國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
27
CHAPTER THREE
METHODOLOGY
This study investigated the effect of repeated reading (RR) and non-repeated reading (Non-RR) on EFL elementary school students’ oral
reading rate and accuracy rate. A quasi-authentic experiment was conducted
to answer the research questions:
1. Is assisted non-repeated reading as effective as assisted repeated reading in improving EFL beginners’ reading rate and accuracy rate?
2. Is there a significant difference between assisted repeated reading and assisted non- repeated reading on higher-level readers’ reading rate and
accuracy rate?
3. Is there a significant difference between assisted repeated reading and assisted non-repeated reading on lower-level readers’ reading rate and
accuracy rate?
This section presents the methodology in four parts: participants, instrument, procedure, and data analysis.
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
28
Participants
The participants in this study were 59 fourth graders from two classes, aged from ten to eleven, studying in a public elementary school in downtown Taipei, northern Taiwan. They were instructed by the same teacher in grade three and four. These participants were selected for two following reasons.
First, they were at a transitional stage in reading (Chall, 1983; Dowhower, 1987). Based on Chall’s stages of reading development (1983), students who
can read words accurately but have not achieved automaticity and have to deliberately decode words are transferring from reading stage 1 to stage 2, i.e.
from deliberately decoding to automatic reading. These participants had started formal English courses since the second grade. As second graders, they received one period of English class per week, forty minutes per class, focusing on letter names, and letter sounds. As third graders, they had two periods of English classes every week, emphasizing basic phonics skills and some daily dialogues. By the end of grade three, they were able to blend and decode some simple CVC words and had a word bank of about 100 words.
Moving toward grade four, the participants had two periods of formal English classes and two periods of additional classes each week. They were expected to be able to read simple passages consisting of three to four short
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
29
sentences. This was the year they started to read from word lists to short passages. What they needed was to keep practicing and mastering their word recognition abilities by reading the stories they already knew or rereading the same passages (Chall, 1983). According to the instructor who worked with the participants since grade three, most of the participants were not able to read short decodable word and high frequency word list fluently. Instead, they usually chopped the sentences inappropriately because they had to make an effort to decode the words or recognize the high frequency words. Once they had the automatic word recognition skills, they were able to read short
passages easily and shift their attention to the meaning of the texts. After acquired automatic reading, they were able to read for learning. Therefore, fourth graders were selected as the participants, for they were transferring from slow word decoding to automatic reading.
Second, according to the English teacher of the two classes, their English proficiency and learning motivation were very close. The two classes not only performed similar learning outcomes on the formative and summative English tests in grade three but also held positive motivation toward learning English.
Both classes were interested in learning new things and enjoyed interacting with the instructor as well as their peers.
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
30
Instruments
The instruments in this study include two sets of graded readers as reading materials, pretest and posttest materials, and scoring sheets.
Graded readers
The target learning materials in this study were chosen from two different series of graded readers published by Scholastics: Scholastic Phonics
Readers and Scholastic Sight Word Readers. Considering that the textbook used by the participants was designed by topics with related vocabularies, sentence patterns and dialogues but with very few readings, graded readers can meet the need for them to be exposed to connected and interesting texts.
Scholastic Phonics Readers and Sight Word Readers have been proved as effective reading fluency materials for beginning readers (Slavin, Lake,
Chambers, Cheung, & Davis, 2009). Furthermore, the decision of selecting the two series as supplemental reading materials for fourth graders was made by all the English teachers in the participants’ school.
The two sets of readers are short stories with colorful pictures enriching the context and topics relevant to students’ experiences. Both sets contain
CD-ROM recorded by native speakers. The first set of selected readers,
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
31
Scholastic Phonics Readers, includes 72 books with difficulties gradually leveling up. Each book deals with different sound features. For example, readers 7 to 12 focus on the consonant ― m, l, t, s‖ and short vowels ―a and o‖
sound. Except for decodable sound features, the stories also contained high frequency words overlapped in each story. The second set of learning
materials, Scholastic Sight Word Readers, contains 25 readers with similar difficulties. Each Sight Word Reader focuses on two different high frequency words, for example, ―are‖ and ―at‖ in book two. Meanwhile, each sight word
reader contains basic decodable words as well.
