CHAPTER III
RESEARCH METHODOLOGY
A. Research Design
This research used quantitave approach. It was designed as an
experimental research. According to Johnson, in an experiment, the researcher’s
goal is to establish a cause-and-effect relationship between two phenomena.1
Thus, Cresswell states that experiment is test of an idea (or practice or procedure)
to determine whether it influences an outcome or dependent variable. 2
Independent variables are variables selected by the researcher to determine their
effect on or relationship with the dependent variable.3 This research was designed
as a quasi experimental research with nonequivalent control group design.
Nonequivalent control group design is one of the most widespread experimental
designs in educational research that involves an experimental group and a control
group both given a pretest and posttest, but in which the control group and the
experimental group do not have pre experimental sampling equivalence. 4 There
were two variables in this research, independent variables and dependent variable.
Independent variables are sticker stories and dependent variable was students’
writing ability on descriptive paragraph.
1
Donna M, Jhonson, Approaches to Research in: Second language Learning, New York
& London: University of Arizona, Longman, p. 165
2
Jhon W.Cresswell, Educational Research: Planning, Conducting, and Evaluating
Quantitative and Qualitative Research. (New Jersey: Pearson Education, 2008), p. 299
3
James Dean Brown, Understanding Research in Second language Learning: A
Teachner’s Guide to Statistics and Research Design, Cambridge University Press, 1988. P. 10
4
Donald T. Campbell and Julian Stanley, Experimental and Quasy Experimental Design
for Reseacrh, USA: Rand McNally and Company, 1963. P. 47
30
31
In this research, there were two classes used. One class was as an
experimental group and the rest was as control group. The experimental group
was treated by using sticker stories and control group was treated by using
conventional teachnique. Therefore, in this research, the writer used pre-test and
post-test to measure change or effect in individuals in the groups, So the design of
this research can be illustrated as follows:
Table III. 1
Nonequivalent Control Group Design
Group
Pre – test
Treatment
Post – test
E
Test 1
X
Test 2
C
Test 1
-
Test 2
E = Experimental Group
C = Control Group
T1 = Pre – Test to Experimental Group and Control Group
X = Receive the treatment using outlining technique
-
= no treatment
T2 = Post – Test to Experimental and Control Group
B. Location and Time of the Research
The research was conducted at the second grade students of SMPN 1
Ujungbatu at Sudirman Street no. 125, Ujungbatu. In this school there were two
English teachers. And in this school, English was taught twice a week with
duration 40 minutes per period. The research was done for 8 meetings, starting
from Agustus 22th until September 20th 2013.
32
C. Subject and Object of the Research
The subject of the research was the second grade students of SMPN 1
Ujungbatu. The object of this research was sticker stories and writing ability on
descriptive paragraph.
D. Population and Sample of the Research
The population of this research was the second grade students of SMPN 1
Ujungbatu in 2012-2013 academic year. It had 8 classes in which the number of
students of SMPN 1 Ujungbatu has 242 students. The specification of the
population can be seen on the table below:
Table III. 2
The Total Population of the Second Grade
Students at SMPN 1 Ujungbatu 2012-2013
No
1
2
3
4
5
6
7
8
Class
VIII 1
VIII 2
VIII 3
VIII 4
VIII 5
VIII 6
VIII 7
VIII8
Total
Total
31
31
30
30
30
30
30
30
242
The population was large enough to be taken all as sample of the
research. The writer took two classes by using simple random sampling.
Random sampling is the gold standard to which other sampling techniques
aspire5. To be a random sample, two conditions have to be satisfied6:
5
Sugiyono, Statistika untuk Penelitian. (Bandung: Alfabeta, 2011), P. 64
Jeremy Miles and Philip Banyard, Understanding and Using Statistics in Psychology.
(Los Angeles: SAGE Publication, 2007), P. 55
6
33
1. Every member of the population must have an equal chance of
being selected.
2. The selection of any one member of the population should not
affect the chances of any other being selected.
