國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
69
CHAPTER FIVE
DISCUSSION AND CONCLUSION
The present study addressed three research questions in evaluating the effectiveness of repeated reading and non-repeated reading as a method for increasing EFL elementary students’ oral reading rate and accuracy rate. This
chapter discusses the findings based on the results of the analyzed data presented in Chapter Four. Section 1summarizes the answers to the three research questions. Section 2 compares the findings between the present study and previous studies with possible explanations. Section 3 provides some pedagogical implications derived from the findings of the present study.
Section 4 describes the limitations of the present study. Last, in section5, some suggestions for further research are illustrated.
Answers to the Research Questions
Based on the results of the pretest and the posttest on the participants’
oral reading rate and accuracy rate, the answers to the three research questions are summarized in this section.
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
70
Research Question 1: Is assisted non-repeated reading (Non-RR) as effective as assisted repeated reading (RR) in improving EFL beginners’
reading rate and accuracy rate?
The results demonstrated that the RR and Non-RR groups both improved their reading rate and accuracy rate significantly, indicating that both repeated reading and non-repeated reading approaches are effective approaches for the students’ oral reading fluency.
Meanwhile, there was no significant differences between the two groups’
reading rate and accuracy rate. The results suggest that non-repeated reading is as effective as repeated reading approach in helping students improve their
oral reading fluency. It indicates that repeated reading and non-repeated reading approaches are equally beneficial for improving students’ reading
fluency ability in terms of reading rate and accuracy rate.
Research Question 2: Is there a significant difference between assisted repeated reading and assisted non- repeated reading on higher-level readers’
reading rate and accuracy rate?
It is interesting to note that the higher level students in the RR group did
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
71
not improve their accuracy rate significantly; instead, they improved their oral reading rate significantly only. On the other hand, the higher level students in the Non-RR group significantly improved not only their reading rate but also their accuracy rate.
Nevertheless, no significant differences were observed between the higher level students in the two groups, which indicate that both RR and Non-RR approaches assisted the higher level students in a similar manner.
Research Question 3: Is there a significant difference between assisted repeated reading and assisted non-repeated reading on lower-level readers’
reading rate and accuracy rate?
The posttest results showed that the lower level students in each group
improved their reading rate and accuracy rate significantly, indicating that both RR and Non-RR approaches are helpful for the lower level students’ reading
fluency improvements.
With the comparison of the lower level students’ posttest between the two
groups, we can clearly see that no significant differences existed between the two groups’ reading rate and accuracy rate. In other words, the two assisted
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
72
reading approaches delivered similar growth on the lower level students’
reading rate and accuracy rate.
To sum up, based on the answers to the three research questions proposed, assisted repeated reading and non-repeated reading improved students’ reading fluency significantly after the 16-week instruction. Further, RR and Non-RR are equally beneficial for the EFL young beginners’ oral
reading fluency regardless of different reading levels.
Discussion
The present study aims to compare the effect of repeated reading and non-repeated reading approaches on EFL elementary school students’ oral
reading fluency. Comparing with previous studies, consistent and inconsistent findings obtained in the present study are presented as follows.
Consistent Findings
A contribution of the present study is that it extends previous literature to EFL elementary learners. The following findings of the present study show agreement with previous research.
First, assisted reading approaches, including repeated reading and
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
73
non-repeated reading are beneficial for students’ reading fluency regardless of
reading levels. In the present study, the two groups were instructed with assisted reading approaches such as choral reading, echo reading, and partner reading. After the instruction, both groups significantly improved their reading rate and accuracy rate. Such finding shows agreement with NRP’s
conclusion that assisted and guided oral reading approaches contribute to clear and agreeing improvements on students’ reading fluency abilities. (NRP,
2000).
