تحقیق با موضوع test، reading، In، Statistic
Part 5: ten four-option multiple choices cloze, with an adapted-authentic text drawn from a variety of sources. The text was of a factual or narrative nature.
3.2.2. The reading posttest
After the completion of the treatment during a course of 10 sessions, each 40 minutes, the reading section of a PET test, a different version from the one used before the treatment, was administered as a posttest to both groups in order to measure their reading performance after the treatment. The reliability and item characteristics of this test were checked to be appropriate in a pilot study beforehand. The test turned out to be as high as 0.89….in reliability, and as the result of the item analysis item facility was 0.40-0.26. And ID was 0.41-0.56. The inter-reliability of writing was 0.9387.
3.2.3. Instructional Material
The textbook that was used in this study was ‘Select Reading’ by Lee and Gundersen (2001). There are 14 chapters in this book. Each chapter includes a 2-4 page passage. The new words of each passage are defined by a phrase. Each passage followed by some exercises, in the first exercise there are scramble sentences based on what in the passage that must be ordered. In the next exercises there are 3-4 questions for a free discussion. In the third exercise there a 3-multiple choice question, which must be answered by using the context. The following exercise includes sentences with a blank that must be filled based on the information in the passage. In the following pages some grammatical points such as phrasal verbs and past perfect tense. In the next pages there are two or three open-ended stories that can be discussed. Finally there is crossword puzzle that can be solved by using the words of the passage. The researcher chose the exercise one, three, and four. Because they are related to the comprehension of the text. The researcher taught 9 chapters of it. She did not follow the chapters in order, because some students usually read the chapters at home, and check the new words by dictionary; to avoid the interference of this to the results of study, the researcher selects the passages randomly; she taught chapters 1, 3, 4, 6, 7, 9, 11, 12, and 14.
3.3.1. Piloting the PET Test
The first step was to pilot the PET test by which the research could identify problematic items if any. For this reason, the researcher administered the reading and writing sections of the PET test to a sample group of 30 participants who shared the same characteristics (gender, proficiency level, age) with the participants of the main research. Then the item characteristics and reliability were calculated and due to the acceptable range of IF and ID 0.39-0.60 and 0.40-0.55 respectively no items has been deleted. Finally, the internal consistency of PET test was 0.87 and the correlation coefficient was 0.9867.
The results of piloting posttest: Item facility of these items is between 0.40-0.62 and item discrimination of these items also is 0.41-0.56. Therefore no items were deleted. The reliability of test id 0.89. The correlation coefficient between two rates is 0.9387.
3.3.2. Homogenizing the Participants
After piloting the PET test on 30 participants of the same characteristics of the main sample, the modified PET test was administered to 95 students as a language proficiency test to make sure that the students were similar in terms of their English proficiency level and reading abilities. In this language proficiency test the students had to answer the reading and writing questions in the allocated time, and finally the researcher compared their mean scores through t tests and made sure that there was no significant difference between them in those regards. It should be mentioned that the teacher administer the test. They have 1.30 minutes to finish the exam.
3.3.4. The Treatment
The characteristics and class conditions of the two groups were the same including gender, average age range of 15-17, the number of sessions, hours, the time of test, the PET questions. In the control group the researcher first taught the key vocabularies (as specified in the textbook). The researcher wrote the key words on the board and asked d students whether they knew the meaning or not, if they did not know, she wrote the definitions provided by the author, then taught and practiced the two reading strategies of skimming and scanning. After that she asked the students to read the text and answer the determined questions. This procedure was followed all throughout the course for each unit of the textbook.
In the experimental group the researcher taught the key vocabularies like the control group, and then used the scaffolding model, developed by Dansie (2001):
Model of a scaffolding reading comprehension adopted from Dansie (2001 p.66).
This model has four parts:
Building the field. In this phase the researcher set the scene by asking a question, using a shared experience based on the topic of the passage. This phase lasted 3-5 minutes. It adopted for each unit of the book.
Modeling. In the second phase the researcher models how she thinks and read the passage line by line. ”At first I read the first and the last sentence of the paragraph. I guess the meaning of the words I don’t know, if I cannot guess the meaning I continue reading to understand the gist of the text. I do not focus on words begun with capital letters; they are usually the name of place and persons. Then I write two or three sentences for the paragraph as a summary”.
