Development of Reading Fluency During an Extensive Reading Course Incorporating Reading Fluency Tasks

Abstract

Reading approaches in Japan and other EFL settings have typically focused on accuracy at the expense of fluency. However, reading rate and fluency are important components of reading that are currently being neglected. A review of the literature reveals growing interest in investigating the effectiveness of reading rate and fluency enhancement approaches such as extensive reading (ER), speed reading, and oral fluency training. Results of previous research indicate the effectiveness of these approaches for increasing reading speed and fluency as well as comprehension, but the overall body of evidence in this area remains weak, particularly in EFL contexts.With these gaps in mind, the purposes of this study were to investigate the effectiveness of the ER approach in developing reading fluency, investigate appropriate ways to measure reading fluency in adult EFL readers, investigate whether reading fluency increases are accompanied by reading comprehension increases, and identify the patterns of reading rate development that occur during an ER course. To this end, a longitudinal quasi-experimental study was conducted with 77 first-year students at a nursing college in Japan, all of whom participated in one 15-week semester of ER with an added fluency training component (ERFT) and one 15-week semester of intensive reading (IR). Silent reading rate, oral reading rate and fluency, reading comprehension, and written receptive vocabulary size were measured at the beginning and end of both semesters. Silent reading rate, oral reading rate, oral reading fluency, and reading comprehension were measured using both simplified and authentic texts. Approximately half of the students took ERFT in the first semester and IR in the second semester, and the other half did the opposite, so analyses were conducted separately for these two groups. Repeated-measures ANOVAs showed that for the IR-before-ERFT group, oral reading rate increase was significantly greater for the ERFT treatment (9.79 wcpm) than the IR treatment (3.05 wcpm) (p < .001) for simplified texts; silent reading rate increase was significantly greater for the IR treatment (8.26 wpm) than the ERFT treatment (-2.86 wpm) (p < .001) for authentic texts; and oral reading rate increase was significantly greater for ERFT (15.18 wcpm) than IR (4.37 wcpm) (p < .001) for authentic texts. For the ERFT-before-IR group, silent reading rate increase was significantly greater for ERFT (19.22 wpm) than IR (-2.50 wpm) (p < .001) for simplified texts; oral reading rate increase was significantly greater for ERFT (12.55 wcpm) than IR (4.81 wcpm) (p < .001) for simplified texts; silent reading rate increase was significantly greater for ERFT (21.86 wpm) than IR (-7.64 wpm) (p < .001) for authentic texts; and oral reading rate increase was significantly greater for IR (13.06 wcpm) than ERFT (6.70 wcpm) (p < .001) for authentic texts. Repeated-measures ANOVAs also showed that for the IR-before-ERFT group, reading comprehension increase for authentic texts was significantly greater for IR (1.00 logits) than ERFT (.17 logits) (p = .010); and vocabulary size increase was significantly greater for IR (.36 logits) than ERFT (-.19 logits) (p = .001). For the ERFT-before-IR group, reading comprehension increase for authentic texts was significantly greater for ERFT (1.05 logits) than IR (.11 logits) (p = .003); and vocabulary size increase was significantly greater for ERFT (.44 logits) than IR (-.07 logits) (p = .001). No significant correlations were found between changes in any of the reading rate and fluency measures (silent reading rate, oral reading rate, and oral reading fluency) and reading comprehension change. For the ERFT-before-IR group, analysis revealed a significant and moderately high negative correlation (r = -.422) between initial silent reading rate and silent reading rate change (on simplified texts) during the ERFT treatment. Analysis of individual students’ silent reading rate change patterns during the ERFT semester revealed a jagged increase to be the most common (22/77; 28.57%) pattern of rate increase. These results add to a growing body of evidence that ERFT yields greater reading fluency gains than IR, suggest that both ERFT and IR contribute to reading comprehension and vocabulary size gains, suggest that ERFT yields greater reading fluency gains among students with initially slower reading rates, and provide insight into how individual’s reading rate changes over the course of a semester of ERFT.