RESEARCH METHODOLOGY
This chapter presents the procedures used in conducting the research. It covers:
research design, population and sample – including research subject, data collection and
data analysis.
3.1 Research Design
The design of this study was quasi experiment with nonrandomized pretest and
posttest.. Ary has stated (2002:316), “Designs that do not include random assignment are
known as quasiexperimental designs and are used where true experimental design are
not feasible”. The writer used quasi experiment with nonrandomized pretest and post
test because it was impossible to conduct true experiment. Besides it was difficult and
need much time to change the classification of the classes which has been determined by
the school. According to Ary (2002), in many situations in educational research, it is not
possible to conduct a true experiment. Neither full control over the scheduling of
experimental conditions nor the ability to randomize can always be realized.
To begin the experimentation the writer needs the subjects which belong to
different groups but they have similar competence, the two groups are given the pretest
at the beginning and posttest at the end.
The research design
(Nonrandomized Pretest and Posttest)
Control + _ +
Experiment + + +
Where:
+: With treatment
: Without treatment
3.2 Population and Sample
defined class of people, events or objects”. The population of this study is the eighth year
students of SMP Negeri 8 Malang. In SMP 8 Malang, there are eight classes of eighth
year with total 336 students for all. There are three teachers who teach in the eighth
grade, they are Mrs. Endang Triaridjumiati, S.Pd who teaches 8A, 8B, and 8C, Mrs. Hj.
Hariati Tinuk, S.Pd who teaches 8D and 8E, and Mrs. Sunne Sunjawati, S.Pd who
teaches 8F and 8G.
According to Ary (2002:163) “the small group that is observed called as sample.”
A good sample must be nearly representative of the entire population, so that the
generalization that is true of the sample will be true of the population. In taking the
sample of this study, the writer has given a pretest for 8A and 8C which is taught by
Mrs. Endang Triaridjumiati, S.Pd. Before the test was given to the students, the test had
been tried out in class 8B. The writer took those classes because only Mrs. Endang
Triaridjumiati, S.Pd who was willing to help the writer to conduct the experiment at SMP
Negeri 8 Malang.
test. Then, the scores were calculated to know the mean score for each class. After
knowing the mean score, the writer decided class 8C as an experimental group which was
taught by using semantic mapping technique and 8A as a control group which taught by
using the English teacher’s technique.
3.3 Data Collection
Data is an absolute requirement of research. Data are the means of proving the
hypothesis. To collect the data about the implementation of semantic mapping to improve
the student’s vocabulary in reading comprehension, the writer used test as the research
this study was the students’ score of the test.
3.3.1 Research Instrument
According to Black (1984), instrument is a process of selecting or developing
devices and method appropriate to give evaluation. The research instrument used in this
study was pretest and posttest. Referring to the title of the study, the suitable instrument
to get the data was the test of vocabulary. Both of pretest and posttest consist of 25
multiple choice items which were cumulating scored 100. The score for each correct
answer was 4 and for incorrect answer was 0.
The writer also used semantic mapping in getting the data that was conducted for
three meetings in reading comprehension subject.
These were the schedule of the study:
Treatment 1 to the experimental group by
2. Snake
using semantic mapping
Treatment 2 to the experimental group by
3. Amphibians
using semantic mapping
Treatment 3 to the experimental group by
4. Monarch Butterfly
using semantic mapping
3.3.1.1 Try Out
Before the tests were given to class 8A and 8C, the pretest and posttest had been
tried out in class 8B in order to know the validity and reliability of the instrument
items of the try out test. Then, the writer discarded 15 items which considered as the
invalid and not reliable items.
