The Normality Test RESEARCH FINDING
The Reading Test Data Analysis
Before the data analysis is described, the writer would like to re-emphasizes the hypothesis of the research. The hypothesis are described as below:
H :
There is a significant difference of students‟ reading achievement after they are taught using Think-Aloud.
H
1
: There is no a significant difference of students‟ reading achievement after they
are taught using Think-Aloud. To know which hypothesis is accepted, the writer analyzed the data using
Independent T-test, and the result is shown in the table below.
Table 4.7 The Independent T-test
Post Test Score t
df Sig. 2-
tailed Mean
Difference Std. Error
Difference Equal variances
assumed 3.212
68 .002
8.257 2.570
Equal variances not assumed
3.212 66.386 .002
8.257 2.570
From the table above, it can be seen that the score of sig. is 0.02. The value if sig2-
tailed is lower than α=0.05. Thus, since the value of sig. is lower than α 0.020.05 H
1
is accepted. It indicates that there is a significance differences between control class and experiment class in the post-test. Afterwards to know how
far Think-Aloud influenced the experiment class, the gain difference is conducted to both classes. The gain differences are shown in the table below:
Table 4.8 The Gain Differences Test Result
The Experiment Class The Control Class
Pretest Post-test
Pretest Post-test
52.54 62.68
47.37 54.60
Gain: 0.217446 Gain: 0.137374
0.05 0.1
0.15 0.2
0.25
Experiment Gain Control Gain
The Gain Differences
Experiment Gain Control Gain
The table above shows that there is difference on both the experiment and control classes. It shows that students in the control class that is not taught by using
Think-Aloud also shows an improvement in the reading comprehension test result, and so do students in the experiment class that is taught by using Think-Aloud.
After being calculated, it shows that the gain difference on experiment class is higher than the gain difference on the control class. We can see that the gain achieved
by the control class is 0.137374, while the experiment class achieved the gain as much as 0.217446, which is higher. This result supports the independent t-test that
Think-Aloud significantly influence the reading comprehension skill on the experiment class.
The gain differences between the experiment and control class is illustrated on the bar diagram below.
Picture 4. 3 The Gain Differences on the Experiment and Control Classes The Questionnaire Analysis
Besides giving sets of reading test, to know the effectiveness of Think-Aloud toward students‟ reading skill the writer also gave sets of questionnaires namely
Metacognitive Awareness of Reading Strategies Inventory MARSI. The questionnaires aims at investigating students‟ awareness in using reading strategies
whilst they are reading the text. The detail of the question on the questionnaire will be attached on the appendix of this paper.