Efficient and Effective Classroom Phonological Awareness Practices to Improve Reading Achievement

摘要

International studies of reading achievement demonstrate that significant inequalities in reading outcomes continue to exist among some of the world’s wealthiest countries, despite strong investment in initiatives directed towards raising literacy achievement for all children (United Nations Educational, Scientific and Cultural Organisation—UNESCO, 2009; United Nations Children’s Fund—UNICEF, 2010). One approach towards the elevation of reading achievement is to investigate how key predictors of reading success are incorporated into everyday classroom literacy practices. Phonological awareness (PA) is widely recognised as a powerful predictor and underlying precursor to early reading success for both typically developing and at-risk readers (Al Otaiba, Kosanovich, & Torgesen, 2012; Blachman, Ball, Black, & Tangel, 2000; Goswami, 2001; Pressley, 2006). A majority of research demonstrating the benefits of PA to literacy growth has been conducted under controlled research settings outside of the classroom environment (Ehri, Nunes, Willows, Schuster, Yaghoub-Zadeh, & Shanahan, 2001; Gillon, 2000a, 2005; Gillon & McNeill, 2009), and thus less is known about whether such benefits hold true when integrated into the heterogeneous classroom setting. For this reason, four experiments reported in this thesis investigated whether PA can be efficiently and effectively integrated into the classroom literacy programme with the overarching aim of raising reading achievement and equalising reading outcomes for the majority of children in the first year of formal education.In the first experiment (reported in Chapter 3), time-efficiency and congruency of scores between a computer-based PA screening and monitoring tool (described in Chapter 2) and a paper-based equivalent were examined. Thirty-three children aged between four years 10 months and five years zero months participated in the study, 12 of whom presented with moderate-severe speech delay (MSD). Participants were randomly allocated to either Group A or Group B experimental assessment conditions. A crossover research design was employed where Group A received the paper-based version of the PA assessment followed two weeks later by the equivalent computer-based assessment (CBA). Group B received the same assessments but in the reverse order of delivery. That is, the computer-based PA assessment first followed two weeks later by the paper-based counterpart. Results demonstrated that: 1) the CBA generated comparable scores to the paper-based equivalent for both children with typical development and children with MSD, and 2) CBA took 31 per cent less time than paper-based administration. These results demonstrate that CBA can provide educators with a time-efficient approach to the screening and monitoring of PA development in the classroom while maintaining equivalency of scores with paper-based testing. Having established the time-efficiency of CBA, the next step was to investigate the use of the computer-based PA screening and monitoring tool as part of the beginning classroom reading programme.In the second experiment (reported in Chapter 4), the influence of a short and intensive period of teacher-implemented classroom PA instruction on reading outcomes in the first year of education was investigated. One-hundred and twenty-nine children aged five-years participated in the study. Using a quasi-experimental design, thirty-four children in two classrooms received 10 weeks of PA instruction from their teachers, as an adjunct to the ‘usual’ reading programme. Ninety-five children from 10 classrooms continued with the ‘usual’ reading programme, which included phonics instruction but did not target PA. Results demonstrated that children exposed to classroom PA instruction performed significantly higher on reading and spelling measures compared to children who received the ‘usual’ reading programme only. Of importance, the number of children experiencing word decoding difficulties after one year of schooling reduced from 26 per cent among children who followed the ‘usual’ reading programme to 6 per cent among children who received classroom PA instruction. These results provide evidence that a short and intensive period of classroom-wide PA instruction in the first year of schooling can have a positive influence on raising reading achievement.In the third experiment (reported in Chapter 5), the effect of classroom PA instruction on raising reading achievement and reducing inequality in literacy outcomes for children with spoken language impairment (SLI) was examined. The data from 129 five-year-old children who participated in the second experiment were extracted and analysed. End-of-year reading outcomes between children with SLI who received classroom PA instruction (n = 7) was compared to: 1) children with typical language development (TD) who received classroom PA instruction (n = 27), 2) children with SLI who followed the ‘usual’ reading programme (n = 21), and 3) children with TD who followed the ‘usual’ reading programme (n = 74). Children with SLI who received classroom PA instruction showed significant improvements in PA, reading and spelling acquisition immediately and up to six months following PA instruction. However, this cohort, in comparison to children with TD, appeared less able to transfer their enhanced PA knowledge to reading and writing tasks. Of importance, children with SLI who received PA instruction performed significantly higher than children with SLI who followed the ‘usual’ reading curriculum; and on par with children with TD who followed the ‘usual’ reading programme. Children with TD who received classroom PA instruction significantly outperformed all other cohorts in this experiment on end-of year reading measures. These results indicate that both children with TD and children with risk for reading difficulties can benefit from classroom-wide teacher-directed PA instruction. These findings have positive implications for elevating reading achievement and reducing inequality between good and poor readers.In the fourth experiment (reported in Chapter 6), the validity and reliability of the computer-based PA screening and monitoring tool was investigated and established. Using a longitudinal research design, the responses of 95 children to test items in the CBA at the start, middle and end of the first year at school were collated and analysed to provide evidence of content, construct and criterion validity, in addition to test-retest and internal consistency reliability. A number of statistical analyses were employed including Rasch Model analysis, exploratory factor analysis and multiple regression analysis. Results demonstrated that the majority of test items were appropriate for five-year-old children in the first year of school and sampled a spectrum of ability levels that would be present in a typical classroom environment. Rhyme oddity, initial phoneme identity and letter-knowledge tasks were most appropriate at school-entry while tasks of final phoneme identity, phoneme blending and phoneme segmentation became more suitable by the middle and end stages of the first year at school. Importantly, performance on the CBA predicted end-of-year reading status with 94 per cent accuracy, and in conjunction with language abilities accounted for 68.9 per cent of the variance in end-of-year reading performance. These findings indicate that the computer-based PA screening and monitoring tool developed and applied in this thesis has sufficient validity and reliability to be used confidently as a time-efficient assessment tool in the classroom.The results from the experiments reported in this thesis provide evidence that PA can be efficiently and effectively integrated into the beginning classroom reading programme from two complementary perspectives: 1) through use of computer-based screening and monitoring of PA skills, and 2) through implementation of a short and intensive period of teacher-directed classroom-wide PA instruction. The results reported in this thesis demonstrate that the evidenced-based integration of key predictors of literacy success, such as PA, into existing classroom programmes can support national and international initiatives that seek to raise reading achievement and reduce inequalities in literacy outcomes for all children.