Exploring the relations between speech production and word reading in children with dyslexia, speech production deficits, or both

Exploring the relations between speech production and word reading in children with dyslexia, speech production deficits, or both

First Author: Kelly Farquharson -- Florida State University
Additional authors/chairs: 
Tiffany P. Hogan; Kathryn Cabbage; Jennifer Zuk; Jennifer Thomson
Keywords: Dyslexia, speech, Word reading, Nonword Repetition
Abstract / Summary: 

Poorly-specified phonological representations have been implicated for both children with dyslexia (Boada & Pennington, 2006; Elbro et al., 1998; Goswami, 2000; Swan & Goswami, 1997) and children with speech production deficits (Anthony et al., 2011; Edwards, Fourakis, Beckman & Fox, 1999; Rvachew et al., 2003; Sutherland & Gillon, 2005; Sutherland & Gillon, 2007) though these conditions do not always co-occur. In this symposium, four studies are reported in which the common aim is to increase our understanding of the relations between speech production and word reading. We report findings of white matter tract differences in children with early speech production deficits and later word reading difficulties (Zuk et al.). We further considered the influence of various stimuli, such as nonwords (Cabbage et al.; Farquharson et al.; Hogan et al.); mono- versus multi-syllabic words (Hogan et al.); high versus low neighborhood density and list length (Farquharson et al.), on task performance in children who have dyslexia or speech sound disorders.

Symposium Papers: 

The potential role of speech sound production in facilitating reading development among children at risk for reading impairment

First Author/Chair:Jennifer Zuk -- Harvard University
Additional authors/chairs: 
Jade Dunstan; Elizabeth Norton; Xi Yu; Ola Ozernov-Palchik; Yingying Wang; John D.E. Gabrieli; Tiffany P. Hogan; Nadine Gaab

Purpose: Among kindergarteners with behavioral risk for reading impairments (RI), what distinguishes those who subsequently develop typical reading abilities from those who develop RI? At-risk children with subsequently typical reading skills have shown superior speech and language abilities, and compensated children with RI have shown enhanced right-hemispheric neural engagement; yet no study has examined how these behavioral and neural factors in kindergarten relate to subsequent reading outcomes.

Methods: 63 Kindergarteners were selected from a longitudinal investigation and characterized as those with behavioral risk who subsequently developed typical reading abilities (RiskTYP), at-risk who developed RI (RiskRI), and children without risk who developed typical reading abilities (NoRiskTYP). A comprehensive battery of standardized assessments and parent questionnaires were acquired in kindergarten and second grade. Neuroimaging was acquired in kindergarten, and reading-related white matter tracts were estimated bilaterally due to left-hemispheric atypicalities associated with RI and implications of compensatory mechanisms on the right.

Results: Speech production accuracy and SES were found to be significantly higher in RiskTYP compared to RiskRI. Differences in vocabulary emerged from kindergarten to second grade. Partial correlations based on risk status examined relationships between white matter in kindergarten and subsequent word reading measures controlling for age, IQ, gender, and SES. There were no significant correlations with NoRiskTYP, whereas RiskTYP and RiskRI showed positive correlations between word-level reading measures and properties of right-hemispheric white matter tracts.

Conclusions: RiskTYP showed significantly higher speech accuracy, SES, and vocabulary development compared to RiskRI, and showed compensatory/protective mechanisms in right-hemispheric white matter pathways.

Characterizing nonword repetition production error patterns in children with dyslexia

First Author/Chair:Kathryn Cabbage -- Brigham Young University
Additional authors/chairs: 
Tiffany Hogan; Shelley Gray; Mary Alt; Samuel Green; Nelson Cowan

Purpose: Nonword repetition is difficult for many children with dyslexia (see Melby-Lervag & Lervag, 2012 for review) and has been used as part of a diagnostic workup to identify dyslexia in children (Wagner, Torgesen, & Rashotte, 1999). Children with dyslexia have difficulty encoding auditory (and visual) information in correct serial order (Cowan et al., 2017; Peter et al., 2017;). In this study, we tested the hypothesis that nonword repetition errors made by children with dyslexia would primarily be characterized as transpositions (e.g., producing sounds in reverse order), as opposed to substitutions or omissions.

