Sunteți pe pagina 1din 9

LUTFIZUL H S (2016121009)

BAB III
Research Method

A. The Method of The Research

In this research, the writer uses the quantitative. The quantitavie methods emphasize
objective measurements and the statistical, mathematical, or numerical analysis of data
collected. The data used in this study are interval data in form of student’s score of
Integrated Reading and Structure Material test. The research method used in this study
was a quasi-experimental method. Quasi-experimental or quasi-experimental is an
experimental method that does not allow the researcher to fully control the variables and
experimental conditionswhich random assignment to treatment groups is not used. Quasi-
experimental designs are similar to randomized experimental designs in that they
involvemanipulation of an independent variable but differ in that subjects are not
randomlyassigned to treatment groups.
The first group and the second group used the experimental method, which aims
to compare the learning achievement of students using the learning method Problem
Based Learning (conventional learning) and the learning methods of Discovery Based
Learning. This quantitative experimental research consists of three steps. The first step is
pre-test, the second step is the treatment and the last is post-test. This research needs two
groups to apply, one group as an experimental group and the other one is the control
groups.The researcher acted as the teacher in giving treatment of Inquiry Based Learning
and the researcher only gave the pre-test and posttestin control group.

Table 3.1 Non-randomized control group pretest-posttest design


Group Pre-test Treatment Post-test
Experimental Group Y1 X Y2
Control Group Y1 - Y2

Y1 : Observation in Pretest (Test was given before treatment


Y2 : Observation in Posttest (Test was given after treatment)
X : Treatment using Discovery Based Learning in the teaching Integrated
Reading and Structure Material.
B. Variables

The variables of the research consisted of independent variable and dependent


variable.

Table 3.2Variables of the Research


Independent Variables Dependent variable
Learning Models Student’s achievement on
 Discovery Based Learning Integrated Reading and Structure
 Conventional Learning (cognitive strategy) Material.

According to Ary (1985), there are three important characteristics in experimental


research,
such as:

(a) The manipulated independent variable


Manipulating variables is the actions taken by researchers on the basis of
scientific consideration. This treatment can be justified
open to get effect differences in related variables.

(b) Other influential variables are controlled to remain constant


According to Gay (1982), control of the researcher is to remove the problem
the influence of any other variable than the independent variable that ought to affect
performance on a dependent variable.
In other words, controlling is an attempt by researchers to move
the influence of other variables that might affect related variables.

C. Procedure of the Research

Data were collected through several strategies. First, the researcher taught the
class, she consulted to the teacher and asked the book that was used, then the material
was taught, the method and the problem existed in the class.
Second, the researcher came to the class and started to teach the students. The
researcher took two different classes and put them as experimental group and control
group. The researcher taught the experimental group with Discovery Based Learning, and
she taught the control group with cognitivestrategy.
Third, the researcher gave pre-test to both classes. This test was given before the
researcher taught them with the new method and software. The result of the test had been
analyzed to know whether there was any difference between teaching speaking using
Discovery Based Learning and conventional strategy.
Fourth, the researcher taught the class and gave the explanation about the material
that had been planned to be delivered clearly.
Last, the researcher gave post-test to the students. Post-test is given to the
students, after they were taught using Discovery Based Learning and conventional
strategy. The post test was held to measure the students’ ability and result in
comprehending materials which was given in the class. Finally, the researcher analyzed
the score.

D. Treatment
The treatment of design method was using quasi-experimental, we compared two
groups and each group received different treatment. The researchers gave treatment to
increase the students’ Integrated Reading and Structure Materialand the control group did
the activity like usually.

E. Subject of the Study


The subject of this study was the students in the Second grade of SMA NEGERI
2. Each class consists of about 20 students. The researcher takes two classes of many
classes in Blitar as sample of study. These are 11 A and 11 B of SMA NEGERI 2 which
become an experimental group and Control group.

