Effects of decoding accuracy variables and (morpho)syntactic competence on the reading of Italian children with and without developmental dyslexia

Abstract

**Abstract:** Developmental Dyslexia (DD) is a heterogeneous disorder that affects reading accuracy and speed to different extents. Different decoding errors can depend on impairments to different components of the dual-route model of reading, thus resulting in different types of DD (Friedmann & Coltheart 2018) even in a shallow-orthography language like Italian (Traficante et al. 2017). Moreover, there is evidence of the effect of morphological information on reading fluency (Burani 2010) and accuracy (Angelelli et al. 2014) of Italian subjects. The present study explores the influence of different measures of decoding accuracy and morphosyntactic competence on the reading of Italian children with and without DD. Forty-two Italian children from the 2nd to the 5th grade of primary school [(14 with a diagnose of general DD and 28 with Typical Development (TD)] were tested on standard measures of reading (Cornoldi & Colpo 2011) and syntactic comprehension (Bisiacchi et al. 2005). A non-standardized test (Arosio et al. 2014) for the syntactic production of 3rd-person direct object clitic pronouns was also administered, and a decoding error analysis based on an adaptation of the coding scheme by Friedmann & Coltheart (2018) was performed. Generalized linear mixed models were analyzed, with the double level of group (DD and TD), syntactic comprehension, and syntactic production as fixed effects, and age and 11 types of reading errors as random effects. The outcomes show that (morpho)syntactic comprehension and production significantly predict reading accuracy. Moreover, negative predictive effects of morphological, semantic, and phonological-output-buffer decoding errors on reading accuracy emerged, thus showing that reading accuracy is significantly affected by morphosyntactic competence and by some decoding errors that have to do with morphology to different extents. Reading speed appeared to be a specific problem of DD children and was significantly affected by the decoding of complex multi-letter compounds. Reading comprehension was not affected by any particular error category but was predicted by standard measures of decoding accuracy and speed, with a stronger effect of the latter. A positive predictive effect of age on written comprehension but not on written decoding was probably due to the acquisition of cognitive strategies and general knowledge, which strengthen children’s contextualization and expectation skills. DD children showed a specific difficulty in decoding longer and morphologically complex words. The significant co-occurrence of some error types suggests the presence of diverse comorbidity profiles among different types of DD. A weak positive effect of age, matched with a negative effect of syntactic comprehension, on errors in decoding complex multi-letter compounds might be evidence of the use of the lexical route of reading as children’s age increases, even by resorting to inadequate compensation strategies. To conclude, morphosyntactic skills predict reading accuracy in children with and without DD. (Morpho)syntactic comprehension is particularly related to the recognition of complex multi-letter compounds, which, in turn, affects reading speed. This suggests that oral morphosyntactic training might help reading accuracy and, indirectly, reading speed. This might be particularly useful for DD children, whose reading speed, as well as accuracy in decoding longer morphologically complex words, was specifically impaired. At the same time, a fine-grained analysis of decoding errors made by DD children might help the assessment of impairments to specific sections of the reading model, thus allowing for ad hoc intervention programs.