EXPERIMENTAL DESIGN TO ANALYZE THE EFFECT OF WRITING FORMAT ON STUDENT READING COMPREHENSION ON PAPER MEDIA USING COMPLETELY RANDOMIZED FACTORIAL DESIGN

Abstract

Paper writing with the existing writing format causing excessive paper usage. Excessive paper usage could lead to environmental destruction. Therefore, it is necessary to find an alternative in the writing procedure by doing an experimental design based on student reading comprehensive. This study uses Completely Randomized Factorial Experimental Design. After analysis of variance and the multiple comparisons test done, it results a new writing format alternative that use less paper than the existing writing procedure. In other hand, it results greater student reading comprehensive than the existing writing procedure. With the results, the application of the alternative procedure can save the paper usage and make student easier to understand the paper.

Full text

JOURNAL OF ENGINEERING AND MANAGEMENT IN
INDUSTRIAL SYSTEM VOL. 6 NO. 1 YEAR 2018
e-ISSN 2477-6025
DOI 10.21776
Cite this Article As Yuwana Adam, R., & Hadi, Y. (2019). EXPERIMENTAL DESIGN TO ANALYZE THE
EFFECT OF WRITING FORMAT ON STUDENT READING COMPREHENSION ON PAPER MEDIA
USING COMPLETELY RANDOMIZED FACTORIAL DESIGN. Journal of Engineering and Management
in Industrial System, 6(2), p95-104. Retrieved from https://jemis.ub.ac.id/index.php/jemis/article/view/299
Paper Submitted : Apr, 13th 2018
Paper Published : Okt, 11th 2018 95
THE EFFECT OF WRITING FORMAT ON STUDENT READING
COMPREHENSION ON PAPER MEDIA USING COMPLETELY
RANDOMIZED FACTORIAL DESIGN
1)Rexy Satrya Yuwana Adam, 2)Yuswono Hadi
1), 2) Industrial Engineering, Faculty of Science and Technology, Universitas Ma Chung
Abstract Paper writing with the existing writing format causing excessive paper usage. Excessive paper
usage could lead to environmental destruction. Therefore, it is necessary to find an alternative in the writing
procedure by doing an experimental design based on student reading comprehensive. This study uses
Completely Randomized Factorial Experimental Design. After analysis of variance and the multiple
comparisons test done, it results a new writing format alternative that use less paper than the existing writing
procedure. In other hand, it results greater student reading comprehensive than the existing writing procedure.
With the results, the application of the alternative procedure can save the paper usage and make student easier
to understand the paper.
Keywords: excessive paper usage, writing format, completely randomized factorial experimental design
1. Introduction
In this digital era, study activities in
University still can’t avoid paper usage. For
study activities, a student will face hundreds of
assessments that need a lot of paper, especially
when a student facing internship report and
essay. A student needs approximately one rim or
more of paper in the process of internship report
or essay writing. This shows that paper writing
with the existing writing format leads to
excessive paper usage. Excessive paper usage
causes so many effects to the environment, start
from the paper production waste to the paper
after use waste. Not only the waste, but also its
needs of tree as its base material. A five years old
tree only for one rim paper production.
Therefore, it’s clearly needed an alternative for
paper writing format that use less paper.
From figure 1, it is shown that writing
format in text book use less paper than writing
format in essay. In essay format, the text height
is 2,72 cm, in text book format, the text height is
1,18 cm. Using text book format leads paper
usage saving up to 56,61%. Only by changing
the format it could decrease the paper usage.
The objective of this research is to provide
a new writing format that not only use less paper
but also could results greater reading
* Corresponding author. Email :
411410013@student.machung.ac.id
Published online at http://Jemis.ub.ac.id
Copyright ©2018 JTI UB Publishing. All Rights Reserved
comprehension for student. This research will
use Completely Randomized Factorial
Experimental Design.
Experimental design is conducted to get
responses from the objects in a certain condition
that designed by the experiment conductor. In
experimental design, there are three important
things, objects response, given treatment, and
environmental condition that could affect the
experimental results [1].
2. Research Methodology
There are some steps in this research.
Those steps arranged systematically in a
flowchart. Figure 2 is the flowchart of this
research.
2.1 Experimental Design
Experimental design started with
determining desired responses and will be
followed with determining factors and levels and
what kind of experimental design that will be
used in this research.
2.1.1 Determining Desired Responses
This research desired responses are the
level of reading comprehension. Reading
comprehension only occurred if there is
interaction between a reader and a text. In order
to understand a text, the reader have to read the
text first [2]. Therefore, appropriate text
choosing is needed. This research objects are