Second, Non-repeated reading approach is as effective as repeated reading on students’ oral reading fluency. In addition, in regard with students
with different levels, the present study indicates that RR and Non-RR
approaches are equally effective for the participants. The comparison between the higher achievers of the two groups showed no significant differences, and the comparisons between the lower achievers in the two groups demonstrated the same results. Such findings are consistent with literature on the
comparison of repeated reading and non-repeated reading. Previous studies (Homa, et al., 1993; M. Kuhn, 2004; Perfetti, 1985) show that repeated reading and non-repeated reading approaches yield to similar learning outcomes. As Rashotte, and Torgesen (Perfetti, 1985) claimed, repeated reading was not
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
74
more effective than non-repeated reading on improving students’ reading fluency. Likewise, it is congruent with Homan, Klesius, and Hite’s (1993) report
that both repeated and assisted non-repetitive reading on connected texts improved students’ reading rate and accuracy rate. Also, it echoes Kuhn’s
(2004) conclusion that guided repeated and assisted non-repeated reading on basal readers or other books lead to equivalent effects on reading fluency.
Meanwhile, the finding of the present study is in agreement of O’Connor, White, and Swanson’s study (2007) that the assisted repeated reading and
non-repeated reading are equally effective for students’ word identification and
reading rate.
Two possible reasons are provided to the finding of the non-significant differences between RR and Non-RR groups’ oral reading fluency. First, it
seemed that repetition occurs in both repeated and non-repeated reading approaches. According to Samuels (1979), rereading the same short passage
enhances students to read from inaccurately to accurately and finally became fast and automatic reading. Samuels (1979) stated, ―One important function of
repeated reading is that it provides the practice needed to become automatic (P. 379).‖ Later, Moyer (1982) further explained that repeated reading provides
―the highest possible level of redundancy‖ (p.622). Once students are
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
75
instructed the skills to read the words accurately, they need time to practice the skills by rereading the same words so that they are able to gradually achieve automaticity. In sum, it can be inferred that the RR group improved their oral reading fluency by practicing the similar word recognition skills repeatedly in the same short passage until they achieved mastery of the skills.
Considering that the function of repeated reading is providing ―redundancy‖
(Moyer, 1982) of practices on the same word recognition skills, reading different passages may in other ways provide students chances on practicing the same reading skills by reading different passages with similar difficulties.
Similar difficulty-level readings require similar decoding skills and contain similar high frequency words. In the present study, some of the decodable words and high frequency words used in the phonics readers also appeared in the sight word readers. The words overlapped in both sets of readers provided the Non-RR group chances to practice the same word recognition skills. They first practiced those words in the phonics readers and then reread the words again in the sight word readers. As a whole, although the Non-RR group read each book without too much repetition, they actually repeatedly practiced similar decoding skills and high frequency words through similar reading difficulty level materials with overlapped words. In other words, repeated
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
76
reading improves students’ reading fluency by repeatedly practicing the same
word recognition skills on the same short passage, while non-repeated reading assisted students by practicing similar word recognition skills on different passages with similar difficulties. Thus, non-repeated reading can help students achieve automaticity the way repeated reading does.
Another possible reason for the non-significant differences between the RR and Non-RR groups’ reading rate and accuracy rate may be that the
students received similar amount of reading time and reading volumes.
Rashotte and Torgesen (1985) stated that with the same amount of reading volume, both repeated reading and non-repeated reading benefit students’
reading speed and accuracy. Likewise, In Kuhn’s (2003) review of studies relating with fluency instruction, she argued that ―it is not the repetition that leads to the effect but the amount of time spent reading connected text‖ (p.17).
Later, Kuhn (2004) concluded in her study that when readers were provided
the same amount of connected texts, repeated reading was not necessarily a more effective method than non-repeated reading in enhancing readers’ gain
on reading speed and accuracy. Other than the importance of reading volumes, O’Connor, White, and Swanson (2007) stated that time is a critical element .
They found that good readers, compared with poor readers, usually spend
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
77
more time on reading. Those who read more tend to become good at reading with automaticity and comprehension. Reversely, those who are good at reading usually involve in reading more.
In short, when the difficulties of reading materials, reading volumes, and reading time are comparable, assisted repeated reading and non-repeated reading are both effective approaches. In other words, it can be inferred that enough amount of appropriate readings and reading time are two major factors that contribute to their reading speed and reading accuracy gains.
