The population of this research is year-8 students of SMP Kartika XVII-01 Sungai Raya.

Population itself
is defined as the largest participants or groups of sample being chosen (Cohen et al., 2007). In this grade,
the students have been divided into two classes which are class A and class B. class A consists of 18 students
while class B consists of 17 students.

The sample of the research is twenty students among the population. Cohen et al. (2007) states that sample
is a group or a subset of the population as the representative of the total. The writer in selecting the sample
of the research uses simple random sampling as the sampling technique. The writer took ten students each
class as the sample of the research and thus twenty students become the sample of the research.

This research was conducted in SMP Kartika XVII-01 which is located in Jalan Adisucipto, Sungai Raya-
Kubu Raya West Kalimantan. This research was conducted on Friday, 14 November 2018 to Monday, 24
November 2018. On Friday, November 14th 2018 the writer did the pretest. On Monday, November 17th
2018 the writer did treatment 1. On Friday, November 21st the writer did treatment 2. On Monday,
November 24th the writer did the posttest.

In collecting the data of the research, the writer used performance assessment. According to Blerkom
(2007), performance assessment is an alternative assessment technique used when it is inappropriate to do
assessing through paper-based or pencil tests. The performance assessment is in form of reading aloud
monologue text. The students have to read aloud in front of the writer 3 simple texts consisting 4 to 5
sentences each. Each text has 5 words with sound /ʃ/.

Tools in collecting the data are pronunciation test through monologue text, sound recorder and scoring note.
These tools were used for the administered pretest and posttest. The tests are 3 numbers or 3 short monolog
texts. Each text contains 5 words which sound /ʃ/ appears in. The writer consistently puts 5 words in each
text for the writer expects to test the students’ articulation of sound /ʃ/ by their consistency. The students
are tested by how consistent they are in pronouncing sound /ʃ/ in slightly different cases. The consideration
to use term “consistency” is based on five rating scales category in pronunciation rubric used by Rui Ma
(2015) in his research over speaking test ratings cited from Educational Testing System.

The writer used sound recorder to record students’ pronunciation when they were being tested. The actual
purpose of using sound recorder is to get valid result of the test. Moreover, sound recorder enables the
writer to recheck the result for the sake of data analysis.

The writer also used a scoring note to help the writer score the students’ pronunciation. The scoring note is
in form of a table consists the names of the students being tested. The scoring note could ease the writer in
managing the test result before it is actually transferred to the table of pretest and posttest for further
analysis.

The writer analyzed the test items administered as the pretest and posttest as to maintain the validity and
the reliability. Below is the specific explanation:

The writer provides a specification table to know the content validity of the test items. According to Heaton
(1988), validity is the extent to which it measures what it is supposed to measure. Therefore the writer
provides a specification table of the administered test as to maintain its validity. The following specification
is for both pretest and posttest.

Reliability refers to the consistency of a test score. That is how consistent test scores results are from one
measurement to the other. According to Creswell (2012), reliability concerns range in which an experiment
reaches the same result upon repeated implementation. Cohen (2007) adds that reliability in quantitative
research is essentially similar in meaning to dependability, consistency and replicability over time, over
instruments and over groups of participants. Thus, through SPSS, the writer analyzed the result of the try-
out to make certain whether the administered test instruments are reliable or not. The following is the
guidelines that can be used (Cohen, 2007, p. 506):

Based on the results of Cronbach’s Alpha in the table of reliability analysis, obtained 0.384 which is
qualified as average. Therefore the administered test instruments have average reliability.

Level of difficulty is to find out how easy or difficult test items based on students’ point of view who
already took the test. The level of difficulty (LD) of the test instruments was calculated by the writer
using SPSS. The following is the criteria of level difficulty:

The writer analyzed the data of the administered pre-test and post-test through statistical analysis. Since
pronunciation is a sort of performance assessment, the writer needed to use a scoring rubric as to make the
test more objective (Blerkom, 2007). Thus, the writer firstly scored the students’ performance using five
scales category as already used by Rui Ma (2015). The specification is as follow:

Table 10. Five Scales Category of Pronunciation Scoring

Category Consonants Pronunciation

5 Pronounces consonants correctly all the time.
4 Pronounces consonants correctly most of the time.
3 Makes inconsistent consonant errors.
2 Pronounces some consonants incorrectly consistently.
1 Consonant errors are frequent.
After scoring the individuals, the writer analyzed the score through t-test and effect size formula. The writer
used t-test to answer the first question. In calculating t-test, the writer used the following formulas proposed
by Siregar (2015) as to find out the correlation of pretest and posttest, mean scores before and after the
treatment, variance before and after the treatment and standard deviation of pretest and posttest. The writer
then calculated the correlation of pre-test and post-test. The students’ mean score of pre-test (X) and post-
test (Y) were calculated using the following formula: The variance score of pre-test and post-test were
calculated by these following formulas: The result of standard deviation of pre-test and post-test will be
calculated by these following formulas: The writer then calculated the correlation coefficient score. The
coefficient correlation (r) was calculated through this following formula: The writer used effect size formula
to answer the research question number two. An effect size itself identifies the strength of the conclusion
about differences or the relationships among variables in quantitative studies (Creswell, 2009). The formula
of effect size is as follows.