Method: Second-grade children were administered a nonword repetition task of 16 words that increased in length from 2-syllables to 5-syllables. Children’s responses were transcribed and analyzed according to type of error (e.g., omission, substitution, or transposition). We compared the errors made on a nonword repetition task by children with dyslexia with errors made by age- and grade-matched typically-developing children without dyslexia.

Results: Preliminary analyses of a subsample of our total sample shows that as compared to typically-developing children (n=20), children with dyslexia (n=14) exhibit proportionately more transposition errors as word length increased.

Conclusion: These findings lend further support to deficits of serial ordering in children with dyslexia and may be used clinically to differentiate children with dyslexia from children with other nonword repetition deficits.

Dyslexia versus childhood apraxia of speech: similar speech errors with different etiologies?

First Author/Chair:Tiffany Hogan -- MGH Institute of Health Professions
Additional authors/chairs: 
Jenya Iuzzini-Seigel; Jennifer Zuk; Kathryn Cabbage; Jordan Green

Purpose: Children with dyslexia make speech errors characterized by reversing sounds in words (e.g., ‘aminal’ for ‘animal’) and idiosyncratic sound substitutions. Moreover, multisyllabic words can be especially difficult. Interestingly these errors don’t always translate to a clinically diagnosed speech sound disorder. Children with apraxia of speech (CAS) make similar errors, although they do not always have a word reading impairment. In this study we test the hypothesis that children with CAS and dyslexia will show similar speech errors, yet those with CAS will evidence these errors regardless of word length whereas those with dyslexia will evidence these errors only when their system is taxed (e.g., in longer words).

Method: Nine children with dyslexia and 8 with CAS were matched on age, language and intelligence, all within normal range. Those with dyslexia had poor word reading and age-appropriate speech. Those with CAS had poor speech production and good word reading. Children were asked to produce real and nonwords that varied in length.

Results: Children with dyslexia had a similar number of ‘apraxia-like’ features as their peers with CAS. Children with CAS and dyslexia demonstrated limited speech errors when producing monosyllabic real and nonwords; however, as word length increased, children with CAS evidenced an increasing number of errors in both real and nonwords. In contrast, children with dyslexia only showed increased errors when producing multisyllabic nonwords.

Conclusions: Discussion will explore how these groups show similar speech errors resulting from different underlying causes: phonological deficit in dyslexia and motor planning deficit in CAS.

The influence of lexical features and list length on nonword repetition skills for children with speech sound production errors

First Author/Chair:Kelly Farquharson -- Emerson College
Additional authors/chairs: 
Tiffany P. Hogan; Annie B. Fox; John E. Bernthal

Purpose: This study examined relations between neighborhood density, a lexical feature, and list length on performance on an nonword repetition task for children with speech sound disorders (SSD). Children with SSD are at risk for reading disorders, and often experience phonological processing deficits similar to children with dyslexia.

Method: A nonword repetition task was administered to 40 children in second through fifth grades. Twenty of the children were typically developing and 20 had SSDs. A comprehensive speech, language, and literacy battery was administered along with five working memory tasks, including nonword repetition. The nonword repetition task increased in list length from 1 to 4 words. Stimuli in the nonword repetition task were manipulated for neighborhood density. A mixed effects logistic regression analysis was used to determine the extent to which neighborhood density and list length differentially influenced task performance within the sample.

Results: Preliminary results indicate that children with SSD are less accurate on a nonword repetition task, particularly in the 4-nonword list length; a similar pattern was seen in the 2-nonword condition. Neighborhood density did not contribute to task performance, neither as a continuous, nor as a categorical predictor.

Conclusion: Follow-up analyses will examine the influence of additional factors, such as real and nonword reading accuracy in this population.

Considering the implications of speech production on word reading skills in children with and without dyslexia

First Author/Chair:Jennifer Thomson DISCUSSANT -- University of Sheffield

Purpose: To synthesize research results presented in the four symposium talks with an eye toward future research directions and clinical implications.
Method: We will synthesize the outcomes of the four studies. Implications - both for research and practice - will be discussed.
Results & Conclusions: Outcomes from the four studies provide robust relations between the study of speech production and its influence on literacy. Speech production ability has unique relations with word reading ability. Implications for sensitive assessment measures, specific to the type of stimuli (e.g., real/ nonword; mono- and multisyllabic words; list length; high/ low neighborhood density) will be discussed.