Inconsistent Findings with Previous Studies
In spite of some consistent findings compared with previous studies, inconsistent findings were observed in the present study. The present study revealed that the higher achievers in the RR group improved their oral reading speed significantly, but they did not improve their oral reading accuracy
significantly, which indicates that RR did not help the higher achievers to decrease their miscues significantly. Such finding shows disagreement with
literature (Begeny, et al., 2009a; Dowhower, 1987) that repeated reading is effective for students’ oral reading rate and accuracy rate regardless of
proficiency and reading levels. Possible reasons are discussed as follows.
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
78
Samuels’ (1979) proposal of repeated reading argued that repeatedly
practicing the word recognition skills in the same short passage enables students to gradually achieve automaticity and read the same passage faster with accuracy. With fast and accurate word recognition skills, learners’
attention will be freed to comprehending the texts.
Nevertheless, when compared with non-repeated reading, repeated reading is often criticized as tedious and restricted, which may be the reason for the non-significant improvements on the higher level students’ accuracy
rate after the treatment. Critics (O'Connor, et al., 2007; T. V. Rasinski, 1990) stated that repeatedly reading the same text may bore students and lower their motivation in each repetition. To some readers, it may even be taken as a punishment for not reading well the first time (Homa, et al., 1993). Therefore, it is possible that in the present study, repeated reading approach bored the higher achievers in the RR group so that they took the overly-emphasized repetition as a tedious job and tried their best to complete it as fast as they could without carefully reading the passages. As a result, their accuracy rate did not improve significantly as their oral reading rate significantly raised.
Conversely, in non-repeated reading, students were not required to repeat the same short passage overly; instead, they read more passages in each
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
79
period of class. By reading different passages, they may not be aware of the fact that they were actually repeatedly practicing similar word recognition skills.
For instance, in the ―wh-‖ word families, students in the RR group read the words in the Phonics Reader, ―where, why, what‖, yet the students in the
Non-RR group read more words consisting with the same phoneme in the Phonics Readers and Sight Word Readers, such as ―where, why, what, whale, and whole‖. With different reading passages, their motivation in reading may
not be lowered by tedious repetition. Accordingly the higher level students in the Non-RR group not only improved their reading rate significantly, but also improved their accuracy rate significantly after the treatment.
Homan, Klesius, and Hite (1993) suggested that there seems to be a need to provide the higher achievers in the RR group some specific purposes other than increasing their reading rate such, as (a) reading to the whole class, (b) reading to younger learners (c) dubbing for the story, or (d) presenting in the form of dramatic reading such as reader theater. Rereading the same passage then becomes a purpose for completing the tasks rather than being viewed as
a punishment.
Second, it seems reasonable that because the higher achievers’
motivation in rereading the same passage was not raised, they paid little
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
80
attention to monitoring their peers’ reading. Due to the limit of time, it is challenging for the teacher to monitor each students’ reading. Also, for the
greater ease of use, when RR is applied in the whole class activities, the reading rate and accuracy rate are mostly recorded by the students
themselves and hardly by the teacher or assistant (Yurick, et al., 2006). One concern of peer-monitoring is that if the students do not take the responsibility to monitor their peers’ reading and point out the errors, students’ miscues can
hardly be reduced. For example, the instructor found out that if the two
students working together had the same problem of missing the ending sounds of plural nouns, they may not be able to help each other and point out the errors. Hence, Hoffman and Rasinski (Timothy V Rasinski & Hoffman, 2003) suggested that helping students to learn strategies to monitor themselves is
important for their reading ability development. Likewise, Yetta Goodman (1996) proposed a strategy called ―retrospective miscue analysis‖ to develop students’ ability for self-monitoring. Students directly discuss their errors with
their teachers helps them to analyze their errors and then be able to develop their strategies for self-monitoring.
