Research Findings
A. Research Findings
1. Data Description
The data of this research was the score of students‟ post-test. The researcher had given post test of both samples where the students were asked to create an analytical exposition text. The number of the students who were involved in the post test was 50 students. Those students were divided into two classes, 24 students for experimental group and 26 students for control group. The data of this research were students‟ score in post-test. The researcher taught writing to the students by using Guided Writing Strategy in experimental class and using conventional strategy in control class for six meetings. At the end of the meeting, the post-test was given to the students. The writing test was the same where the students were asked to make composition of analytical exposition text. In scoring the test, the researcher used Jacob‟s criteria that can be seen in Chapter II.
All of the data were analyzed to find out the maximum and minimum scores, mean score and Standard Deviation (SD) of post-test of experimental class and control class.
Table 4.1 The Score of Writing Test of Experimental Group and
Control Group
Total Standard Class
score Deviation Experimental 24
The total score of writing test of both groups was significantly different. The total score of experimental group was 1535, the highest score was 80, the lowest score was 40, and standard deviation was 11.31. On the contrary, the total score of control group was 1465, the highest score was 80, the lowest score was 39, and standard deviation was 9.32.
2. Descriptive Data Analysis
a. Experimental Class
K: 1+3.3 Log n Note: P: Interval R: Range K: number of classes
X min : 40
P: R/K
R: X max -X min
: 80-40= 40
K: 1+3.3 Log n : 1+3.3 Log 24 : 1+3.3 (1.38) : 5,554 :6
P: R/K : 40/6 : 6.667 :7
So, interval of students‟ writing score is 7. Then the interval data of experimental class post-test score can be seen in the table below:
Table 4.2 The Interval Data of Experimental Class Post Test Score
Interval
No (Students’ Writing Scores) Frequency
From the table above, it was found that students‟ writing post- test score in the experimental class in interval 40-46 there was 2 students, while the interval 47-53 there were 1 student, at interval 54-
60 there were 8 students who got score at that interval, at interval 61-
67 there were 4 students, at interval 68-74 there were 4 students who 67 there were 4 students, at interval 68-74 there were 4 students who
Graph 4.1 The Interval Data of Experimental Class Post Test Score
Table 4.3 Calculation Process of Mean and Standard Deviation of Writing Test Experimental Group
2 ∑X 2 1 ∑F 1 ∑X 1 ∑F 1 X 1 ∑F 1 X 1
F 1 X 1 X 1535
F 1 24
1 X 1 ) = 1535 = 2.356.225
Based on table and formulation above, researcher found that mean of writing test experimental group is 63.95 and standard deviation is
b. Control Class
K: 1+3.3 Log n Note: P: Interval R: Range K: number of classes
X min : 40
P: R/K
R: X max -X min
K: 1+3.3 Log n : 1+3.3 Log 26 : 1+3.3 (1.41) : 5,65 :6
P: R/K : 39/6 : 6.667 :7
So, interval of students‟ writing score is 7. Then the interval data of experimental class post-test score can be seen in the table below:
Table 4.4 The Interval Data of Control Class Post Test Score
Interval
No (Students’ Writing Scores) Frequency
From the table above, it was found that most of students‟ writing scores of post-test in the control class about 39-45, where
there were 1 students got score at that interval, at interval 46-52 there there were 1 students got score at that interval, at interval 46-52 there
Graph 4.2 The Interval Data of Control Class Post Test Score
67-73 74-80
frequency
Table 4.5 Calculation Process of Mean and Standard Deviation of Writing Test Control Group 2 X 2
F 2 X 2 X 1465
2 X 2 ) = 1367 = 2146225
2 56417 S 2 =
S 2 = 86 . 79
S 2 = 9.32
Based on table and formulation above, researcher found that mean of writing test experimental group is 56.35 and standard deviation is 9.32.
To explain more about Guided Writing Strategy in improving students‟ writing skill, it can be seen from the comparison of students‟
mean scores both experimental and control group in several indicators, such as: content, organization, vocabulary, language use, and mechanic. The calculation of those aspects can be explained as table below:
Table 4.6 The Calculation of Comparison of Means Post-test of Experimental and Control Class in Content, Organization, Vocabulary, Language Use, and Mechanic
No Aspects/ Post-test (Exp) Post-test (Con) Difference Components
4 Language Use
0.89 From the table of calculation of comparison of means post-test
and control class can be explained that: Moreover, the comparison of means post-test of experimental and control class in content, organization, vocabulary, language use, and mechanic could be seen in the following graphs:
Graph 4.3 The Calculation of Comparison of Means Post-test of Experimental and Control Class in Content, Organization, Vocabulary, Language Use, and Mechanic
Content Organization Vocabulary Language Use Mechanic
experiment class
control clas
The explanation above showed that the students‟ writing skill in both experiment and control class, particularly in aspects of content, organization, vocabulary, language use, and mechanics has really developed and the use of Guided Writing Strategy for Experimental have Higher performance than Conventional Strategy. It means that the hypothesis of this research was accepted.
3. Inferential Data Analysis
1) Prerequisite Hypothesis Testing
The prerequisite is necessary to determine whether the analysis of data for hypothesis testing can be continued or not. Some data analysis techniques demanding test prerequisite analysis. Analysis of variance requisite that data come from a population with normal distribution and group compared to homogeneous of data.
a. The normality of distribution test
Normality test had an objective to know population normal or not. In this research, to do the normality test the researcher used Kolmogrov Smirnov and Shapiro Wilk. Test was performed in SPSS test. Testing criterion and distributed normal if the data was more than 0.05. The class was normal. The summary of the result of test of normality and homogeneity of experimental group and control group is presented in the table below:
Table 4.7 The Result of Testing Normality Writing Post-Test Tests of Normality
Kolmogorov-Smirnov a Shapiro-Wilk
Kelas
Statistic
df Sig.
Statistic
Df Sig.
24 .131 Control
Nilai * Experimental .100 24 .200 .936
26 .073 *. This is a lower bound of the true significance.
a. Lilliefors Significance Correction
b. The homogeneity of variance test
To check the homogeneity of variance of the data, Levene‟s test was conducted. The result of calculate using Levene test is as
follow:
Table 4.8 The Result of Testing Homogeneity Writing Post-Test
Test of Homogeneity of Variances
guided writing strategy
Levene Statistic
Based on the table above, it can be concluded that two groups were Based on the table above, it can be concluded that two groups were