JOURNAL OF ENGINEERING AND MANAGEMENT IN
INDUSTRIAL SYSTEM VOL. 6 NO. 2 YEAR 2018
e-ISSN 2477-6025
DOI 10.21776
96
Fig 1. Comparison Between Essay Format in Universitas Ma Chung and Text Book Format
Fig 2. Text Conformity Test by Fry Graphic
active students of Universitas Ma Chung, which
is start from the first year students up to nearly
graduated students. In order to make sure all
students capable understanding the text, this
research use Fry Graphics to check if the text
suitable or not. Fry Graphics could determine
text readability from the sum of syllables and
sentences in the text [3]. There are some
procedures to use Fry Graphics [3] :
1. Choose 100 words from a text
2. Count the sum of sentences in the chosen 100
words, if the last word ends in the middle of
a sentence, it will be counted in decimal
forms
3. Count the sum of syllables in the chosen 100
words, if the text in Bahasa Indonesia, times
it by 0,6. The number 0,6 shows the
comparison between syllable in Indonesian
words and syllable in English words [4].
4. Look the Fry Graphic, in the graphic, line y
shows the average number of sentences every
100 words, line x shows the average number
of syllables every 100 words.
In this research, a text titled “Museum Nasional
Start
Experimental Design
Determining Sample Size
Randomization and Experiment
Floorplan Making
Reading
Comprehension Test
Validity and
Reliability Test
Normality Test
Statistical Model and
Hypothesis Making
Hypothesis Test
Conclusion
End

JOURNAL OF ENGINEERING AND MANAGEMENT IN
INDUSTRIAL SYSTEM VOL. 6 NO. 2 YEAR 2018
e-ISSN 2477-6025
DOI 10.21776
97
156,33 ; 4.0
Indonesia” is taken from a website [5]. This is
the calculation of sentences and syllables sum in
“Museum Nasional Indonesia” text.
Table 1. Calculation of Sentences and Syllables in
“Museum Nasional Indonesia” Text
100 words
Sentences
Syllables
1
3,36
158
2
4,92
155
3
3,73
156
Average
4,0
156,33
After got the results above, it will be continued
by testing the text conformity using Fry Graphic
as fig 3.
Based on the results above, it is shown
that text “Museum Nasional Indonesia” is
appropriate for at least High School Student on
Grade 12. It means the desired experiment
objects will be able to understand the text.
2.1.2 Determining Factors, Levels, and
Experimental Design Method
Reading comprehension is affected by the
text legibility [2]. Text legibility is affected by
font type, font size, line spacing, and line length
[3]. In the font type, the most readable font type
is Serif type, this font has little leg in the lower
part that looks like connecting each character.
This can improve the reading speed and reduce
eye strain [6]. In the font size, many publishers
said that size 10 or 11 is the smallest size that can
be used for printing a book [3]. Line spacing is
the vertical distance between lines in a text.
spacing, there are single, 1.5, double and
multiple [7].
In this research, three factors are chosen,
there are font size, line spacing, and line length.
Font type isn’t included because the existing
writing format already use serif font type. For the
line length, paper size used to control the line
length. After determining the factors, it will be
continued with determining the levels of each
factors. For font size, three levels are chosen, 10,
11, and 12. For line spacing, two levels are
chosen, single and 1.5 spacing. For the paper
size, two levels are chosen, A4 and A5. Based on
the factors and levels chosen, this research will
use 2x3x2 Completely Randomized Factorial
Experimental Design. Therefore, there are 12
combination that will be used as the treatment of
this experiment:
1. A = Combination of paper size A4, font size
12, and line spacing 1.5.
2. B = Combination of paper size A4, font
size 10, and line spacing 1,5.
3. C = Combination of paper size A4, font
size 12, and line spacing 1.
4. D = Combination of paper size A4, font
size 11, and line spacing 1,5.
5. E = Combination of paper size A4, font
size 11, and line spacing 1.
6. F = Combination of paper size A4, font
size 10, and line spacing 1.
7. G = Combination of paper size A5, font
size 10, and line spacing 1,5.
8. H = Combination of paper size A5, font
size 11, and line spacing 1.
9. I = Combination of paper size A5, font
size 12, and line spacing 1,5.
10. J = Combination of paper size A5, font
size 10, and line spacing 1.
11. K = Combination of paper size A5, font
size 11, and line spacing 1,5.
12. L = Combination of paper size A5, font
size 12, and line spacing 1.
Fig 3. Text Conformity Test by Fry Graphic