3.3.1.2 Reliability and Validity
A good test possesses reliability and validity. The reliability of a measuring
instrument is defined as the ability of the instrument to measure consistently the
phenomenon it is designed to measure (Heaton, 1983). According to Ary (2002: 249),
“reliability is the extent to which the test is consistent in measuring whatever it does
measure.” In this study, the writer used SplitHalf Reliability which is splitting the test
into two halves and correlates the individuals’ scores on the two halves. The most
common procedure, however, is to correlate the scores on the oddnumbered items of the
test with the scores on the evennumbered items. The formula is:
Where: rb = The obtained correlation coefficient
n = Number of items in the test
X = Oddnumber
Y = EvenNumber
the entire test, the obtained correlation coefficient (rb) is then entered into Spearman –
Brown formula:
2 x r b
r11 =
1 + r b
According to Sutrisno (1955:25) the entire items will have a good reliability or
not is based on the description below:
0, 80 – 1, 00 Very high
0, 60 – 0, 80 High
0, 40 – 0, 60 Middle
According to Heaton (1983), the validity of a measuring instrument is defined as the
property of a measure that allows the researcher to say that instrument measures what it is
supposed to measure. In this study, the writer used content validity. Based on O’Malley
and Pierce (1996), consideration of content or curriculum validity is particularly
important to ensure correspondence between local curriculum objectives and the content
of the assessment. The writer matched the tests material with the curriculum at SMP 8
Malang. In Reading subject, the material given by the teacher was descriptive text about
animals. Therefore, the writer used descriptive text about “Rhinoceros” for the pretest
and “Frogs” for the posttest. So that, the content of the vocabulary test was suitable with
the curriculum.
3.3.1.3 Item Analysis
To know the reliability of each items of the tryout test, the writer needed to
analyze each of those items. Therefore, the writer calculated the item difficulty and item
discrimination. Item difficulty reflects the percentage of students who answer the item
correctly. According to Oller (in Nurgiyantoro,1987:128), the formula of item difficulty
is:
FH + FL
IF =
N
Where: IF = Item Facility (level of difficulty)
FH = Frequency High Group
FL = Frequency Low group
N = the number of students
(Oller in Nurgiantoro, 1987). Item difficulty was calculated to know whether the items of
the test were too difficult or too easy for the population being tested.
According to Oller (in Nurgiantoro, 1987), to calculate the discriminating power
of each item, we can use the following formula:
FH - FL
ID =
N
Where: ID = Item discrimination
FH = Frequency High Group
FL = Frequency Low Group
N = the number of students in half
The items which discrimination index was 0.1 to 1 were included in the test (Oller in
Nurgianto, 1987). Item discrimination was calculated to compare the amount of the right
answer between the higher group and the lower group. Logically, if the amount of the
higher group who can answer the item correctly is greater than the lower group, it means
that the item is suitable to be tested.
3.3.2 Procedures of Data Collection
To collect the data, the writer conducted the following procedures:
1. Doing pretest for the experimental group and control group.
2. Collecting the result of pretest.
3. Implementing semantic mapping technique in experimental group for three
meetings.
4. Doing posttest for experimental group and control group.
5. Collecting the result of pretest.
6. Analyzing the data from pretest and posttest.
pretest and posttest), the writer used scoring criteria below based on “Buku
Pedoman Penyelenggaraan Pendidikan UMM (1991: 8)”
The Percentage of Mastery Meaning
80100 Excellent
6579 Good
5564 Fair
4554 Poor
044 Very Poor
3.5 Data Analysis
After collecting and classifying the data, the writer will analyze the data in order
to answer the statement of this study and to prove the hypothesis of this study.
The procedures of data analysis are as follows:
1. Analyzing the result of pretest.
2. Analyzing the result of posttest.
3. Combining the result of pretest between experimental and control group.
4. Combining the result of posttest between experimental and control group.
5. Calculating the result of posttest of the experimental group and control group.
The result will be analyzed by using the statistical calculation that is ttest.
According to Hatch and Farhady (1982), the procedures of the ttest formula is as
follows:
1. Counting the standard deviation of differences by using the following formula:
where: SD = the standard deviation of differences
D = Difference
n = number of pairs
2. Counting the standard error of the differences between two means by using the
following formula:
S D
S D =
n
SD = the standard deviation of differences
n = number of pairs
3. Counting the t value by using the following formula:
X 1 - X 2
t =
S D
where: t = the ttest value
X 1 = The mean score of experimental group
X = The mean score of control group
4. Counting the degree of freedom = (n1)