The pre-test was held on November 14th, 2018. The pre-test was given before conducting the treatment to
the year-8 students of SMP Kartika XVII-01 Sungai Raya. The objective of the test is to find out the
students’ prior knowledge of pronunciation of sound /ʃ/ before the treatment was applied. In the pre-test,
each student was asked to read aloud 3 short monologue texts in front of the writer and they were being
recorded. Allocation time for the test was 60 minutes.

The treatment was applied two times, on November 17th and November 21st, 2018. The treatment was the
implementation of tongue twister technique in teaching the students pronunciation of sound /ʃ/. In
implementing the treatment, the students were involved in a drilling activity of tongue twister containing
sound /ʃ/. The writer firstly explained the students material such as giving example of monologue texts
about travelling and teaching them how to say it using WH questions. The writer emphasized phrases and
words related to sound /ʃ/. After being involved in monologue texts, students were asked to do tongue
twister drill. All twenty students had to stand in a line to recite 5 tongue twisters one-by-one. The students
had to return to the back of the line once they had done reciting one tongue twister. The process was
continuously repeated.

The post-test was the last step of the research. It was held on November 24th, 2018. In the pre-test, each
student was also asked to read aloud 3 short monologue texts in front of the writer and being recorded.
The test is the same in terms of the tested words and the numbers, yet the rests are slightly changed by the
writer. The allocation time for the posttest was 60 minutes.

This research was conducted to find out whether tongue twister technique effective as well as to
what extent the effect is in teaching pronunciation of sound /ʃ/. The data of the research was
collected by giving students 3 short monologue texts to read as the administered pretest and
posttest. The participants of the research are year – 8 students of SMP Kartika XVII-01 which
were chosen using cluster random sampling. Thus, 20 students become the sample of the
research.
To collect the data, students were asked to read, with clear voice, 3 short monologue texts about
personal experience and travelling. The writer recorded the students’ pronunciation and scored
them. Then, the data was calculated to analyze their significance score of pretest and posttest and
the effect size of the treatment.
The students’ significance score in the pre-test and post-test was analyzed by using t-test
formula. It is started by calculating mean score, variance score, standard deviation score and
correlation score of pre-test and post-test. The calculation is served in the following phases:

The students individual score of pre-test and post-test

The writer first analyzed the score of each student’s pronunciation of sound /ʃ/ as to know the

different score of pre-test and post-test whether there is a sign of increases or decreases of the

scores. The difference of students’ score of pre-test and post-test can be seen in the following

graph.

Students’ Mean Score of Pretest

The students’ mean score of pre-test (𝑋) was calculated as follow:

∑ Ẋ1 1,000
𝑋= = = 50
𝑁 20
 The pretest was given before the treatment. The highest score of the students’ pre-

test was 67 and the lowest score was 40. The total score of the students’ pre-test was

∑X1= 1,000 and the mean score of students’ pre-test was 𝑋 = 50

Students’ Mean Score of Posttest

The students’ mean score of pre-test (Ῡ) was calculated as follow:

∑ Ῡ1 1,006
Ῡ= = = 58
𝑁 20

The posttest was given after the treatment. The highest score of students’ posttest
was 80 and the lowest score was 40. The total score of students’ posttest was ∑X1=
1,160 and the mean score of students’ posttest was Ῡ = 58

Calculation of the Variance

The result of the variance of pre-test and post-test was calculated by these following

formulas:

(𝑋1−𝑋)² 1,006
S2x =∑ = = 52.94
𝑁−1 20−1

(𝑌1−Ῡ)² 2,970
S2y=∑ = 20−1 = 156.32
𝑁−1

Calculation of the Standard Deviation

The result of standard deviation of pre-test and post-test was calculated by these

following formulas:

∑(𝑋𝑖−𝑋̅ )2 1,006
SX√ = √20−1 = 7.27
𝑛−1

2
∑(𝑌1−Ῡ) 2,970
SY√ = √20−1 = 12.50
𝑛−1

Calculation of the Correlation Coefficient

 The result of correlation score in pre-test and post-test were calculated by these
following formulas:

The score of coefficient correlation (r) was calculated by this following

formula:

𝑛(∑ 𝑋𝑌)−(∑ 𝑋)(∑ 𝑌)

r=
√[𝑛(∑ 𝑋 2 )−(∑ 𝑋)2 ][𝑛(∑ 𝑌 2 )−(∑ 𝑌)2 ]

