38
a. Validity
Acording to Fraenkel and Wallen 2008, p. 453, “Validity refers to the
appropriateness, meaningfulness, and usefulness of the inferences a researcher makes.” In other words, the questionnaire is valid if the statements of the
questionnaire are able to reveal what is claimed to measure. There are three main types of validity which are content-related evidence of validity, criterion-related
evidence of validity and construct-related evidence of validity Fraenkel and Wallen, 2008. Content-
related evidence of validity is “validity that is estimated by testing the content of the instrument by using rational analysis or personal
judgem ent” Azwar, 2001. In other words, the content that researcher wants to
test should be the representative sample. According to Fraenkel and Wallen 2008, criterion-related evidence of validity is related to the relationship between
scores obtain using the instrument and one or more other instruments or measures. It refers to the validity of the instruments if there is a relationship between two
instruments. Constructed-rela ted evidence of validity refers to “the nature of the
physiological construct or characteristics being measured by the instrument” Fraenkel and Wallen, 2008. It means the procedure of data collection whether
the variable being measured in the questionnaire underlies the current theories or not. Therefore, the instruments which are measured in the research are implied by
theoretical definition. PLAGIAT MERUPAKAN TINDAKAN TIDAK TERPUJI
39 In this research, Pearson‟s product moment correlation was used to get the
validity of the questionnaire items Azwar, 2001 and the formula is shown below.
Figure 3.3 Pearson’s Product-moment Correlation Formula
Notes: r
xy
= the Pearson‟s product-moment correlation X
= the scores of Metacognitive Learning Strategies distribution Y
= the scores of students‟ writing skills distribution N
= the number of paired Metacognitive Learning Strategies and students‟
writing skills scores subjects
b. Reliability
According to Fraenkel Wallen 2008, the reliability relates to “the
consistency of the scores obtained-how consistent they are for each individual from one set of item
to another” p. 453. It refers to the reliability of the questionnaire whether the responses of the subjects are consistent from time to
time. Furthermore, there are many procedures to test the reliability of a questionnaire, which are retest method, equivalent-forms method, and internal-
consistency methods Fraenkel and Wallen, 2008. In this research, in order to find out the reliability of the questionnaire, the researcher use Cronbrach‟s
coefficient alpha to analyse the data which is included in internal-consistency. According to Fraenkel Wallen 2008, in calculating the reliability of items in
some essay tests where it is possible to have more than one answer and that are not scored right versus wrong, the Cronbrach‟s coefficient alpha is used. In other
words, the Cronbrach‟s coefficient alpha is used to find out the reliability of an � =
Σ − Σ Σ . Σ − Σx . Σ − Σ
40 instrument that the items scores are not 1 or 0, for example the questionnaire or
the essay tests. Below is the formula of Cronbrach‟s coefficient alpha.
Figure 3.4 Cronbrach’s Coefficient Alpha Formula
Notes: r
11
= the instrument reliability k
= the number of items ∑
= the sum of the variances of the items score = the sum of the variance of the test score all k items
E. Data Analysis Technique
The aim of this research is to find out the relationship betwee n students‟
writing skills and metacognitive learning strategies in class 8A of SMP BOPKRI 1 Yogyakarta. The data analysis which was used to know the relationship between
stu dents‟ writing skills and metacognitive learning strategies in class 8A of SMP
BOPKRI 1 Yogyakarta is by using the Pearson‟s Product-moment Simple Correlation formula. Therefore, in order to analyse the data, the researcher did
normality test and linearity test as the data analysis technique.
1. Normality Test
The normality test was used to see whether the data of the variables were normality distributed or not. The normal distribution of the variables would be
tested using Kolmogorof-Smirnov test.
Figure 3.5 Kolmogorof-Smirnov Tests
� = �
� − 1 1 − ∑ �
�
� �
�
�
= � �
�
− � PLAGIAT MERUPAKAN TINDAKAN TIDAK TERPUJI
41 Notes:
D
n
= the maximum deviation F
nx
= the sample distribution function F
x
= the specific population distribution function
2. Linearity Test
The linearity test was used to see whether the relationship between the two variables was linear or not. The formula for linearity test is shown below:
Figure 3.6 Linearity Test
Notes: F
reg
= the value of the regression line N
= the sample size m
= number of predictor R
= the correlation coefficient
F. Research Procedure
In order to conduct this research, the researcher did several procedures in this research. Below is the explanation of how the research was conducted.
1. Observing the teaching and learning activities in SMP BOPKRI 1
Yogyakarta
Before conducting the research, the researcher did the observation during the teaching practice in SMP BOPKRI 1 Yogyakarta. The observation was
conducted on Wednesday, 12
th
August 2015 to record the teaching and learning activity that might help the researcher in the research.
�
���
= � − − 1
1 − � PLAGIAT MERUPAKAN TINDAKAN TIDAK TERPUJI