A Meta-Analysis of Reading Fluency Interventions for Students with Reading Difficulties: Using IRD Index

Abstract

Reading is a process that involves word cognition and comprehension

therefore, it is an essential requirement for students. However, some students still have difficulties reading and cannot comprehend the texts they are presented. Hence, researchers should seek effective reading intervention methods for these students. Until now, many researchers discussed what skills those students with reading difficulties must utilize to develop their reading abilities and what methods can enhance these skills. As a result, the importance of reading fluency has been emphasized. Reading fluency is the ability to read text rapidly, accurately, and with prosody. Significant research has suggested that reading fluency is one of the core characteristics. Thus, research related to reading fluency has been implemented both at home and abroad. Meanwhile, the Response to Intervention (RTI) model is emerging as a diagnosis and intervention model for students with learning difficulties. One of the main features of the RTI model is the provision of interventions according to the students performance level from Tier 1 to Tier 3. The higher tier students are provided with more intensive interventional tools. Therefore, there may be a huge difference in intervention methods between Tier 1 and Tiers 2–3. In short, different instruction strategies can be applied to each tier. The main purpose of this study is to identify effective intervention characteristics for the improvement of reading fluency, focusing on the intervention within Tiers 2 and 3. This study employed the meta-analysis method to derive the general and comprehensive conclusions from the individual studies, and tried to discover the effective instruction methods for small groups (or those learning in a one-to-one environment) by synthesizing the results of single-subject research on reading fluency for students with reading difficulties. Accordingly, the overall effects of the reading fluency intervention were calculated and each effect size was yielded according to student-related variables, intervention-related variables, implementation-related variables, and measurement-related variables. In addition, in this meta-analysis, effect sizes were calculated using the IRD (improvement rate difference) index. Until now, the effect size has mainly been computed using the PND (percentage of non-overlapping data) index. However, recent research has proved that the IRD is a better index for calculating the effects of single-case research, since IRD can discriminate the effect sizes more exactly than the PND index and yield confidence intervals with effect size. Therefore, this study implemented the meta-analysis using the IRD index, presenting the results and suggesting the implications for reading fluency intervention.