Robin van Rijthoven

145 General discussion 6 in line with the lexical quality hypothesis (Perfetti, 2007) and lexical restructuring hypothesis (e.g., Metsala & Walley, 1998) to boost their reading via the phonological pathway and spelling directly. This confirmed the suggestions that semantic abilities help to circumvent phonological decoding deficits as mentioned by Haft et al. (2016) Besides the compensatory effects of semantic abilities, we also hypothesized that a phonics through spelling intervention could help children with dyslexia to strengthen their phonology-orthography bi-directional connection by stimulating reading and spelling to the same extent in a phonics through spelling intervention. Most phonics interventions so far focused mainly on reading and thus on strengthening the orthography-phonology connection (see Galuschka et al., 2014). Since spelling is more difficult than reading (Bosman & Van Orden, 1997), more time on instruction and practice of spelling development seems necessary. This time can be used to form the more detailed and specific orthographic representations that are necessary to learn to spell. According to the lexical restructuring hypothesis this could benefit phonological development as well. So, by combining reading and spelling in a phonics through spelling intervention both orthography-phonology and phonology-orthography relations grow stronger. The addition of a great deal of spelling therefore is important to arrive at better-defined bi-directional orthography-phonology connections. Our results showed that such an intervention for children with dyslexia indeed has a significant impact on word reading, pseudoword reading, and spelling development. Additional analyses regarding the response to intervention revealed that the results of the intervention were robust to individual differences in the phonological deficit, verbal learning, and verbal consolidation. We found that spelling results were influenced by individual differences in semantic abilities in the sense that more defined semantic abilities were related to a better response to the phonics through spelling intervention. Our findings contribute to knowledge about the complex interplay between multiple cognitive factors in defining difficulties in learning to read and spell as described in the Multifactorial causal model of dyslexia (Catts et al., 2017; McGrath et al, 2020; O’Brien & Yeatman, 2021; Pennington, 2006; Protopapas, 2019; Van Bergen, 2014b). We contribute to the existing literature by pointing out that a strength in semantic representations can foster children’s reading and spelling development in a rather transparent language such as the Dutch language, and may possibly compensate for weak bi-directional phonology-orthography relations. Children with dyslexia benefit from stronger semantic-phonology relations when learning to read and spell. In addition, interventions that focus onboth readingand spelling stimulate thebi-directional relationshipbetween phonology and orthography by building more specific orthographic representations. Although these compensatory sources matter for children with dyslexia, the effects are rather small. Consequently, these strengths did not prevent them from becoming weak