20(58,653)−(1,000)(1,160)
=
√[20(51,006)−(1,000)2 ][20(68,970)−(1,160)2 ]

1,173,060−1,160,000
=
√[1,020,120−1,000,000][1,379,400−1,345,600]

13,060 13,060
= =
√[20,120][33,800] √680,056,000

13,060
= 26,077.88 = 0.500

Calculation of T-test

T-test (t) was calculated by this following formula:

𝑋̅ − 𝑌̅
t =
2 𝑆 2
𝑆 𝑆 𝑆
√ 𝑥 + 𝑦 −2𝑟( 𝑋 )( 𝑌 )
𝑛1 𝑛2 √𝑛1 √𝑛2

50− 58
= 52.94 156.32 7.27 12.50
√ 20 + 20 −2(0.500)( )( )
√20 √20

−8
= 7.27 12.50
√2.64 + 7.81 − (1)( )( )
4.47 4.47

−8 −8
= =
√(10.45)−(1)(4.51) √(6.77)−(4.51)
−8
=
√(2.26)
−8
= 1.50
 = - 5.33

Analysis of The Effect Size

The formula of effect size is as follow:

1
ES = t √𝑁

1
= -5.33√20

= -5.33 x √0.05
= -5.33 x 0.22
= 1.17
Based on the computation above, the writer found out the effect size of the

treatment (ES) is 1.17. Based on the effect size criteria proposed by Cohen, Manion

and Morrison (table 5), 1.17 is categorized as strong effect (ES > 1.00). It means that

tongue twister technique is strongly effective in teaching the students’ pronunciation

of sound /ʃ/.

HYPOTHESIS TESTING

The calculation of t-test above shows that the tratio (t-test) was -5.33. The writer

determined the value of tcritical, where the significance level (α) = 0.05 because of two –

tailed test, the value of α/2 = 0.05/2 = 0.025 and the degree of freedom: df = n-1, df = 20 –

1 = 19. Then the tcritical value is t(α,df) = t(0.025,19) = 2.093 (appendix 111). The result of the

calculation shows that the tratio is higher than the tcritical (5.33 > 2.093). It proves that tongue

twister technique is effective in teaching pronunciation of sound /ʃ/ of year –8 students of

SMP Kartika XVII-01. Thus, the Null Hypothesis (Ho) which says “Tongue twister

technique is not effective in teaching students’ fricative pronunciation.” is rejected and the
 Alternative Hypothesis (Ha) which says “Tongue twister technique is effective in teaching

students’ fricative pronunciation.” is accepted.

RESEARCH DISCUSSION

The finding of this research shows that tongue twister technique has significant effect
toward pronunciation of sound /ʃ/ of the year – 8 students. Based on the data analysis, the mean
score of pre-test is 50 and mean score of post-test is 58. In this case, the students’ score has shown
a sign of different achievement between the pretest and posttest. That means there are increases on
students score after implementing tongue twister technique as the treatment.

For the t-test, the writer calculated the data and obtained t-ratio and t-critical. The
alternative hypothesis is accepted because the tratio is higher than tcritical (5.33 > 2.093) and the result
of the effect size is 1.17 (strong effect) which indicate that the use of tongue twister technique is
highly effective to increase the students’ pronunciation of sound /ʃ/.

Earlier moment, when implementing tongue twister in the classroom, the students sort of
were confused and doubtful about what to do with the tongue twisters they had to recite. Such
occurred since the students were new to tongue twister itself and some never even heard one. Even
when the writer gave them Indonesian examples, prior to any demonstration at the front, confusion
still seemed to be running in their head. Thus the writer had to really demonstrate the tongue
twisters word by word to make them understand the pronunciation. Right then, the writer
demonstrated how to play tongue twister drill and set the students’ position.

The tongue twister drill, as it was done by the students, the writer could see that the students
were enthusiastic to do the activity. Once tongue twister was given to them, for the first recitation,
they could not pronounce it completely correct. Then the writer told them to do it slowly for the
beginning and said they might recite it in fast once they got it in complete. However most of them
seemed to be uncontrolled in terms of the repetition they made. That means the students wanted to
keep reciting it until they were capable. Therefore while doing tongue twister drilling activity,
writer himself tried to control all the students so that all of them had the same opportunity to recite
tongue twisters-that is three times per one tongue twister.

In relation to the result of the pretest and posttest, it can be judged that there are changes
upon the students’ achievement. As mentioned earlier, mean score of the students’ pretest is 50 and
posttest is 58. The quality has increased about 8 (eight) points (interval). In the pretest, among 15
items in the three monologs, there are about 6 to 9 words the students did not really understand to
pronounce. Words with sound /ʃ/ in medial; vacation, sunshine and in final; fish, wish, wash, trash
seemed to be continuously incorrectly pronounced by the students. Words such as “show”,
“special”, and “station” are slightly easier for them to pronounce since they are more commonly-
used ones. However the rest of words seemed to be really difficult for them to pronounce.