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
81
Pedagogical Implications
The following four pedagogical implications were drawn from the present study. First, both assisted repeated reading and non-repeated reading are recommended to be used alternatively due to the finding that the RR and Non-RR groups under assisted reading approaches improved their reading fluency significantly after 16 periods of classes. With scaffolding, beginners are able to read books in their instructional level independently and gradually read as well as reduce their miscues.
Second, if the benefits of reading are taken into consideration,
non-repeated reading is recommended more because non-repeated reading improves the students’ oral reading fluency the way repeated reading does
without the drawbacks of repeated reading. During the same amount of time, students are exposed to more texts, ideas, and wider genres. Also,
non-repeated reading is closer to how good readers read outside the classroom; that is, when reading, readers do not usually overly reread the same short passage (O'Connor, et al., 2007); rather, readers tend to read continuously connected texts. In addition, too much the overly repetition on the same short passage, readers’ motivation toward reading will not be lowered.
Moreover, non-repeated reading is appropriate to be implanted in class with
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
82
multi-levels. For the higher level students, reading a wider range of texts and genres enables the students to gain the interest of reading. For the lower level students, through non-repeated reading on the passages with similar
difficulties, they are able to repeatedly practice similar word recognition skills as well as gain the interest of reading.
Third, it is advised that repeated reading to be administered as a means for specific purposes such reading to younger kids, or dubbing for the stories owing to the finding that repeated reading did not help the lower achievers reduce their miscues. However, with specific purposes, students will not take repeatedly reading the same short passage as meaningless and tedious.
Lastly, exposing students, regardless of their reading levels, to abundant reading time and appropriate reading volume is crucial to their reading fluency.
With accurate word reading, students still need time to practice their word recognition skills on connected texts. Consequently, providing students with time on appropriate connected text helps students to become good readers.
Limitations of the Study
The present study demonstrated that the repeated and non-repeated reading approaches had desirable effects on the EFL fourth graders’ oral
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
83
reading fluency. Moreover, it is suggested that non-repeated is as effective as repeated reading approach on improving students’ oral reading rate and
accuracy. Nevertheless, several limitations existed and it is hoped that they are noted for further investigation.
The first limitation concerns with the duration of the instructional time. The instruction was implemented once a week, 16 weeks in total, adding up to 16 periods of class. This may not be long enough for the participants to build up their reading skills. Yet, 16 weeks was close to a semester. Although the results showed that RR and Non-RR are equally beneficial for the participants’
oral reading fluency. It is uncertain whether longer period of instruction results in similar findings. Therefore, a longer-period instruction is suggested to
investigate the long-term effects.
Second, the participants’ attitudes toward RR and Non-RR approaches
are not clear. In the present study, the data collected included pretest and posttest assessments on the participants’ oral reading rate and accuracy rate.
Questionnaires exploring the students’ attitudes toward the two reading
approaches were not included. As a result, questionnaire investigating the students’ attitude toward RR and Non-RR was suggested in future studies.
The third limitation of the present study concerns with the effect of students’
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
84
self-monitoring. During the reading practice session, students were monitored mainly by their partners due to the limit of time and resources. It is challenging for the teacher to monitor each student’s reading and providing feedbacks in
each reading session. It was worried that in peer-monitoring, some errors may be neglected and some feedbacks may be inappropriate. This may reduce the effect of RR and Non-RR in helping students to reduce the miscues. Hence, it is suggested that the students were instructed with retrospective miscue analysis (Goodman, 1996) to help them be more aware of their own errors.
Suggestions for Future Studies
Based on the limitations of the present study, some suggestions for further studies are presented as follows. First, a longer-period instruction is
recommended to examine the long-term effect of RR and Non-RR. Second, questionnaires investigating the students’ attitudes toward RR and Non-RR are suggested to be included in future studies to understand the students’
perspective more. Third, approaches helping students’ to be more aware of
their own errors such as retrospective miscue analysis proposed by Yetta Goodman (Goodman, 1996) is recommended to be instructed before the experimental instruction to avoid the drawbacks of peer-monitoring. Fourth, the