F. Instrument of the Research


The study was quasi-experimental, pre-test and post-test would use pre and post-test
that were given to both control and experimental group. The score from the pre-test were
used to see that the initial ability of both groups was similiar before conducting the
treatment. On the other hand, the scores from post-test were used to measure whether
there is influence between the two variables. A good test should have a validity and
reliability.
G. Validity
Validity is the most important characteristic to consider when researcher is
constructing or selecting a test measurement technique. Ross (2005) stated that a valid
test is a test which can measure what it is intended to measure. Validity must be examined
with respect to the use which is to be made of the values obtained from the measurement
procedure. There are three important types of validity in educational research : content
validity, criterion-related validity, and construct validity.
1. Face Validity
Face validity is the appearance of the test. Face Validity is the physical appearance
of the test, which covers the format of presenting the test like a layout and
paragraphing.
2. Content Validity
Ross (2005) stated that this type of validity refers to the extent to which a test
measures a representative sample of subject-matter content and behavioral content
from the syllabus which is being measured. For example, consider a test which has
been designed to measure “Competence in Using the English Language”. In order
to examine the content validity of the test one must initially examine the subject-
matter knowledge and the behavioaral skills. They are required to complete the
test. Then after this examination, compare these to the subject-matter knowledge
and behavioral skills which are agreed to comprise correct and effective use of the
English language. The test will have high content validity if there was a close
match between these two areas.
3. Construct Validity
Ross (2005) stated that this type of validit is concerned with the extent to which
test performance can be interpreted in terms of certain psychological constructs. A
construct is a psychological quality which is assumed to exist in order to explain
some aspect of behavioer. For example, “ Reasoning Ability” is a construct. The
test must have a quality associated with individuals that can be properly called
reasoning ability. It can account to some degree for performance on the test.
4. Empirical Validity
Empirical validity means that the results of the test has a high score of correlation
compared with another test. Empirical validity is divided into predictive and
concurrent validity
To measure the empirical validity of the test, the researcher applies the factorial
analysis; where the score of every number of the test were correlated with the total
scores. The formula which was applied to correlate the validity is as follows:
𝑁 ∑ 𝑋𝑌 − (∑ 𝑋)( ∑ 𝑌)
𝑅𝑥𝑦=
√(𝑁 ∑ 𝑥 2 – (∑ 𝑥 2 )}(𝑁 ∑ 𝑌 2 − (∑ 𝑌)2 )

Where:
X = the scores of each number of the test
Y = the total scores
N = the total number of the students
Rxy = the coefficient of the correlation showing the degree of the validity of the test

Table 3.3Validity Result


Item Faktor R Xy Hargathitung Hargartabel Keputusan

I 0,969 16,6437 0.4438 Valid


2 0,963 15,1598 0.4438 Valid

According to the table above, the result of validity test is valid because all of the
scores of Rxy are >the the score of rtable.

H. Reliability
Ary (2010) stated that reliability of a measuring instrument is the degree of
consistency in which it measures it is measuring. This reliability is essential in any kind
of measurement since it has to do with the stability of the score for the same individuals.
According to Ross (2005), reliability refers to the degree to which a measuring
procedure gives consistent result. A reiable test is a test whih would provide a consistent
set of scores for a group of individuals. Reliability is a necessary but not sufficient
condition for validity. A test which provides totally inconsistent result cannot possibly
provide accurate information about the behavior being measured. Thus low reliability can
be expected to restrict the degree of validity that is obtained, but high reliability provides
no guarantee that a satisfactory degree of validity. Note that reliability refers to the nature
of the test scores and it is not to the test itself. Any particular test has a number of
different reliabilities, depending on the group involved and the situatuon in which it is
used. Some various ways of measuring or estimating the reliability have been formulated
by statisticians. The five ways to measure the reliability in educational research are:
1. Test-Retest Method
It is the method of measuring the reliability of the test by giving the test twice to
the same students and computing the two sets of the scores are calculated by
using correlation formula
2. Alternate or Equivalent-Form Method
This method requires two kinds of different the test but the same of balanced
weight, quality, integrity and times of conducting the test. Then the two sets of
the scores are calculated by using correlation formula .
3. Split-Half Method
This method is used by the researcher just by giving a test only once then
computing the separated scores from each half of the test. It can be the score
from odd and even number or from upper and lower number. To gain the
reliability, the researcher could apply the correlation formula .
4. Kuder Richardson Formula
It is a method of analyzing the reliability of the test based on the correlation of
internal consistency.
To measure the realiabity of the test, the writer used the split-half technique. She
correlated the first half and the second one by using person product moment formula. The
formula is as follow:
(∑ 𝑋)( ∑ 𝑌)
∑ 𝑋𝑌 −
𝑛
𝑟1⁄ 1
2 ⁄2 = (𝑥)2 (∑ 𝑦)2
√{∑ 𝑥 2 − } {𝑥 2 − }
𝑛 𝑛