16 Scholastic Phonics Readers were selected for both repeated reading (RR) and non-repeated reading (Non-RR) group, while 16 Scholastic Sight
Word Readers were selected for Non-RR group only, meeting the requirement of the participants’ instructional level from 90% to 97% of accuracy rate. The
book selection was carried on in the pilot study by the 12 randomly selected fourth graders in the same school with the participants without participating in the main study. Based on the accuracy rate gained by the 12 students in the
pilot study, Phonics Readers served as a reference for book selection. Phonics Readers 9 to 13 were within the students’ instructional level, while reader 14 was out of the students’ instructional level, i.e. below the readers’ 90% of
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
32
accuracy rate. Considering the feedback from the students in the pilot study and the requirement of the difficulty level within the students’ instructional level,
reader 13 was selected as the first book for the present study. Because the intervention lasted for 16 weeks, Phonics Readers 13 to 28 were selected as their learning materials, adding up to 16 books. The similar procedures went with the 16 Sight Word Readers selection.
Caldwelll (2007) claimed that reading materials can be leveled as follows:
independent, instructional, and frustration level. The materials that students can read with above 98% of accuracy rate are at the students’ independent level, which students can work on by themselves. Materials with students’
accuracy rate between 90% and 97% are within learners’ instructional level,
which means with appropriate support and guidance, readers can accomplish the job. However, materials less than 90% of accuracy rate are beyond
readers reading ability. Even with support, they still cannot comprehend the material. Reading below the instructional level might lead to struggle and frustration. In sum, Scholastic Phonics Readers 13 to 28 and the other 16 Scholastic Sight word readers (see Table 3.1) were preferred because they were designed for the beginning readers and the difficulty levels were within the participants’ instructional level.
‧
Table 3.1 The 16 selected Phonics Readers and Sight Word Readers
Weeks
Scholastic Phonics Readers (for RR and Non-RR groups)
Scholastic Sight Word Readers (for Non-RR group)
1 book 13 I, see
Pretest and posttest materials
The last Phonics Reader, reader 28, was presented as the pretest as well as the posttest material. The texts of reader 28 were copied on white A4-size sheets with larger font and without pictures attached. When being recorded at pretest and posttest, the students read the sheet instead of the book to avoid orthographic affect. In order to help students associate the meaning with the words, some of the decodable words in the Phonics Readers are printed with small pictures right above the words. If the students read the book while
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
34
assessing, it is unknown whether they sound out the words because of their reading skills or because of the trace of the pictures. Therefore, in the pretest and posttest, the students read the passage on the sheet without the presence of the pictures.
Scoring system
The scoring system of oral reading fluency was based on the DIBELS accuracy and fluency scoring system. Dynamic Indicators of Basic Early
Literacy Skills (DIBELS) is a standardized assessment assessing K-6 learners’
five literacy skills: phonemic awareness, alphabetic principle, accuracy and
fluency, vocabulary, as well as comprehension. DIBELS is widely adopted by teachers and researchers to assess students’ oral reading fluency due to the
easy administration and empirically proved adequate procedure. (Fuchs, L. S., Fuchs, D., Hamlett, C. L., Walz, L., & Germann, G., 1993; Shinn, 1989;
National Reading Panel, 2000).
In the fluency and accuracy assessment, two scores were assessed: the
oral reading rate and the accuracy rate (see Table 3.1). The oral reading rate was scored to understand the participants’ reading speed, while the accuracy
rate was scored to understand whether the students decrease their miscues
‧
after the intervention. The oral reading rate was based on the words per minute (WPM) the participant read with error words subtracted. The accuracy rate was gained through the total accurate words divided by the total words read (see Table 3.1).
total words read– error words= reading rate
Accuracy rate
total words read – error words
total words read
Based on the DIBELS scoring system, self-corrected words, insert words, and repeated words were counted correct. However, mispronounced words, omitted words and change of word order were incorrect. If a students struggles
in a word over three seconds, it was demonstrated but the word would be slashed as incorrect on the examiner’s sheet (see Table 3.2 for samples of
scoring).
