The Effect of Computer-Assisted Oral Reading While Listening on L2 Speaking Fluency

Abstract

This study investigated the effects of 10 times of once a week computer-assisted oral reading while listening (ORWL) on listening comprehension, objective measures and subjective rater judgment of L2 oral reading fluency and L2 rehearsed speech fluency. In addition, how listening score gains relate to working memory, L2 oral reading fluency gains, or L2 rehearsed speech fluency gains were examined. ORWL is a task of listening, speaking and reading almost simultaneously and is usually incorporated with shadowing or oral reading instruction, but rarely be a focus of study. Forty-six first- and second-year, non-English major, low to intermediate English proficiency Japanese college students (Comparison group n = 24; Experimental group n = 22) participated in this study. Over the course of the semester, the comparison group received reading comprehension instruction twice a week (total of 16 times) whereas the experimental group received reading comprehension instruction once a week (total of 6 times) and ORWL instruction once a week (total of 10 times). In order to enhance the effects of ORWL, pronunciation analyses and self-evaluation of recording of oral reading were also conducted during the ORWL instruction. Data were obtained from conducting pre- and post-listening dictation tests, Momotaro oral reading pre- and posttest, Kaguyahime oral reading posttest, rehearsed speech pre-and posttest, and listening span (working memory) test. Before conducting the quantitative analysis, the dichotomous Rasch analysis was conducted to check the validity and reliability of the listening tests. The results showed that the experimental groups’ listening scores did not significantly improve compared to the comparison group. Regarding the effects on L2 oral reading, the experimental group significantly improved the gain scores of the objective measures of fluency compared to the comparison group. The significant improvement was also found for the mean length of runs and number of pauses per minute between the same passage pretest and posttest, but not between the two different passages. Furthermore, the significant difference was found for the subjective rater judgment of speed, pausing and prosody between the same passage pretest and posttest. Regarding the effects on L2 rehearsed speech, no significant difference was found between the comparison and experimental groups on the gain scores of the objective measures of fluency. On the other hand, the significant difference was found for the subjective rater judgment of speed, pausing and prosody between the L2 rehearsed speech pretest and posttest. Regarding the relationship between the listening score gains and working memory, L2 oral reading fluency gains, and L2 rehearsed speech fluency gains, the moderately strong significant negative correlation was found between the listening score gains and the gain scores of the number of pauses per minute. The results suggest that the computer-assisted ORWL instruction can contribute to pronunciation research because it improved students’ L2 oral reading and made their rehearsed speech more comprehensible by improving the impression of speed, pausing, and prosody. ORWL can also contribute to L2 speech processing research because it improved students’ ability to read aloud an L2 text with familiar vocabulary, and the improvement of this ability was found to be important for listening comprehension.