35 researcher could know what the students thought about using jumbled words in
the class. The last instrument was interview. The researcher chose four participants
to conduct the interview. The researcher chose the participants randomly. There were no specific purposes of conducting the interview. It was aimed to know
further the students’ opinions toward the use of jumbled words.
F. Data Analysis Technique
The most important instrument was the students’ writings. The researcher gained the students’ writing in the preliminary study April 23, 2012, in the first
cycle April 30, 2012, and in the second cycle May 7, 2012. The preliminary study started in April 23, 2012 after the school had its mid-term test. There were
16 students in XB class. However, there was an obstacle in the research that in the preliminary study, there were four students who did not join the class. Therefore,
the researcher gained 12 students’ writing. In the first cycle, there were three students who moved to other schools and two students who did not join the class
because of a particular reason. Therefore, the researcher only gained 11 writings. In the second cycle, there were two students who did not join the class. Therefore,
the researcher only gained 11 writings in the second cycle. However, it did not influence the research result because the researcher analyzed the data from the
students who attended the class. Besides, the two students who were absent in the first cycle and second cycle were the same persons.
36 After the researcher obtained the students’ writing, she analyzed sentence
by sentence. Then the researcher gave error score in each sentence. The criteria of error used by the researcher to give error score in the students’ writing are
following: The error score: 1
The students constructed incomplete simple sentence which has no subject or verb, or both Mallery, 1944, p. 112. It is included in
sentence fragment. The error score: ½
The students constructed good sentence, but there was inappropriate tense simple past tense and misspelling that was difficult to
understand. The researcher gave error score ½ because she gave explanation to tenses, especially past tense which was related to the
topic. The error score: 0
The students constructed a simple sentence correctly. If there were other mistakes in dictions, spellings, passive voices, and articles, the
researcher did not pay attention to those since the researcher focused on the good simple sentences. If there were misspellings, but they were
still understandable, it would be acceptable. After the researcher gave score in each sentence, she summed up the error
score. Then, she counted the error percentage in each student’s writing. The error percentage was counted by using formulation as the following:
37
Note: X
: the percentage mistakes in students’ composition N
: the number of wrong simple sentences that the students made N
: the sum of simple sentences that the students made In order to figure out that the research was successful or not, the researcher
counted the average of error percentage from the students’ writing in XB class in the preliminary study until cycle 2. After that, she filled it in the table to see the
improvement of the students’ ability to construct simple sentences. Therefore, the researcher could see whether the average of error percentage of the students’
writing decreased or not. There were two criteria of success in this research.
More than half of the students improved their ability to construct simple sentences that could be shown from the error percentage of the students’
writing. The average of error percentage of the students’ writing decreased.
G. Research Procedure