F. Technique of Data Analysis
Data analysis was the last procedure of the experimental design used by the writer.  In order to obtain the result of this study,  the data gained  w a s   analyzed
using statistical analysis. 1.  Normality Test
Normality  test  was  used  to  know  whether  the  data  came  from  the  normal distribution or not. In this study, the writer used SPSS 20 to find out the normality
of the data by following these steps: a.  Open SPSS program
b.  Input the data to the data view by first fill the variable view with Score as the score of pre-test or post-test and Class as the kind of class.
c.  Click Analyze  Descriptive Statistic  Explore d.  Drag the Score to the Dependent List and Class to the Factor List
e.  Click Plot  checklist Normality plots with test  ok The criteria of determining the normality of the test are as follows:
a.  If  L
value
was  smaller  than  L
table
L
value
L
table
,  it  means  that  the  data  were distributed normally.
b.  If  L
value
was  greater  than  L
table
L
value
L
table
,  it  means  that  the  data  were  not distributed normally.
2.  Homogeneity Test
Homogeneity  test  is  used  to  know  whether  the  data  come  from  the homogeneous variance or not. To calculate the data, the writer used SPSS version
20 as follows: a.  Open SPSS program
b.  Input the data to the data view by first fill the variable view with Score as the score of pre-test or post-test and Class as the kind of class.
c.  Click Analyze  Compare Means  One-way ANOVA d.  Drag the Score to the Dependent List and Class to the Factor List
e.  Click Plot  checklist Homogeneity of variance test  ok
The criteria of determining the homogeneity of the test are as follows: a.  If t
value
was smaller than t
table
t
value
t
table
, it means that H was accepted and H
1
was rejected. b.  If t
value
was greater than t
table
t
value
t
table
, it means that H was rejected and H
1
was accepted.
After testing the normality and the homogeneity of the data, the writer used statistical  calculating  of  t-test  to  find  out  the  difference  scores  of  the  student
s‟ achievement  learning  writing  using  PWIM  as  medium  compared  without  using
PWIM. Data processing was the step to know the result of both experimental class using PWIM as variable X and controlled class without using PWIM as variable
Y,  and  their  differences.  The  data  that  had  been  collected  from  the  pre-test  and post-test were analyzed by the following steps:
8
a.  Determining Mean of Variable X, with formula:
M
1
= the average of gained score mean of variable X ∑X  = sum of gained score variable X
N
1
= number of students variable X b.  Determining Mean of Variable Y, with formula:
M
2
= the average of gained score mean of variable y ∑X = sum of gained score variable y
N
2
= number of students variable y c.  Determining standard deviation score of variable X, with formula:
SD
1
= standard deviation of variable x
8
Anas Sudijono, Pengantar Statistik Pendidikan, Jakarta: PT. Raja Grafindo Persada, 2003, pp. 298
– 299.
∑X
2
= sum of squared gained score of variable x N
1
= number of students variable x d.  Determining of standard deviation of variable y, with formula
SD
2
= standard deviation of variable y ∑X
2
= sum of squared gained score of variable y N
2
= number of students variable y e.  Determining of standard Error of mean variable x, with formula:
SE M
1
= standard error mean of variable x SD
1
= standard deviation of variable X N = number of students
f.  Determining of standard Error of mean variable y, with formula:
SE M
2
= standard error mean of variable y SD
2
= standard deviation of variable y N = number of students
g.  Determining standard error from mean of variable X and Y, with the formula:
h.  Determining t-observation t
o
with formula:
i.  Determining t-table t
-t
in significant level 5 with degree of freedom df, with the formula:
df = N
1
+N
2
– 2 df = degree of freedom
N = Number of students
G. Statistical Hypothesis