Kaani, Bestern (2014-12). The Influence of Orthographic Opacity on Reading Development among Nyanja-English Bilinguals in Zambia: A Cross-Linguistic Study. Doctoral Dissertation.
Thesis
Learning to read is critical to school success and also plays an important role in everyday life. Several studies show that reading proficiency among students differ significantly according to the orthographic depth of the language of instruction. Students taught in transparent orthographies acquire reading skills almost effortlessly and faster than their counterparts taught in opaque orthographies. The English language is considered to have one of the most orthographically opaque writing systems, while Finnish is highly transparent. Accordingly, studies show that students taught to read in English face significantly more challenges than Finnish students. The purpose of this study was to evaluate the effects of orthographic depth on reading proficiency by comparing Zambian bilingual students in a local highly transparent orthography, Nyanja, and English. Equivalent Nyanja and English versions of the Zambia Achievement Test (ZAT) were administered to 240 grades four to six participants drawn from five basic (elementary) schools in Lusaka, Zambia. The ZAT consisted of five linguistically comparable measures of letter discrimination, phonological awareness, word reading, pseudoword decoding, and reading comprehension skills. Of the 240 participants comprising the sample, 119 students received the assessments in Nyanja, while 121 were tested in English. The samples were relatively evenly distributed across the three grade levels and gender. The age of the participants varied widely ranging from 8 to 18 years. The results revealed that participants tested in Nyanja out-performed their English counterparts, and the mean reading proficiency difference was statistically significant on all measures except phonological awareness. Model analysis showed that the English data fitted the reading comprehension model better than the Nyanja data, as all the four model fit indexes used met the required thresholds for the English data, with only two meeting the threshold for Nyanja. The four measures--letter discrimination, phonological awareness, word reading, and pseudoword decoding--accounted for 58% and 49% of the English and Nyanja comprehension variance respectively. Generally, the findings reflect trends in the existing literature that acquiring reading skills is relatively easier in transparent than reading in opaque orthographies. However, in comparison to cross-national monolingual studies, the mean reading differences are slightly moderated probably by the effects of cross-linguistic transfer between Nyanja and English languages. As skills students acquired in one language may have been applied in learning to read in the other language.