Where:
X = the scores of the first half
Y = the scores of the second half
n = the total number of the students
𝑟1⁄ 1 = the coefficient of the correlation of two halves of the scores
2 ⁄2

Because the result of the correlation index got still indicate the relationship of the
two halves of the test, so in order to get reliability index of the test, the researcher used
the Spearman Brown formula:
2.𝑟 1⁄ 1⁄
𝑟1.1 = 1+𝑟1⁄2. 1⁄2
2. 2

Where:𝑟1.1 = the coefficient reliability of the test


𝑟1⁄ 1 = coefficient of correlation of two halves’ score.
2 ⁄2

Table 3.4Result of the Reliability


𝑟1.1(observed) 𝑟𝑡𝑎𝑏𝑙𝑒 at the significant level of
5% 1%
0.4438
0,931503 0.3783

From the table above, it shown that the result of r1.1 =0,931503 is bigger than the
rtable . It means that the test is reliable and applicable. Therefore, the test can be used as
instrument to collect the data.

I. Place and Time of Research

1. Place of Research
The research was carried out at SMA NEGERI 2. The subjects of the research
were the second grade students of SMA NEGERI 2 from class A and B.
2. Time of Research
This research was held out in second semester of the 2018/2019 academic year.
There were four steps in this research, preparation, implementation, analysis of
the data and research report. The details of each activity are as follows:
a. Preparation on 1st April 2019
b. Implementation
 Research Permission on 4th April 2019
 Data collection on 6thApril – 23th April 2019
c. Analysis of data on 27th April 2019
d. Research report on 13th Mei 2019

J. Schedule of Conducting Research


The time of collecting data in accordance with the permission of the Headmaster
at SMA NEGERI 2 was held onApril 2019.
Table 3.5 the Timetable of the Data Collecting
No Activities Date
1 Observing the school April 4th, 2019
2 Observing the classroom April 4th, 2019
3 Interviewing English teacher (Sharing April9th, 2019
about syllabus, lesson plan and the
test)
4 Doing pre-test for the experimental April5th, 2019
group
5 Teaching explanation text using April10th, 2019
Discovery Based Learning
for the experimental group
6 Doing pre-test for control group April 7th, 2019
7 Teaching explanation text using April 10th, 2019
cognitive strategy for the control group
8 Doing post-test for the experimental April 12th, 2019
group
9 Doing post-test for the control group April 14th, 2019
10 Analyzing the data obtained April 27th, 2019
11 Finishing the written report May 5th, 2019

K. Research Procedure
The procedures of the research can be described as follows: a) deciding the
subject of thye research, b) preparing the research instrument, c) doing pre-test, d)
teaching using discovered based learning experiment group, e) giving the test to get data
of the student’s achievement by administering the post test, f) scoring the students’
answers, and g) making the suitable technique of analyzing the data and statistical
operation to know results of both groups h) concluding remarks from the results of
analysis. It will give information whether the hypothesis (Ho) is received or refused.

L. Method of Collecting Data


In research, data collection techniques are an important factor for the success of
research. This study the researchers used a written test. Then, the researcher got the score
from the written test to measure the effectiveness of learning objectives

M. Technique of Analyzing the Data


Since non-randomized pretest-posttest control group design was used in this
research, the researcher applied ANCOVA formula to prove the hypothesis as suggested
by Pallant (2000, chapter 20: 1), who states that ANCOVA can be used when you have
two group pretest/posttest design (e.g., comparing the impact of two different
intervention, taking before and after measure of the groups). Further he states that the
pretest scores are treated as a covariate to control for pre-existing differences between
groups.
ANCOVA is also handy when a research has been unable to randomly assign the
subject to the different groups, but instead having had to use existing groups (e.g., classes
of students). As these groups may differ on a number of different attributes, ANCOVA
can be used in an attempt to reduce some of these differences (Stevens (1996: 324-327 in
Pallant 2000, chapter 20: 2). In order to gain accurate and correct data, the researcher has
calculated the data by using SPSS for window version 20.

S-ar putea să vă placă și