To take the sample by using simple random sampling, the writer named
cards based on every second grade students’ class in SMPN 1 Ujungbatu. After
mixing these cards, the writer took two cards randomly as a sample of this
research, so Class VIII.5 was an experimental class and VIII.6 was an control
class. Those are as the sample of the research with 60 students.
Table III. 3
The sample of the second grade at SMPN 1 Ujungbatu
No
Class
Male
Female
Total
1
VIII 5 (Experimental class)
11
19
30
2
VIII 6 (Control class)
15
15
30
Total
60
E. Technique of Collecting Data
In this research, the writer used test (pre-test and post-test) as instrument
to collect the data. As Brown stated that a test is a method of measuring a person’s
ability, knowledge or performance in a given domain.7 Test was used to find out
the students’ writing ability in descriptive paragraph. Pre test was given before the
treatment and post test was given after the treatment. In pre-test and post-test, the
7
Douglas Brown, Language Assessment: Principal and Classroom Practices
(San Fransisco:Longman ,2003), p. 3
34
students are given the same topics. The pre-test was done in order to know
students ability in writing before taught by using sticker stories and post test was
done in order to know the effect of using sticker stories toward students’ writing
ability on descriptive paragraph.
The tests was given to both experimental class and control class. The
students are required to write descriptive paragraph based on the topics provided.
Then, the students’ test was measured based on indicators of writing descriptive
paragraph. There were two raters giving scores for the students’ test. The writing
assessment of writing descriptive paragraph used by the teacher in SMPN 1
Ujungbatu as follows:
Table III. 4
ASSESSMENT ASPECS OF WRITING DESCRIPTIVE PARAGRAPH
No
1
2
3
4
5
Aspects Assessmend
1
Score
2
3
Content
Organization:
a. Description
b. Identification
Vocabulary
Grammatical Features
a. Using simple present tense
b. Using adjectives and compound adjectives
c. Using attributive has and have
d. Using linking verbs/relating verbs
Mechanics (spelling, punctuation and capitalization)
Total
Maximum Score
20
4
35
Explanation of Score:
1 = incompetent
2 = competent enough
3 = competent
4 = very competent
Final Score = Total Score X 80
Maximum Score
According to Arikunto, there are 5 components to categorize students’
writing ability. Each component has 20 as the highest score and the total of the
components are 100. In this research, the writer took 80 as the highest score.
Then the score was interpreted into the following category: 8
1. 80 – 100 = A (Very good)
2. 66 – 79 = B (Good)
3. 56 – 65 = C (Enough)
4. 40 – 55 = D (Less)
5. 30 – 39 = E (Bad)
F. Validity and Reliability of the Test
In order to know the validity of writing ability test, the writer used
content validity. Content validity is partly a matter of determining if the
content that the instruments contains is an adequate sample of the domain of
content it is supposed to represent.9 Thus, the test was given based on the
material studied by the students. The material of the test was taken from the
textbook used by the second grade students of SMPN 1 Ujungbatu.
8
Suharsimi Arikunto. Dasar-dasar Evaluasi Pendidikan : Edisi Revisi. (Jakarta: Bumi
Aksara, 2009), p. 245
9
Jack R. Fraenkel & Norman E. Wallen, How to Design and Evaluate a Research in
Education, New York: McGraw-Hill Companies Inc, 2006, p. 153
36
Reliability is the degree to which a test consistently measure whatever it
is measuring.10 The testing of students’ writing ability had to have reliability
in order to get the same scores obtained when the tests done were more than
once. There are five types of reliability: stability, equivalence, equivalence and
stability, internal consistency, and rater agreement. In this research, the writer
used the rater agreement type of reliability concerned with interpreter
reliability as the scores were given by two raters.
The writer used Pearson Product Moment formula by using SPSS 16
version to obtain the correlation between scores from rater 1 and rater 2. Then,
to know the level of the correlation, the ro is process through Spearman-Brown
Prophecy formula as follows:11
rtt =
(
Where,
,
)
,
rtt
= inter-rater reliability
n
= the number of raters whose combined estimates the final mark for
the examines
r A.B = the correlation between raters, or the average correlation among all
raters if there are more than two.