JOURNAL OF ENGINEERING AND MANAGEMENT IN
INDUSTRIAL SYSTEM VOL. 6 NO. 2 YEAR 2018
e-ISSN 2477-6025
DOI 10.21776
98
2.2 Determining Sample Size
Sampling is conducted if it’s impossible
to do the experiment for the whole population.
Slovin formula could help researcher to
determine the sample size of a population with
desired level of significance, here is the Slovin
Formula[8].
  
 (1)
Population that chosen in this research
consists of 1142 active students. With the level
of significance 0,1, this is the numbers of
minimum sample for this research.
  
 (2)
The numbers of minimum sample have
been determined, after that, it will be divided
into 12 difference treatments, the result becomes
the numbers of replication of each treatment.
The result is 8 students for each treatment, then
it’s rounded up into 10 students for each
treatment. Therefore, total number of students
that will be involved in this research will be 120
people.
2.3 Randomization and Experiment
Floorplan
Based on the factors and the levels, it is
known that there are 12 combinations between
factors and levels. Every combination will be
used as treatment in this experiment. Every
treatment will be replicated 10 times, so there are
120 experiment units. Then it will be continued
with randomization. In randomization, it will be
started with generating 3 digits random numbers
as many as the number of experiment units.
Then, generated random numbers will be ranked
start from smallest number up to the largest.
After that, continue it with determine the
treatment, here are the illustration of
randomization [1].
Rank in the randomization will become
the sequence number of the experiment, for
example, based on the table, rank 1 got treatment
C, then the first object should receive treatment
C [1]. After do the randomization, the floorplan
of this research become like Table 3.
Table 2. Illustration of Randomization
Random Numbers
Rank
Treatment
652
4
A
899
6
A
370
2
B
478
3
B
232
1
C
674
5
C
Table 3. Experiment Floorplan
1
2
3
4
5
61
62
63
64
65
D
D
E
K
I
B
H
L
I
D
6
7
8
9
10
66
67
68
69
70
B
G
E
G
E
D
A
J
K
C
11
12
13
14
15
71
73
73
74
75
H
G
H
A
A
F
I
K
E
A
16
17
18
19
20
76
77
78
79
80
F
E
D
J
G
H
K
J
I
B
21
22
23
24
25
81
82
83
84
85
J
C
H
L
A
C
L
A
B
C
26
27
28
29
30
86
87
88
89
90
L
E
F
K
I
C
E
L
C
G
31
32
33
34
35
91
92
93
94
95
F
B
B
A
G
D
A
K
J
D
36
37
38
39
40
96
97
98
99
100
K
G
B
J
C
C
I
D
C
H
41
42
43
44
45
101
102
103
104
105
L
E
K
C
F
L
J
G
I
J
46
47
48
49
50
106
107
109
109
110
I
D
L
J
E
F
L
I
K
F
51
52
53
54
55
111
112
113
114
115
F
G
A
I
F
A
G
D
B
H
56
57
58
59
60
116
117
118
119
120
L
H
F
B
J
E
H
K
B
H

JOURNAL OF ENGINEERING AND MANAGEMENT IN
INDUSTRIAL SYSTEM VOL. 6 NO. 2 YEAR 2018
e-ISSN 2477-6025
DOI 10.21776
99
2.3 Reading Comprehension Test on Student
of Universitas Ma Chung
Reading comprehension test use text
“Museum Nasional Indonesia”, each object will
receive the same text but with different writing
format. After finish reading the text, the object
will get 10 questions to test their understanding
of the text, the answer will be written in a sheet
of paper and will be given to the conductor after
the test session. Reading comprehension test will
be conducted based on the experiment floorplan
in previous chapter. The test was conducted in
April 2018, with the objects are active student of
Universitas Ma Chung start from the first year
student to the last year student. Table 4 is the
result of reading comprehension test.
Table 4. Result of Reading Comprehension Test
No.
Treatment
Score
No.
Treatment
Score
No.
Treatment
Score
1
D
72,5
41
L
75
81
C
85
2
D
67,5
42
E
80
82
L
75
3
E
80
43
K
75
83
A
92,5
4
K
82,5
44
C
62,5
84
B
70
5
I
72,5
45
F
60
85
C
77,5
6
B
65
46
I
77,5
86
C
85
7
G
82,5
47
D
77,5
87
E
75
8
E
80
48
L
85
88
L
70
9
G
60
49
J
100
89
C
70
10
E
80
50
E
80
90
G
100
11
H
72,5
51
F
65
91
D
80
12
G
82,5
52
G
100
92
A
92,5
13
H
72,5
53
A
75
93
K
100
14
A
65
54
I
77,5
94
J
82,5
15
A
82,5
55
F
60
95
D
90
16
F
57,5
56
L
67,5
96
C
65
17
E
80
57
H
72,5
97
I
65
18
D
77,5
58
F
85
98
D
90
19
J
75
59
B
70
99
C
70
20
G
82,5
60
J
100
100
H
67,5
21
J
92,5
61
B
75
101
L
60
22
C
77,5
62
H
72,5
102
J
100
23
H
77,5
63
L
75
103
G
100
24
L
65
64
I
72,5
104
I
85
25
A
82,5
65
D
80
105
J
100
26
L
57,5
66
D
90
106
F
67,5
27
E
72,5
67
A
90
107
L
70
28
F
85
68
J
57,5
109
I
77,5
29
K
82,5
69
K
100
109
K
82,5
30
I
65
70
C
77,5
110
F
75
31
F
65
71
F
67,5
111
A
92,5
32
B
75
73
I
85
112
G
82,5
33
B
85
73
K
100
113
D
72,5
34
A
82,5
74
E
80
114
B
62,5
35
G
100
75
A
90
115
H
77,5
36
K
100
76
H
90
116
E
90
37
G
100
77
K
82,5
117
H
90
38
B
70
78
J
100
118
K
75
39
J
92,5
79
I
72,5
119
B
57,5
40
C
85
80
B
85
120
H
90

JOURNAL OF ENGINEERING AND MANAGEMENT IN
INDUSTRIAL SYSTEM VOL. 6 NO. 2 YEAR 2018
e-ISSN 2477-6025
DOI 10.21776
100
2.4 Validity and Reliability Test
After do the reading comprehension test,
it is necessary to test the instrument, is it valid
and reliable to obtain the desired responses or
not.
2.4.1 Validity Test
Validity means the accuracy of an
instrument to obtain desired responses [9]. In
order to determine validity of instrument, it can
be done by using Biserial Point Correlation. This
is the formula for Biserial Point Correlation: [10]
  
 
 (3)
After got the result of Biserial Point
Correlation (table 5), the result will be compared
to r table, if the biserial point correlation result
smaller than the r table, so the instrument is not
valid. This is the result of validity test based on
Biserial Point Correlation.
From the result above, it can be concluded
that all questions in this experiment instrument
is valid, which means that the instrument is
accurate to obtain desired responses.
2.4.2 Reliability Test
Reliability means the consistency of an
instrument to obtain desired responses.
Instrument is reliable if the results is consistent
for repeated measurements [9]. This is formula
K-R 21 to determine the reliability of an
instrument [11]:
  

  (4)
After the result obtained, then it will be
continued by comparing the result with the
reliability level categories table 6 [11].
In this experiment, the reliability level of
the instrument is :
  

 
 
 
   (5)
By the result, it shows that the instrument
reliability is 0,513 which categorized moderate.
2.5 Normality Test
Normality test is needed to find the
distribution of this experiment responses data.
One of the method is Kolmogorov Smirnov test.
In this method, data is compared with the
standardized normal distribution data. If the
experiment results data have the same
distribution, it can be concluded that the
experiment results data is distributed normally
[12].
In this experiment, the normality test use
Kolmogorov Smirnov by using Minitab 17. Fig
4 is the result of normality test.
Based on the result, it shows that KS value is
0,081, it is smaller than KS critical value 0,123.
In the other hand, the p value is 0,053. The KS
value is smaller than the KS critical value and
the p value is larger than the level of
significance, this shows that the data of student
reading comprehension is normally distributed.
Table 5. The Results of Biserial Point Correlation
Problem no -
1
2
3
4
5
6
7
8
9
10
Correct
96
95
92
96
94
98
92
98
100
94
Wrong
24
25
28
24
26
22
28
22
20
26
Mi
80,23
80,45
81,44
80,86
80,77
81,40
80,57
80,99
80,43
80,58
Mx
79,17
79,17
79,17
79,17
79,17
79,17
79,17
79,17
79,17
79,17
Sx
11,57
11,57
11,57
11,57
11,57
11,57
11,57
11,57
11,57
11,57
P
0,80
0,79
0,77
0,80
0,78
0,82
0,77
0,82
0,83
0,78
rpbis
0,184
0,22
0,36
0,29
0,26
0,41
0,22
0,33
0,24
0,23
r tabel
(df=118)
0,15
0,15
0,15
0,15
0,15
0,15
0,15
0,15
0,15
0,15
Results
Valid
Valid
Valid
Valid
Valid
Valid
Valid
Valid
Valid
Valid
Table 6. The Level of Reliability Categories
r11
Category
0,8 ≤ r11 ≤ 1,00
Very Reliable
0,6 ≤ r11 ≤ 0,79
Reliable
0,4 ≤ r11 ≤ 0,59
Moderate
0,2 ≤ r11 ≤ 0,39
Low
0,00 ≤ r11 ≤ 0,19
Very Low

JOURNAL OF ENGINEERING AND MANAGEMENT IN
INDUSTRIAL SYSTEM VOL. 6 NO. 2 YEAR 2018
e-ISSN 2477-6025
DOI 10.21776
101
2.5 Statistical Model
The basic statistical model for factorial
design which consist of three factors (A,B,C)
and use completely randomized design as the
experimental environment is as follow [13]:
   
   (6)
For this experiment, there are three
factors, with two levels on the first and third
factor and three levels on the second factor, here
is the statistical model:
   
 (7)
2.6 Hypothesis
Hypothesis that will be tested in this
experiment are:
1. H0 : (abc)ijk = 0, which means there is no
effect of interaction between paper size, font
size, and line spacing to the objects reading
comprehension.
H1 : at least one (abc)ijk ≠ 0, which means
there is an effect of interaction between paper
size, font size, and line spacing to the objects
reading comprehension.
2. H0 : (ab)ij = 0, which means there is no effect
of interaction between paper size dan font
size to the objects reading comprehension.
H1 : at least one (ab)ij ≠ 0, which means there
is an effect of interaction between paper size
dan font size to the objects reading
comprehension.
3. H0 : (ac)ik = 0, which means there is no effect
of interaction between paper size and line
spacing to the objects reading
comprehension.
H1 : at least one (ac)ik ≠ 0, which means there
is an effect of interaction between paper size
and line spacing to the objects reading
comprehension.
4. H0 : (bc)jk = 0, which means there is no effect
of interaction between font size and line
spacing to the objects reading
comprehension.
H1 : (bc)jk at least one ≠ 0, which means there
is an effect of interaction between font size
and line spacing to the objects reading
comprehension.
5. H0 : ai = 0, which means paper size is not
affect objects reading comprehension.
H1 : at least one ai ≠ 0, which means paper
size affect objects reading comprehension.
6. H0 : bj = 0, which means font size is not affect
objects reading comprehension.
H1 : at least one bj ≠ 0, which means font size
affect objects reading comprehension.
7. H0 : ck = 0, which means line spacing is not
affect objects reading comprehension.
H1 : at least one ck ≠ 0, which means line
spacing affect objects reading
comprehension.
2.7 Analysis of Variance
In order to make an analysis of variance,
it should be started first by calculating the sum
of squares. This are the formula to calculate the
sum of squares [13]:
      








 (8)
 


 
 (9)
 


 
 (10)
 


 
 (11)
  
  

 


   (12)
  
  

 


  
 (13)
  
  

 


  
 (14)
  
   

 




 
    (15)
After do the calculation of sum of squares,
then it will be continued with make the analysis
of variance table, table 7 is the table of analysis
of variance.
2.8 Hypothesis Test
1. H0 : (abc)ijk = 0, which means there is no
effect of interaction between paper size, font
size, and line spacing to the objects reading
comprehension.
H1 : at least one (abc)ijk ≠ 0, which means
there is an effect of interaction between paper
size, font size, and line spacing to the objects
reading comprehension.
Based on the analysis of variance, the result
of F for interaction between factor ABC is
1,69 and the F table is 3,07 for level of
significance 0,05. Therefore F results < F
table, it shows that there is no effect from
interaction between three factors.
2. H0 : (ab)ij = 0, which means there is no effect
of interaction between paper size dan font
size to the objects reading comprehension.
H1 : at least one (ab)ij ≠ 0, which means there
is an effect of interaction between paper size
dan font size to the objects reading

JOURNAL OF ENGINEERING AND MANAGEMENT IN
INDUSTRIAL SYSTEM VOL. 6 NO. 2 YEAR 2018
e-ISSN 2477-6025
DOI 10.21776
102
comprehension.
Based on the analysis of variance, the result
of F for interaction between factor AB is
19,94 and the F table is 3,07 for level of
significance 0,05. Therefore F results > F
table, it shows that there is effect from
interaction between two factors.
3. H0 : (ac)ik = 0, which means there is no effect
of interaction between paper size and line
spacing to the objects reading
comprehension.
H1 : at least one (ac)ik ≠ 0, which means there
is an effect of interaction between paper size
and line spacing to the objects reading
comprehension.
Based on the analysis of variance, the result
of F for interaction between factor AC is 0,04
and the F table is 3,92 for level of
significance 0,05. Therefore F results < F
table, it shows that there is no effect from
interaction between two factors.
4. H0 : (bc)jk = 0, which means there is no effect
of interaction between font size and line
spacing to the objects reading
comprehension.
H1 : (bc)jk at least one ≠ 0, which means there
is an effect of interaction between font size
and line spacing to the objects reading
comprehension.
Based on the analysis of variance, the result
of F for interaction between factor BC is 1,06
and the F table is 3,07 for level of
significance 0,05. Therefore F results < F
table, it shows that there is no effect from
interaction between two factors.
5. H0 : ai = 0, which means paper size is not
affect objects reading comprehension.
H1 : at least one ai ≠ 0, which means paper
size affect objects reading comprehension.
Based on the analysis of variance, the result
of F for interaction between factor A is 8,457
and the F table is 3,92 for level of
significance 0,05. Therefore F results > F
table, it shows that there is effect from factor
paper size on objects reading comprehension.
6. H0 : bj = 0, which means font size is not affect
objects reading comprehension.
H1 : at least one bj ≠ 0, which means font size
affect objects reading comprehension.
Based on the analysis of variance, the result
of F for interaction between factor B is 3,072
and the F table is 3,07 for level of
significance 0,05. Therefore F results > F
table, it shows that there is effect from factor
font size on objects reading comprehension.
7. H0 : ck = 0, which means line spacing is not
affect objects reading comprehension.
H1 : at least one ck ≠ 0, which means line
spacing affect objects reading
comprehension.
Based on the analysis of variance, the result
of F for interaction between factor C is 5,911
and the F table is 3,92 for level of
significance 0,05. Therefore F results > F
table, it shows that there is effect from factor
line spacing on objects reading
comprehension.
Based on the Hypothesis Test Results, it
can be concluded that both three factors affect
the results of objects reading comprehension.
However, there is interaction between factor font
size and paper size that affect the results of
objects reading comprehension, so that the main
effects for both paper size and font size ignored.
The main effect of paper size and font size
ignored or become invalid because there is
interaction effect between them that make their
main effect not consistent.
Therefore, to determine the best treatment for
factor A and factor B, it should be based on the
interaction effect. For factor C, the best
treatment can be determined from the main
effect result.
Based on the interaction plot of factor font
size and paper size, it is clearly shown that there
are different effect when 1 level of factor A
combined with 1 level of factor B. In addition,
the best reading comprehension results is
obtained in treatment paper size A5 and font size
10, and both of them are use less paper.
Therefore, alternative format for paper size is A5
and for the font size is 10.
Based on the main effects plot, it is shown
that in line spacing 1.5, that treatment could
obtain a high reading comprehension result.
Therefore for alternative format in line spacing
is 1.5 spacing.
According to the results above, we can
decide the best alternative format that can results
greater reading comprehension and use less
paper than the existing format.

JOURNAL OF ENGINEERING AND MANAGEMENT IN
INDUSTRIAL SYSTEM VOL. 6 NO. 2 YEAR 2018
e-ISSN 2477-6025
DOI 10.21776
103
Table 7. Analysis of Variance Table
Variance
Source
Sum of
Squares
Degree of
Freedom
Middle
Squares
F Results
F Table
α=5%
Treatment
6041,67
11
A
775,21
1
775,21
8,457
3,92
B
563,23
2
281,61
3,072
3,07
C
541,88
1
541,88
5,911
3,92
AB
3655,10
2
1827,55
19,937
3,07
BC
193,44
2
96,72
1,055
3,07
AC
3,33
1
3,33
0,036
3,92
ABC
309,50
2
154,75
1,688
3,07
Error
9900,00
108
91,67
Total
15941,67
119
Fig 5. Interaction Between Factor Font Size and Paper Size Plot
Fig 6. Main Effects of Line Spacing Plot
Based on the interaction effects plot, we
can choose the paper size A5 and font size 10,
because their interaction effects results the best
comprehension score. In addition, paper size A5
and font size 10 use less paper than the other.
For the line spacing, we can decide the
best line spacing based on the main effects plot,
which shows that line spacing 1.5 results the best
comprehension score. Although line spacing 1.5
not results the least paper usage, it can be
tolerated, because great reading comprehension
is the first objective, and this alternative format
still use less paper than the existing format
Therefore the alternative format is paper size A5,
font size 10, and line spacing 1.5.

JOURNAL OF ENGINEERING AND MANAGEMENT IN
INDUSTRIAL SYSTEM VOL. 6 NO. 2 YEAR 2018
e-ISSN 2477-6025
DOI 10.21776
104
3. Conclusion
The conclusion that can be obtained from
this research are as follows:
1. Based on Analysis of Variance results, three
factors is confirmed have effect on student
reading comprehension. In addition, there is
interaction between factor font size and
paper size. Therefore the main effect on
factor font size and paper size is ignored.
2. Based on the interaction plot and main
effect plot, we can conclude that the
alternative writing format which have great
reading comprehension level is paper size
A5, font size 10, and line spacing 1,5. In
addition, this writing format results less
paper usage than the existing format (paper
size A4, font size 12, line spacing 1,5).
References
[1] Gaspersz V. Metode Perancangan Percobaan.
Bandung: Armico; 1991.
[2] Pardo LS. What Every Teacher Needs to
Know about Comprehension. International
Reading Association. 2004;58: 272-280.
[3] Lee YS. A Study of Reading With a Handled
Computer, Virginia: Virginia Polytechnic and
State University; 2003.
[4] Saroni N, Suwignyo H, Mudiono A. Analisis
Keterbacaan Teks pad aBuku Tematik Terpadu
Kelas V SD Berdasarkan Grafik Fry. Prosiding
Seminar Nasional KSDP Prodi S1 PGSD
“Konstelasi Pendidikan dan Kebudayaan
Indonesia di Era Globalisasi”. 2013: 157-164
[5] NN. Profil Museum Nasional Indonesia.
Available from:
https://www.museumnasional.or.id/
[Accessed 7th November 2017].
[6] Loyd J. Webdesign: typographic readability
and legibility. Available from:
https://webdesign.tutsplus.com [Accessed 7th
November 2017].
[7] Butterick M. Practical Typography: practical
typography. Available from
https://practicaltypography.com [Accessed 7th
November 2017].
[8] Kalimba E, Shukla J, Mbabasize M. Effect of
credit Management System on Financial
Performance of Development Bank in Rwanda:
Case Study of Development Bank of Rwanda.
The International Journal of Business &
Management. 2016;4(4): 521-531.
[9] Wahyuni N. Binus: uji validitas dan
reliabilitas. Available from:
https://qmc.binus.ac.id [Accessed 7th November
2017].
[10] Azwar S. Metode Penelitian. Jogjakarta:
Pustaka Pelajar; 2012.
[11] Arikunto S. Prosedur Penelitian: Suatu
Pendekatan Praktik. Jakarta: Rineka Cipta; 2010.
[12] Stephanie. Statistic how to: Kolmogorov-
Smirnov goodness of fit test. Available from:
http://statistichowto.com [Accessed 15th May
2018].
[13] Montgomery DC. Design and Analysis of
Experiments 8th edition. Tempe: John Wiley &
Sons; 2013.