X100 = accuracy rate
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
36
Procedure
The entire procedure of the present study consisted of three parts: the pilot study, the main study including the pretest, the16-week intervention, and the posttest, as well as data analysis (See Figure 3.1).
In this session, the following procedures were implemented: (1) the pilot study; (2) the main study; (3) the accessing procedure; (4) teaching procedure;
and (5) data analysis.
Table 3.3 Sample of scoring
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
37
Pilot study
To justify the appropriateness of the procedures of the main study, the pilot study was carried out for four periods of classes in one week, adding up to 160 minutes. The students were 12 randomly selected fourth graders from the same school who did not take part in the main study. The procedures and instrument followed the main study except that the students read fewer readers than the participants in the main study because of the shorter period of treatment. The six students in the RR group read four Scholastic Phonics Readers, while the other six students in the Non-RR group read four
Scholastic Phonics Readers and four Scholastic Sight Word Readers.
Before the four reading classes, the students in the pilot study first helped select readers meeting their instructional reading level. During the four weeks, the instruction covered Phonics readers 13 to 16 and the first four Sight Word Readers. Phonics Reader 16 was presented as the pretest and posttest material for the 12 students in the pilot study.
Overall, the students from the two groups improved their reading rate and accuracy rate after the pilot study. The RR group improved their WPM from 63.5 to 82.5 words, while the Non-RR group gained growth from 66.5 to 80.67 words. The RR group’s accuracy rate was 90.16 in the pretest and 95.12 in the
‧
posttest, whereas the Non-RR group’s accuracy rate was 89.73 in the pretest
and 94.92 in the posttest (See Table 3.4). As a consequence, the procedure of the pilot study was considered an appropriate framework for the following main study.
Table 3.4
Statistics and Independent Samples t-test of the Pilot Study
P>.05
Based on the pilot study, two teaching procedures were added in the main study. First, the importance of reading prosody was needed to be mentioned.
The researcher found that if prosody was overlooked, the students only emphasized on the speed which very often made their reading obscured and
RR
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
39
unnatural. Second, some difficult words should be taught before reading, especially the high frequency words that cannot be decoded, such as ―said and here‖.
‧
Figure 3.1 Procedure of the StudyPre-test: oral reading rate and accuracy rate (Material: Scholastic Phonics Reader 28)
Non-repeated reading (Non-RR)
Procedures Echo reading once→
Partner reading twice
(Each participant read 2 times/ per book.)
Learning Materials:
1. Scholastic Phonics Readers 13 to 28
2. Scholastic Sight Word Readers 1 to 16
Posttest: oral reading rate and accuracy rate (Material: Scholastic Phonics Reader 28)
Data analysis A 16-week intervention Repeated reading (RR)
Procedures:
Echo reading twice→
Partner reading five times (Each participant read 5 times/
per book.)
Learning Materials:
1. Scholastic Phonics Readers 13 to 28
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
41
Main study
A pretest was first administrated to ensure that the oral reading rate and accuracy rate of the two classes were homogeneous. Meanwhile, students whose reading rates were the top half in both groups were classified as
higher-level readers, while students scoring on the bottom half were classified as lower-level readers. Next, there was a 16-week intervention, 40 minutes per week, from September 2010 to January 2011. The two classes were randomly assigned to two different treatments: the RR and Non-RR groups. Every week, the RR group read one selected Scholastic Phonics Reader with assisted repeated reading approaches, while the Non-RR group shared the same Phonics Reader and read one additional selected Sight Word Reader with assisted non-repeated reading approaches. During the 16 weeks, both RR and non-RR groups read Scholastic Phonics Readers from 13 to 28 in order. After
the last intervention class, the posttest was carried on immediately after the intervention to measure the progress of the RR and Non-RR groups’ reading
rate and accuracy rate during the 16 weeks. Lastly, the data collected from the pretest and posttest was analyzed qualitativelly with independent samples t-test and paired samples t-test by SPSS.