10
L.R. Gay and Peter Airisian, Educational Research Competencies for Analysis and
Application, (New Jersey: Pearson Education,2000), p.169
11
Grant Henning, A Guide to Language Testing; Development, Evaluation, and
Research, (Boston: Heinle&Heinle, 1987), p.85
37
The writer used the categories of reliability that can be seen from the
following table.
Table III. 5
The Categories of Reliability
No
Reliability
Level of Reliability
1
0.0 – 0.20
Low
2
0.21 – 0.40
Sufficient
3
0.41 – 0.70
High
4
0.71 – 1.0
Very high
(Taken from Tinambunan in Meltiawati in Zelly) 12
Table III.6
Inter-Rater Correlation
Rater
Rater 1
Rater 2
Rater 1
1.000
0.762
Rater 2
0.762
1.000
From the table above, it can be seen that ro (r
obtained)
is 0.762 will be
correlated to rt (r table). It is necessary to find the df (degree of freedom).
df = N – nr
df: degree of freedom
N: Number of cases
nr: number of correlated variable
df = 30 – 2 = 28
12
Zelly Putriani, The Correlation between Reported Speech Mastery and Speaking Ability
of the Second Year Students of SMKN 1 Pekanbaru, (Pekanbaru: Unpublished, 2011), p. 35
38
The writer took df = 28 to be correlated either at level 5% or 1%. At
level 5%, r table is 0.361; while at level 1% r table is 0.463. Thus, the r
obtained higher than r
table,
obtained
is
either at level 5% or 1%, so the writer concluded
that there is a significant correlation between score given by rater 1 and score
given by rater 2. In the other words, the writing test is reliable.
Then, it was calculated by using Spearman-Brown Prophecy Formula as
follows:
rtt =
(
( .
rtt =
rtt =
,
)
,
.
)
.
.
rtt = 0.864
Based on the data obtained above, the writer concluded that the interrater reliability in this research was 0.864 categorized as very high level.
G. Technique of Data Analysis
In order to answer this research questions, writer analyzed the data by
using‘t’ test formula thought SPSS 16 version. In the first step, the writer
obtained the mean scores of experimental class and control class. The formula
is:
39
Mx= ∑X
N
Mx
: Mean scores of Experimental class
∑X
: Total scores of experimental class
N
: Numbers of students
Then, to find the mean scores of control class the writer used the same
formula in which the scores of students in experimental class changed to
scores of students in control class.
After finding each mean score, the finding of Standard deviation was
necessary. The formula is:
SDX = √∑x2
N
Where:
SDx
: Standard Deviation of Experiment Group
∑ 2
: Total Square
N
: Number of students of experiment group
x
Standard Deviation of control class used the same formula as experiment
class; it was only changed the score of experiment class to control class.
Then, to find whether the result was significant or not, the data were
analyzed by using the formula as follows:13
t0 
13
Mx  My
2
 SDX   SDy 


  
 N 1   N 1 
2
Hartono, Statistik untuk Penelitian, (Yogyakarta: Pustaka Pelajar, 2010), pp. 207-208
40
Where:
to : the t-observation
Mx : the mean score of experimental class
My : the mean score of control class
SDx: standard deviation of experimental class
SDy: standard deviation of control class
N
: the number of cases
After calculating the t-test, to know whether the score was significant or
not, the writer should know the distinction between to and ttable. It was
necessary to obtain the degree of freedom (df) in order to get the ttable. The
formula of degree of freedom is:
df = (NE +NC) - nr
where:
df
= the degree of freedom
NE
= number of students from experiment class
NC
= number of students from control class
nr = number of variable
Finally, when the writer knew the result, the writer concluded that if to<
ttable, Ho is accepted. It means that there is no effect of using Sticker Stories
toward students’ writing ability at the second grade of SMPN 1 Ujungbatu. If
to > ttable, Ha is accepted. It means that there is a significant effect of students’
writing ability on descriptive paragraph taught by using Sticker Stories at the
second grade SMPN 1 Ujungbatu.