Joint construction. In this phase learners read the passage themselves, and without any hesitation asked if they have problem. If they asked the question, the teacher would not give the answer quickly; she asked them to think and then supported them to reach the answer themselves. In this phase the teacher checked the summaries next to each paragraph. And asked students how many words they guessed or skipped. In this phase the teacher provided guidance for students.
Independent construction. In this phase the students themselves did the model independently. The teacher handed over the whole burden on the shoulders of students.
For the first two chapters the teacher did the phases one and two [building the field and modeling]. For the next four chapters the teacher did the phases one, two and three[building the field, modeling, and joint construction], and for the rest of chapters, students could did the model independently.
3.3.5. Administration of the Reading Post-Test
The PET test was piloted in which the researcher administered the pet test to the sample group of 30 students. After receiving ten sessions of instruction in reading in both control and experimental groups, all the participants in both groups were given the reading posttest of PET, different from the PET used before the treatment, in order to measure the effect of the treatment and test the null hypothesis of the study.
This study had one independent variable and one dependent variable. The independent variable was scaffolding, and the dependent variable was English reading comprehension ability. The control variable was language proficiency and gender. As there were two distinct groups of control and experimental receiving placebo and treatment respectively and only their posttest scores were compared, and as the sample was not chosen randomly, the study enjoyed a posttest- only quasi experimental design.
3.5. Data analysis
In order to test the hypothesis of the study, some descriptive and inferential data analysis procedures were carried out:
* The internal consistency of the PET test before and after removing the malfunctioning items.
* The normality check of the experimental and control group’s scores on the posttest.
* An independent sample t-test to check the null hypothesis of the study.
Results and Discussions
This study was an attempt to explore the effect of scaffolding on EFL intermediate learners’ reading comprehension achievement. To meet this purpose, the following null hypothesis was formulated:
Scaffolding does not have any significant effect on the intermediate EFL learners’ reading ability.
After homogenizing the participants regarding their general English proficiency and their reading ability and providing an experimental group with scaffolding techniques and a control group with routine strategies and practices of reading, the participants were compared with respect to their reading achievement through a posttest. The following section is the detailed report of the data analyses.
4.2. Results of Data Analyses
First of all the researcher piloted the PET test to be standardized, the item facility of items is between 0.39-0.60. The item discrimination of them is between0.40-0, 55. The reliability of the test is 0.87. The correlation coefficient of the writing part is 0.9867.
As for the homogeneity of the two groups regarding their general English proficiency, the researcher had to compare the two groups’ mean scores on PET through a t-test, but prior to that the assumption of normality of distributions had to be checked. The following table shows the descriptive statistics of the scores obtained by both groups (control and experimental) on the PET test:
Table 4.1. Descriptive Statistics of scores obtained by both groups on PET
Valid N (listwise)
As it is shown in table 4.1, the skewness ratio belonging to the experimental group exceeds the normality range of ±1.96; therefore the assumption of normality of distribution was not met for conducting a t test. The following graphs represent the distribution of the scores:
Figure 4.1: Histogram representing the scores obtained by the control group on the PET test
Figure 4.2: Histogram representing the scores obtained by the experimental group on the PET test
As the normality assumption for a t test was not met, the non-parametric equivalent Mann Whitney U test was run to compare the two means. The following tables show the result:
Table 4.2: Mean Ranks of both groups on PET test
Sum of Ranks
Table 4. 3: Test Statisticsaof PET scores
Asymp. Sig. (2-tailed)
a. Grouping Variable: grouping
As shown in table 4.3, the difference between the rank means of the two groups was not significant (M=1123.5, p=.973.05).
Also, the means of the two groups on the reading part of the PET had to be compared to make sure that they were not significantly different with respect to their reading ability prior to the treatment. First, the normality assumption for t test had to be checked. The following table shows the descriptive statistics of the scores:
Table 4.4:Descriptive Statistics of reading scores of both groups before treatment
reading pretest control
reading pretest experimental
Valid N (listwise)
As displayed in table 4.4, the distribution of the scores belonging to the experimental group was not normal as the skewness ratio is larger than 1.96. The following graphs represent the distributions visually: