RESULTS AND DISCUSSION

4.1. Data Analysis Research

4.1.1 Description Average and Simpang Baku on pretest data

Before implementation process of learning about the material giving opinion in class VII by using audio-
lingual method models first performed in the second pretest to determine the initial ability of students
in the second grade class. The results of the pretest in the second class, the experimental class and a
control class in group 2 contained in appendix 7 and 8. The following shows the frequency distribution of
the value pretest on Tebel 4.1 and Table 4.2.

a. Frequency Distribution pretest Value Class Experiment group 1

Range (R) = Highest Value - Lowest Value

= 60 30

= 30

Many grade interval (K) = 1 + 3.3 log n

= 1 + 3.3 log 12

= 1 + 3.3 (1.08)

= 3.56 = 4

Many classes used interval is 4 Class Interval

(P) = R / K = 30/4 = 7.5 So long interval class set 8

table 4.1

Frequency Distribution pretest Value Class Experiment

Interval F X fx f
30 37 3 33,5 1122,25 100,5 3366,75
38 45 2 41,5 1722,25 83 3444,5
46 53 3 49,5 2450,25 148,5 7350,75
54 61 4 57,5 3306,25 230 13225
Jumlah 12 182 8601 562 27387

Based on table 4.1 above the average values obtained along with the standard deviation as follows:

1

=

= = 48,83

( )
= ()

()(5)
= ()

3236
=

s2 = 96,97

s = 9,84

b. Frequency Distribution Pretest Value Class Control

Range (R) = Highest Value - Lowest Value

= 60 25 = 35

Banyak kelas interval (K) = 1 + 3,3 log n

= 1 + 3,3 log 12

= 1 + 3,3 (1,08) = 3,56 = 4

Banyak kelas interval 6

Interval Length Grade (P) = R / K = 35/4 = 8.74. So long interval class set 9

table 4.2

Frequency distribution Value Class pretest control group 2

Interval F X fx f
25 33 5 29 841 145 4205
34 42 3 38 1444 114 4332
43 51 3 47 2209 141 6627
52 60 1 56 3136 56 3136
Jumlah 12 170 7630 456 18300

Based on table 4.2 above the average values obtained along with the standard deviation
as follows:

456
=

= = 38

2
 ( )
= ()

()(456)
= ()

s2 = 88,36

s = 9,4

The following are descriptions of achievement data to learn English students before the
implementation of the learning process on giving material Opinion in Class VII SMP Negeri 2
Langsa in the academic year 2016-2017 by using audio-lingual method models in Table 4.3 as
follows:

table 4.3

Data Description Pretest Results

No Uraian Kelas Eksperimen Kelas Kontrol

1. N 12 12

2. Mean 48,83 38

3. Standar Deviasi 9,84 9,4

4. Nilai Minimum 30 25

5. Nilai Maksimum 60 60

Based on table 4.3 above can be seen that the learning achievement of the second class, the
experimental class and control class prior to the implementation of the learning process looks not much
different that the average grade experiment was 48.83 and the average grade control is 83 with a
standard deviation of the experimental class 9.84 and standard deviation of 9.4 control class. When
viewed from the maximum and minimum values of both classes are no different. This indicates that
both classes initially capabilities are the same.

4.1.2 2 Description Homogeneity Test Data pretest

The results of calculations for the variables of the two sample groups at posttest value,
obtained:

3
S12 = 96,97 n1 = 12

S22 = 88,36 n2 = 12

Maka :

Varians _ terbesar
Fhitung =
Varians _ terkecil

96,97

88,36
1,1

To dk numerator is 1 = 12 1 = 11 and dk denominator is 12 1 = 11. For = 0,05, dk = 11 in table f

numerator is 2,69.

By comparing the value of F with values obtained Ftabel of 1.1 <2,69 meaning of F <F table, it

can be concluded that the two groups of samples are homogeneous, so that the sample used to represent

the population.

4.1.3 Description Average and Standard Deviation in Data Postes

Once implemented achievement test (post test) of the obtained data on achievement belajarbahasa English

using audio-lingual method models contained

in appendix 9 and 10. The following shows the frequency distribution postes values in Table 4.4
and Table 4.5

a. Frequency Distribution Post test Value Class Experiment

Range (R) = Highest Value - Lowest Value

= 95 60

= 35

Banyak kelas interval (K) = 1 + 3,3 log n

= 1 + 3,3 log 12

4
 = 1 + 3,3 (1,08)

= 3,56 = 4

Panjang Kelas Interval (P) = R/K = 35/4 = 8,75, Jadi panjang kelas interval ditetapkan 9

Tabel 4.4

Frequency Distribution Post test Value Class Experiment

Interval f x fx f
60 68 1 64 4096 64 4096
69 77 2 73 5329 146 10658
78 86 4 82 6724 328 26896
87 95 5 91 8281 455 41405
total 12 310 24430 993 83055

Based on table 4.4 above the average values obtained along with the standard deviation as
follows:

= 82,75

( )
=
()

()()
=
()

S2 = 80,38

s = 8,96

b. Frequency Distribution Postes Value Class Control

Range (R) = Highest Value - Lowest Value

5
 =90 55

= 35

Banyak kelas interval (K) = 1 + 3,3 log n

= 1 + 3,3 log 12

= 1 + 3,3 (1, 08) = 3,56 = 4

Panjang Kelas Interval (P) = R/K = 35/4 = 8,75 = 9

Tabel 4.5

Frequency Distribution Postes Value Class Control

Interval f x fx f
55-63 2 59 3481 118 6962
64 72 4 68 4624 272 18496
73 81 5 77 5929 385 29645
82 90 1 86 7396 86 7396
Jumlah 12 290 21430 861 62499

Based on table 4.5 above the average values obtained along with the standard deviation as
follows:

=

= = 71,75

( )
= ()

()()
= ()

= 65,66

s = 8,1

The following are descriptions of achievement data to learn English students on material giving
opinion in class VII SMP Negeri 2 Langsa in the academic year 2016-2017 with audio-lingual method
using the model in table 4.6

6
 table 4.6

Description Data Results Postes

No Uraian Kelas Eksperimen Kelas Kontrol

1. N 12 12

2. Mean 82,75 71,75

3. Standar Deviasi 8,96 8,1

4. Nilai Minimum 50 55

5. Nilai Maksimum 90 95

Based on table 4.6 above can be seen that the experimental class learning achievement higher
than the control class learning achievement. This is evident from the average value of the experimental
class 82.75> average value on a control class 1.75. This can occur because of differences in treatment of
experimental class and control class, in which the experimental class in the process of learning to use
audio-lingual method models, while the control class in the learning process using conventional
methods commonly used by teachers. Differences in average values may occur due to treatment effect
or influence learning audio-lingual method models ..

4.1.4 Test Data Normality and Homogeneity

To determine differences in average values due to the influence of audio-lingual method models
it is necessary to test a statistically significant level by using the t test that the formula contained in the
data analysis. But before the test statistic t is used first tested persaratan is test data normality and
homogeneity. To test the normality of the data used Chi square test at level = 0.05 with degrees of
freedom (df) = k-1.

table 4.7

Normality Test Data Class Experiment

Batas Luas
bawah Luas 0 tiap kls
Interval kls Z -z interval Ei Oi
59,5 -2,35 0,4906 0,0371
60 68 1,1872 1
68,5 -1,68 0,4535 0,1004
69 77 3,2128 2

7
 77,5 -1,05 0,3531 0,1867
78 86 5,9744 4
86,5 -0,43 0,1664 0,2418
87 95 7,7376 5
101,5 1,45 0,4265

Information:

Limit class = lower class boundary - 0.5

= 60 0,5 = 59,5

,,
Nilai z =
= ,
= -2,35

Luas tiap kelas interval = Luas Luas = 0,4906 - 0,4535 = 0,0371

Frekuensi Harapan (Ei) = Luas tiap kelas interval x jumlah responden = 1,1872

Mencari nilai Chi Kuadrat :

( )
=

=

(Sudjana, 1992: 273)

(,) (,) (,) (,) (,)) (,))

= ,
+ ,
+ ,
+ ,
+ ,
+ ,

= , + , + , + , + , + ,

= ,

Nilai hitung dibandingkan dengan nilai tabel yaitu 9.91< 12,59 maka data prestasi belajar siswa

kelas eksperimen berdistribusi normal.

Tabel 4.8

Uji Normalitas Data Kelas Kontrol

Luas
Batas bawah Luas 0 tiap kls
Interval kls Z -z interval Ei Oi
49,5 -2,88 0,498
50 57 0,028 0,868 2

8
 57,5 -1,88 0,4699
58 64 0,116 3,596 4
64,5 -1,05 0,3531
65 71 0,236 7,316 5
71,5 -0,22 0,0871
72 78 0,3129 9,6999 1
78,5 0,6 0,2258

Mencari nilai Chi Kuadrat:

( )
=

=

(,) (,) (,) (,) (,) (,)

= ,
+ ,
+ ,
+ ,
+ ,
+ ,

= , + , + , + , + , + ,

= ,

count value compared to the value of tables of 11.429 <12.59 then student achievement data is
normally distributed control class.

The next requirement test that is testing both classes of data homogeneity of variance calculation
results of the two groups are presented in Table 4.6 using the formula:

F hitung =

,
= ,

= 1,13

To dk numerator is n - 1 = 32-1 = 31 and df denominator is 31-1 = 30. At = 0.05, df = 31, not on

the table, then dk = 31dihitung with interpolation which is located between the dk = 40 and df = 30, so
the F table searched using the formula:

Where,

C: F table is sought

C0: Ftabel value at the beginning of the existing value

C1: The value of F table at the end of the existing value

B: dk value sought

9
 B0: dk value at the beginning of the existing value

B1: dk value at the end of the existing value

So that the value of F table with 31 horsepower obtained,

,,
C = 4,00 +
( )

= 1,84+(0,017)

= 1,83

Furthermore, by comparing the two prices is 1.13 <1.85 ie F arithmetic <F table, it can be
concluded that both groups are homogeneous sample.

4.2 Hypothesis Testing

Learning is a process of change in behavior as a result of their experience. These changes can be
demonstrated in a variety of forms such as changing the knowledge, skills and learning pristasi. Based on
research obtained by the fact that learning by learning model audio lingual method makes students
more active so that student achievement can be increased.

To determine whether there is influence of the English language support problem solving in
students after a given audio lingual learning method by using the t test.



t=

+

As for calculating the S used the following formula

( ) + ( )
S2gab =
+

Based on Table 4.4 and Table 4.6 obtained average value and standard deviation of both groups
of the study sample as follows

Mean x1 = 85,31 Mean x2 = 73,38

= 124,67 = 109,45

n1 = 32 n2 = 31

maka diperoleh nilai S adalah

(),+(),
= +

10
 ,+,
=

,
=

= 117,18

S = 10,82

Selanjutnya nilai t dihitung dengan menggunakan rumus, sehingga diperoleh:

,,
t=

, +

,
=

,

,
= ,
,

,
= ,(,)

,
= = 4,60
,

Pada taraf = 0,05, dengan dk = n1 + n2 2 = 12 + 12 2 =22 dari daftar distribusi t diperoleh t

tabel berada antara dk = 120 dan dk = 60 sehingga nilai t tabel menggunakan interpolasi yaitu:

C = + ( )

,,
= 1,67 + ( )

,
= 1,67 +
(1)

= 1,67 0,0001667 = 1, 66

Furthermore, by comparing between the two prices obtained -ttabel tcount + ttable ie -1.66
4.60 1.66 so that Ha is accepted that there is significant influence of the learning model audio lingual
method on student achievement in the material giving opinion in class VII Negeri 2 Langsa in the
academic year 2016-2017

4.3 Discussion

The results of the data analysis of the pretest that is, before the implementation of learning
both homogeneous class is the second class meiliki same initial capability, so both classes as samples can
be used for further research. With the beginning of the same capabilities that owned both classes the

11
next two classes are given different models of learning, classroom learning is done conventionally VII
and VII grade 2 study conducted with the model audio lingual method.

Data from students' mastery learning can be seen in Table 4.9 below:

Table 4.9 Summary of Results of Mastery Learning Students

Kelas Pretes Postes

Eksperimen 37,21 85,31
Kontrol 38,41 73,83
Selisih 11,48

Student achievement on the material giving opinion as a result of the treatment given for
learning model audio-lingual method these students are more daring to reveal the things she did not
know to his friends freely without any sense of shame and fear of being wrong, because the learning
model audio-lingual method this is a role for teachers and students, together doing repetition of
pengucaan sentences difficult to pronounce in on the content being taught, which is thought to enhance
students' memory. This strategy students play an active role in speaking and repeating phrases and
vocabulary that is difficult. Audio-lingual method is learning strategies through learning activities that
occur between students in a class play an active role. In this strategy the student can listen to the
teacher while pronouncing the sentence, vocabulary, grammar with correct pronunciation, and students
can repeat the words of teachers to feel satisfied and able to remember vocabulary and dialogue, while
the teachers act more as a model as an example, the facilitator who gave ease, and supervising the
conduct of repetition of the words difficult to pronounce. repetition is the guidance given by people
who know better to people who are less or do not know, for example, teachers to students, or students
who are good with other students who are less intelligent. Thus it is natural if student achievement in
giving opinion in class VII with audio-lingual method models have a significant influence.

12