Lastly, since it was only two-times implementations, it was not sufficiently strong to
increase students’ pronunciation of /ʃ/ since the pronunciation deals quite much with mouth and tongue.
Although in this research implementation students had recited certain amount of tongue twisters, it was
not easy for them to get used to pronouncing sound /ʃ/ properly. Continuous repetition in classroom
activity is needed for them to master pronunciation of /ʃ/ in complete.
After conducting the research entitled “Teaching Fricative Pronunciation through Tongue Twister

Technique” to year-8 students of SMP Kartika XVII-01, considering the result of the data analysis,

the writer has drawn the following conclusions:

1. The results of students’ mean score of pretest and posttest are 50 and 58. Meaning that,

the score increases about 8 points (interval). Even though the achievement is sort of

low, but the improvement of the quality can be categorized as average to good (Heaton,

1988, p. 96).

2. The result of t-test is 5.333. with the value of α/2 = 0.05/2 = 0.025, the degree of

freedom(df) = 20 – 1 = 19, and tcritical is = 2.093. Tratio is higher than tcritical (5.33 > 2.093)

which indicates that there is a significant difference between students’ mean score of

pretest and posttest.

3. The result of effect size is 0.994. Based on the criteria of effect size (Cohen, Manion &

Morrison, 2007), 1.17 is concluded as strong effect. This result implies that teaching

sound /ʃ/ through tongue twister technique is effective.

4. Tongue twister technique is considered an effective technique to teach pronunciation

of fricative-sound /ʃ/, particularly to the year-8 students of year-8 students of SMP

Kartika XVII-01.

1. Tongue twister technique can be considered by teachers to use for pronunciation

training especially for fricative pronunciation because tongue twister technique focuses

on sound articulation. Combination of consonantal sounds in short phrases or sentences

can help students accustom their mouth to pronouncing fricative, sound /ʃ/ in particular.
 2. For researchers conducting an experiment or to improve students’ pronunciation of

fricative-sound /ʃ/ using tongue twister, it is suggested to use and combine sounds that

are consonantally similar to sound /ʃ/, because the distraction allows errors will get

students’ tongue accustomed to articulating different sounds. Thus, they become more

capable in pronouncing every word with sound/ʃ/.

3. English teachers should make tongue twister technique an important part of their

pronunciation teaching activity and can be more creative in implementing it in their

classroom.

BIBILIOGRAPHY

Bakarić, D. N. (2016). What makes our tongues twist?: computational Analysis of croatian
tongue-twisters. Journal of American Folklore , 43–54.

Blerkom, M. L. (2009). Measurement and Statistic for Teachers. New York: Routledge.

Brown, H. D. (2001). Teaching by Principles : An Interactive Approach to Language

Pedagogy. New York: Longman.

Carmen, R. (2010). Spoken English: Flourish Your Language. Chandigarh: Abishek

Publications.

Cohen, L., Manion, L., & Morrison, K. (2007). Research Methods in Education . New York:
Routledge.

Cresswell, J. W. (2009). Research Design: Qualitative, Quantitative, and Mixed Methods

Approaches . Los Angeles: SAGE Publications, Inc.

Newman, D. (2013). Sound Partners . David Newman.

Gilakjani. A.P. (2016). How Can EFL Teachers Help EFL Learners Improve Their English
Pronunciation? Journal of Language Teaching and Research , 967-972.

Goodwin, J. (2013). Pronunciation teaching methods and techniques. The Encyclopedia of

Applied Linguistics.

Jones, D. M. (1909). The Pronunciation of English. London, UK: Cambridge University Press.

Kelly, G. (2000). how to Teach Pronunciation. England, UK: Pearson Education Limited.

Kenworthy, J. (1987). Teaching English Pronunciation. UK: Longman Group UK Limited

1987.

Kreidler, C. W. (2014). Teaching English Spelling and Pronunciation. Teachers of English to

Speakers of Other Languages, Inc. (TESOL), , 3-12.

Ladefoged, P. (2001). A Course in Phonetics/Fourth Edition. US: HEINLE & HEINLE.

O'Connor, J. D. (1980). Better English Pronunciation Second Edition. Cambridge, UK:

Cambridge University Press.

Siregar, S. (2015). Statistika Terapan Untuk Perguruan Tinggi. Jakarta: Prenadamedia Group.

Sitoresmi, U. (2016). Tongue Twister in Pronunciation Class. ICTTE FKIP , Vol 1, Nomor 1,
589 - 592.

Yollanda L. Turumi, J. S. (2016). Using Tongue Twister to Improve the pronunciation of Grade
VIII students. e-Journal of English Language Teaching Society (ELTS) , Vol. 4 No. 2
, 1 - 12.