Table 3.3 The Classification of Graded Score Likert scales quoted from Setiyadi, 2006:59
3.5.4 Observation checklist
To support research, I also made an observation checklist. It contained some statements which reflected
the students‟ behavior in the class during the research. There were five statements which described the condition of the class. They were
about the presence of the students, the enthusiasm of the students in teaching and learning
process, the activity of the students during the treatments, the students‟ participation during the treatments, and also the students‟ understanding about the
instruction and the material. For the detail of the instrument, see Appendix 6. 3.6
Try-Out Test
The quality of data, whether it is good or bad, is based on the instrument to collect the data. A good instrument must fulfill two important qualifications, valid and
reliable.
Range of
Mean Student’s
Interest The
Advantages Student’s
Achievement The
Relevancy Difficulty
0.00- 1.00
Low Do not help
the students Low
Not relevant
Very Necessary
1.01- 2.00
Medium Help the
students Medium
Relevant Necessary
2.01- 3.00
High Help the
students very much
High Very
Relevant Not
Necessary
Before the test was used as an instrument to collect the data, it had been tried out to the students in another class. I carried out the try out to 36 students of
VIII B class of SMP Negeri 3 Petarukan in the academic year of 20142015 on January 15, 2015.
After scoring the result of the try-out, I analyzed to find out the validity, reliability, index of difficulty, and index of discrimination of the items of the try
out test. All of them were used to decide which items that would be used in collecting the data. Since the tests were oral tests, I analyzed the reliability of the
test by using inter rater reliability.
3.6.1 Content Validity
A test is said to have high content validity if each item, which is used to gather the data, has relevance to established criteria or objectives and covers representative
material. Brown 2005: 221 explains that test is a representative sample of the content the test is designed to measure. “To the extent that the analysis accords
with the views of recognized authorities in the skills area and the test then reflects such an analysis, it may be said to have content validity” Brown, 2004: 21. To
find out whether the test had content validity, I compared the test items with the
materials dealing with tense concord.
Therefore in making a test, I had to provide the information about the specific materials or skills being tested based on the recent curriculum namely a
School-based Curriculum KTSP. It was mentioned that the eighth-grade students are demanded to be able to express monologue texts in descriptive and recount
format to interact with others. Referring to the recent curriculum, the tests I made were valid.
To find out whether the test had content validity, I compared the test with the materials dealing with tense concord, referred to the School-based curriculum
2006:
Standard Competence Basic Competence
4. Speaking Expressing meaning in oral
texts functional and simple short monologue form of
descriptive and recount to interact with the surrounding
environment 4.2 Expressing meaning in simple
short monologue by using a variety of harsh language accurately,
fluently, and thanks to interact with the environment in the form of
descriptive text and recount.
Table 3.4 Standard Competence and Basic Competence According to school-based curriculum KTSP, the eighth-grade students
are demanded to be able to express monologue texts in descriptive and recount type to interact with others. The instruction of the try out test was
“Describe the following picture orally by using your own words in 3-5 minutes with your
partner”. Therefore, based on the school-based curriculum KTSP, the test had the appropriate content validity.
3.6.2 Inter Rater Reliability