown a page with three items that vary along color, shape, and size dimensions (e.g., a small red teapot, a small red shoe, a large red shoe). Two items match on one dimension (e.g., size), two match on another dimension (e.g., shape), and all three match on the irrelevant dimension (e.g., color). The child is asked to select two pictures that “go together one way” (Selection 1, S1) and then to select two pictures that go together “another way” (Selection 2, S2). On completion of the 15 counterbalanced test items, a cognitive shift accuracy is calculated by establishing the number of correct S2 trials after a correct S1 trial.
Phonological Working Memory Assessment
Items Reversed task: This involved the child listening as the examiner says some words (ranging from 2 to 5 one-syllable, high-frequency nouns) and then pointing to the relevant pictures in backward order after a grid of three to eight pictures is revealed. This nonverbal task was selected to avoid examiner listening errors when testing children with unintelligible speech.
Results
Cognitive Flexibility Performance
On the Woodcock–Johnson, the 13 children with phonological delay had significantly higher concept scores (with a large effect size) than their matched children with phonological disorder.
There was a statistically significant difference with a moderately large effect size in the mean FIST (cognitive shift accuracy) score for children with phonological delay and children with phonological disorder. The results suggest children with phonological disorder have reduced cognitive flexibility (poorer rule abstraction and cognitive shift abilities) compared with children with phonological delay. Twelve out of 13 children with phonological disorder performed more poorly than their matched pair with phonological delay in one or more of the cognitive flexibility measures.
Inspection of WJ-III Concept Formation test scores and FIST shift test scores suggested that the cognitive flexibility performance of children with phonological disorder showed little improvement between 3;6 and 5;0, whereas the same skills in children with phonological delay improved after 4;6.
Phonological Working Memory Performance
There was no statistically significant difference in working memory between children with phonological delay and children with phonological disorder. This suggests phonological working memory performance might not contribute to the presence of atypical errors in children with phonological disorder.
Discussion
Dodd and colleagues (1989) proposed that cognitive flexibility may be associated with accurate speech acquisition by enabling children to abstract regularities in the phonological code and to shift to a new generated rule to override errors. In this study, children with phonological disorder performed significantly more poorly than children with phonological delay on the WJ-III Concept Formation subtest and the FIST (which assesses cognitive flexibility). Both measures required the children to first abstract a rule and then shift between dimensions to abstract a new rule. The poorer performance of children with phonological disorder on both measures may be explained by cognitive flexibility difficulties—they found it difficult to abstract the most relevant information (e.g., color) and then shift and abstract another salient dimension (such as size, shape, or number) once an initial selection had been made. Another possible explanation may have been that children with phonological disorder were unable to hold two abstraction rules in their working memory while completing the tasks. Perhaps it was a combination of both. For this reason, the children were assessed on a phonological working memory measure.
Phonological working memory could support phonological development by allowing a child to hold, compare, and manipulate phonological input (Waring et al., 2017). Phonological working memory is considered a building block of cognitive flexibility (Diamond, 2013). Results from the Items Reversed task revealed no statistically significant difference in the phonological working memory performance of children with phonological delay and phonological disorder, although both groups are known to have poorer phonological working memory ability than children with no speech difficulties (Waring et al., 2017, 2018). These results suggest that phonological working memory performance might not account for the poorer cognitive flexibility performance of children with phonological delay. The authors of this study concluded that children with phonological disorder have a specific cognitive flexibility deficit (involving difficulty abstracting correct rules and then mentally shifting when new information is encountered) compared with children with phonological delay. The child may become “stuck” on a phoneme feature such as friction and then be unable to shift to place and number of sounds aspects to derive a correct new rule or appropriate simplification rule (i.e., cluster reduction).
Findings from this study have implications for researchers and clinicians. Assessment of cognitive flexibility may be useful for confirming phonological disorder and phonological delay diagnoses, understanding the nature of a child’s deficits, and informing individualized treatment programs. From an intervention perspective, if executive function ability and atypical errors are associated, it follows that explicitly targeting rule abstraction/cognitive shift could result in more effective and efficient system-wide phonological reorganization than potentially implicitly influencing executive functions through phonological contrast therapy.