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
42
The assessing procedure
When administering the oral reading fluency test, every examinee was assessed individually. Both the participant and the examiner as the researcher had one A4-size examiner sheet of the reading passage from Scholastic Phonics Reader 28.The participant could not see the running record on the researcher’s copy. As soon as the examiner timed for one minute and said,
―Start,‖ the participant began to read. At the end of one minute, the examiner ended the assessment by saying, ―Stop,‖ and put a bracket right after the last word produced by the examinee. Each participant’s oral reading fluency
assessment was recorded by a pen recorder for later analysis of reading rate and accuracy rate by the other scorer. One third of the data was scored by the researcher and another experienced English teacher to ensure the inter-rater reliability.
Teaching procedure
The teaching procedures for both the RR and Non-RR groups were the same, except for the repeating times on the same text and the variety of reading texts. Each student in RR group read the same Scholastic Phonics Reader for five times, while every student in Non-RR read it only twice and
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
43
repeated all the procedures for the second book, Sight Word Reader, in one class. Hence, during the same reading time, every student read similar amount of texts.
The students were instructed with assisted reading approaches, echo reading and partner reading. These two approaches were chosen because the first approach, echo reading, provided a reading model for students to imitate the pronunciation and prosody from the model, and the second one, partner reading, provided chances to interact with their peers, which made reading more interesting and meaningful to them.
To ensure that reading comprehension and prosody was not neglected, the researcher told the students the importance of reading rate, reading
comprehension and prosody before the intervention. They were instructed that a good reader can read accurately, fast, and with prosody. While reading fast, they had to maintain the intonation, expression, and comprehension as well.
The teaching procedures were presented in three sessions: (a) before reading, (b) during reading, and (c) after reading (See figure 3.2). The first one was intended for the participants to associate their background knowledge and learning experiences with the book they were going to read. The
During-reading session was the time for the students in RR and Non-RR to
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
44
read the reader with different treatments. Session C included some following activities after the reading.
In session a, before reading, the researcher as the instructor first directed the participants’ attention to the book title and previewed the pictures in the
book for them to make some predictions. Then, it was in the pretest that the students paired with their peers themselves and recorded each other’s reading
time and errors on the self-monitor record for each other.
In session b, during reading, the students read the readers with two assisted approaches, echo reading and paired reading. Dowhower (1987) suggested that when applying repeated reading, the ideal number of repetition is between three and five times. Therefore students in RR treatment read each book five times, while the students in Non-RR group read the each book only two times.
The whole class first echoed with either the CD-ROM or the instructor sentence by sentence. The RR group echoed the reading twice, while the Non-RR group echoed the reading just once. Next was paired reading when the students were paired randomly. Two partners took turns to start reading one sentence each time, which means in each partner reading, every
participant only read half of the book. The RR group followed paired reading
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
45
for four times and partners chorally read the whole story chorally for the fifth time. In sum, every participant in RR group read the same short passage twice in echo reading and three times in paired reading, summing up to five
repetitions on the same book. On the other hand, each student in Non-RR group followed paired reading just twice. Hence, every participant in Non-RR group read the same short passage once in echo reading and once in paired reading, adding up to two repetitions on the same book.
In the last session, after reading, some detailed comprehension questions were discussed. Then, it was the posttest. The students took turns to time their partner’s reading. Words read per minute and miscues they made were
recorded on their own self-monitor sheet. The self-monitor records were not for the data collection; instead, they were for the students to monitor their own progress. After the Phonics reader, the Non-RR group read another Sight Word Reader following sessions a, b, and c.
‧
Figure 3.2 Procedure of the Study
b. During Reading
1. Echo reading 2. Paired reading
a. Before Reading 1. Previewing the book
2. Pretest (Students timed for their partners.)
c. After Reading 1. Detailed comprehension questions Each student read 2 times/ a book
Repeat sessions a, b, c for the Sight
Word Reader
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
47
Data analysis
This study aims to answer the three research questions: (1) Is assisted non-repeated reading as effective as assisted repeated reading in improving EFL beginners’ reading rate and accuracy rate? (2) Is there a significant
difference between assisted repeated reading and assisted non- repeated reading on higher-level readers’ reading rate and accuracy rate? (3) Is there a
difference between assisted repeated reading and assisted non- repeated reading on higher-level readers’ reading rate and accuracy rate? (3) Is there a