published_or_final_version / Linguistics / Master / Master of Philosophy

published_or_final_version / Psychology / Master / Master of Philosophy

Lexical compounding refers to the process of word formation through union of lexicalized morphemes. Given that young Chinese readers learn print vocabulary as unanalyzed whole, I am uncertain whether children can effortlessly decompose bound morphemes from disyllabic words for lexical compounding to occur. With this concern, I propose a lexical blending process in parallel with lexical compounding, where words are constructed from previously learnt words that have not yet been decomposed as morphemes. This thesis investigated the mechanisms behind the lexical blending process, as well as its role in word reading among young Chinese readers, in five studies Studies One and Two examined the factors that favor lexical blending to occur. In Study One, I located a high proportion of disyllabic words and bound morphemes within a corpus of Chinese textbooks in Hong Kong. Around 40-50% of disyllabic words in Grade One to Grade Three are composed of one or more bound morphemes, which set a favorable environment for lexical blending to occur. In Study Two, I found that younger readers tended to commit more selection errors, defined as “naming the target character as a character that forms a highly frequent two-character compound word with it” (Shu, Meng, Chen, Luan and Cao, 2005), than older readers during character reading, suggesting that their representations of bound morphemes were not precise. An experiment on morpheme name judgment demonstrated that bound morphemes and low frequency morphemes embedded in high frequency words were most prone to selection errors. I further examined the lexical blending process and its contribution to reading development in Studies Three and Four. Adopting a cross-sequential design in Study Three, I found that lexical blending concurrently and longitudinally predicted Chinese word reading, after lexical compounding and other reading-related variables were partialled out. In Study Four, I located lexical class and structural relation knowledge as significant component skills of lexical blending. The process of lexical blending proceeded first with structural arrangement of words, followed by morphological decomposition and union of morphemes to eventually form a blended word. I also tested Chinese dyslexic readers’ performance on lexical blending in Study Five. Dyslexic readers exhibited difficulties in lexical blending and all the related component skills, when compared with chronological-age (CA) matched controls. Process-wise, the dyslexic readers were weaker than CA controls in both structural arrangement and morphological decomposition, while having particular difficulties in the latter process. I conclude that lexical blending is an important word formation process for young Chinese readers. To aid mastery of lexical blending, readers should be aware of the syntax in phrases and sentences, as it provides cues on structural arrangement of blended words. In addition, I suggest explicit instruction on lexical blending skills in the curriculum, with a particular focus on morphological decomposition, in order to meet the learning needs of dyslexic readers. / published_or_final_version / Psychology / Doctoral / Doctor of Philosophy

The main objective of the present thesis was to examine whether and how do different aspects of oral language skills have important contribution for the development of reading comprehension among young Chinese children. In Study 1, a three-wave longitudinal study (from K2 to P1) was conducted with 91 Chinese children, to whom measures of oral language (vocabulary, grammar and narrative discourse), word-level skills (phonological, orthographic, and morphological skills), and word reading were administered at all three time points, and reading comprehension at K3 and P1. This study found that K2 oral language skills explained considerable variance in subsequent word reading and reading comprehension two years later, and their longitudinal contribution appeared to be more important than that of K2 word-level skills. Moreover, it was found that the first-graders showed rudimentary abilities in some higher-order text comprehension skills in spoken language (e.g., sentential ambiguity detection and monitoring of textual coherence), and these language skills were highly associated with children’s reading comprehension. Results of multiple regression analyses showed that these skills had unique contribution to reading comprehension at P1 over and above that of word reading, word-level skills, and general oral language skills (i.e., vocabulary, grammatical and narrative discourse skills). The longitudinal data further demonstrated the unique prediction of sentential ambiguity detection over time. Overall, SEM analyses revealed that although the role of word reading ability was prominent to beginning reading comprehension, children’s early oral language skills at preschool were found to make an independent path to later reading comprehension through facilitating the subsequent development of higher-order comprehension skills at both sentence- and discourse-level. Since Study 1 demonstrated the importance of sentential ambiguity detection to reading comprehension, Study 2 focused on examining the emerging development of this skill in Chinese children from K2 to P1. A phase model was proposed which hypothesized that children gradually acquired the three sub-skills of sentential ambiguity detection (i.e., homophone detection, lexical ambiguity detection in sentence, and structural ambiguity detection in sentence) through successive phases. Study 2 further demonstrated the contribution of vocabulary, grammatical, and lexical compounding skills for the early acquisition of ambiguity detection in Chinese children. In Study 3, one of its aims was to determine the extent to which different kinds of preschool cognitive skills significantly predicted later word reading difficulties at the end of first grade. The results highlighted the important predicting role of meaning-related cognitive skills (i.e., vocabulary, lexical compounding, and homophone detection skills) in addition to that of phonological and orthographic skills. Apart from focusing on children’s word reading problems, Study 3 further used K-mean cluster analysis to identify a group of children at first grade, whose reading comprehension fall short of their average or good word reading ability. It was found that these unexpected poor comprehenders showed weaknesses in skills that are specifically related to text comprehension: sentential ambiguity detection, comprehension monitoring, and working memory. The practical implications for early literacy instructional approaches and early identification of children with reading difficulties were discussed. / published_or_final_version / Psychology / Doctoral / Doctor of Philosophy

published_or_final_version / Speech and Hearing Sciences / Master / Master of Philosophy

published_or_final_version / Speech and Hearing Sciences / Doctoral / Doctor of Philosophy

published_or_final_version / abstract / Education / Doctoral / Doctor of Philosophy

published_or_final_version / abstract / Linguistics / Master / Master of Arts

published_or_final_version / abstract / Linguistics / Master / Master of Arts

published_or_final_version / Linguistics / Master / Master of Arts