View metadata, citation and similar papers at [Link].

uk brought to you by CORE

provided by Jurnal Universitas Islam Malang

LangEdu Journal ©2019 Universitas Islam Malang

Vol. - No. -, Januari-2020, pp.

Do Students’ Self -Confidence affect to Their Speaking Performance?

: The Correlation Study
Nasruddin Roziq Ali Kutub
English Language Teaching Study Program
Postgraduate School, Universitas Islam Malang, Indonesia
Email: kutuba252@[Link]

Abstract
This study was conducted to see the correlation between students’ self-confidece and their
speaking performance in speaking class of fourth semester students. It is intended to get the data
about whether statistically significant correlation between self-confidence and speaking
performance.
This research design was correlation method. Toward data collection technique in this
research was questionnaire and document study. The questionnaire was to measure students’ self-
confidence and documents study to observe about their speaking performance. Furthermore, the
data was analyzed by applying analysis factor to classify the factor which conducted from self-
confidence (Awareness, Capability, and Believe) and regression analysiss which aimed to see the
regression between three factors as the independent variable and students’ speaking performance
as dependent variable.
b
From regression calculation result showed that the F value 1.760 < 3.91, Sig score .157 >
0.05 and T scores (awareness 0.008, capability 0.533, believe 1.845) < 1.976. Thus, it can be
concluded that self-confidence has low significant correlate with speaking performance. Then,
based on the result, it is suggested that lecturers need the lecturers are suggested to facilitate and
activate students' linguistic intelligence by fortifying their grammatical and vocabulary in learning
process. For the future researcher this topic is quite general the researcher suggests doing deep
investigations about the relationship between self-confidence and speaking performance aims to
get details by using different statistical that such as factor analysis and multiple regressions to
obtain the information more detailed about these relationships with more specific instruments.
Therefore, the researcher also suggests to apply add spoken test as the kind of better research
instrument accompany with make students classification belongs to the level of students’
confidence and speaking performance.
Keywords: Self -Confidence, Speaking Performance

INTRODUCTION
In natural term, language is one of the important tools applied by people to
share an idea, feeling, and information in daily communication. Furthermore,
language becomes necessary that happens in the whole of humans' experience.
According to (Victoria & Robert, 1983; Pamungkas, 2018) which said that people
were apply some words to share information depend on the context and condition.
Dealing with information transformation, in Indonesia, most of the information are
transformed by using Bahasa as the first language (L1) and some of information were
transformed by English as foreign language (L2). In understanding information which
transformed in second language (English) make students need to concentrate in
understanding the information. By those reasons, make students have to learn about
 Kutub
Do Students’ Self -Confidence affect to Their Speaking Performance? : The Correlation Study

English as second language for helping them in understanding information that

transformed by English.
In learning English, students trained to master four language skills, i.e.:
speaking, reading, listening and writing. Among those language skills, speaking is one
of the skills that mostly used in their activity. However, in order to communicate
appropriately within a particular community, students need to have a communicative
competence. This competence will lead someone to use the language with a good
grammar and understandable speech to show what in her/his mind and feeling. This
competence is one of oral performance of learning a second language. Oral
performance refers to speaking skill; it attracts students to perform their ability orally
that makes speaking be challenging material in learning a language.
Speaking is one of challenging skills in language learning, in enhancing
speaking skill, students need to have to enrich enough vocabulary. It is because
vocabulary can support their fluency and accuracy during the speaking performance.
It means that the more vocabulary the students have the fluency and accuracy their
speaking will be. This is in line with (Coady, 1993; Pamungkas, 2018) stated that
vocabulary is the prior element that helps students in mastering speaking skills. This
element can help the students in transforming their idea or feeling effortless. When
they did not have enough vocabulary, it can make them difficult to understand and
share their idea. So that in the language learning process, especially in speaking class
vocabulary, it is an important thing that students should have because it builds
students' ability in transforming the information and the material deeply depend on
the context.
In the other hand, in mastering speaking, the students need to have good self-
confidence. Conceptually, self- confidence is one of the non-linguistic factors that
affect speaking. Self-confidence is from personality factor, which does not correlate
with any other language element, but it can affect the speaking. On this occasion,
(Yashima, Zenuk & Shimizu, 2004; Al hebaish, 2012) stated that self–confidence is an
necessary factor that determines learners’ readiness to practice their speaking
performances in the classroom activity. It can be assumed that the high level of
students’ self-confidence the good speaking performance they have.
There have been several studies which discussing about students' self-
confidence. The first research conducted by (Keefe & Shi, 2017) found that self-
confidence has dominant evidence to increase students’ speaking ability in group
presentation practice.
Another study by (Tello, 2012) indicated that the research process prompted
students’ speaking confidence through the use of technology. The study was also an
opportunity to improve the teacher’s performance and become a facilitator by
 LangEdu Journal ©2019 Universitas Islam Malang
Vol. - No. -, Januari-2020, pp.

steering a pedagogical intervention that allowed students to make decisions to

overcome their lack of speaking confidence in all three aspects of ability, assurance,
and willing engagement. Besides, students testified feeling mediocre of 47% more
confident at the end of the semester than the beginning of semester.
Another side, (Doqaruni, 2014) suggests that due to the regular practice of
spoken English in pair or group work activities, students found English easier to
speak. Similarly, (Vazquez, 2014) noted that to build students' self-confidence it is
come from the good partnership between parent and teacher influenced students’
self-confidence. Since, to construct their self-confidence it cannot stand from
individual factor but a combination with all factors which contribute to a building
inside or outside of the classroom.
Therefore, (Pulford & woodwaed, 2018) in their study compare female and
male students’ self-confidence in academic comparison. From these study presented
the higher level of self-confidence that female and male that they require comes from
their activeness during class activity and it also can be said that not self-confidence
do not correlated to gender.
In another side, (Tridinanti, 2018) in her study about found that self-confidence
has significant correlation with speaking achievement. Moreover self-confidence is
the core elements that encourage students and in practicing speaking English in front
of the class.
Furthermore, (Tuncel, 2015) revealed that the subjects who have high self-
confidence achieved higher success rates than the subjects who have low self-
confidence.
From the previous studies about confidence in language learning, the
researcher is eager to study the confidence that takes part to find the correlation
between self-confidence and speaking performance. Another consideration makes
the researcher choose this topic is that self- confidence affects the students’ learning.
By having self-confidence, the students can successfully execute the desired behavior.
The researcher conducted thesis entitled:
‚The Correlation between the Students’ Self -Confidence and their Speaking
Performance‚

METHOD
This research was conducted in a correlational research design as the
researcher wanted to determine which level students’ self-confidence and speaking
performance correlated. According to (Latief, 2016) the correlation research design
is used to know the relationship between two variables. Furthermore, ( Creswell, 2012)
some of the research methodologies which predict scores and explain the
relationship among variables with statistical investigation tests that called correlation
 Kutub
Do Students’ Self -Confidence affect to Their Speaking Performance? : The Correlation Study

design. Based on theory above, the researcher conclude that the correlation research
design was concerning to observe the relationships between two or more variables by
investigating the score of variables with the statistical investigation.
Furthermore, this research determined the correlation between two variables
they are students’ self-confidence as the independent variable and speaking
performance as the dependent variable.
This research was conducted at University of Islam Malang. The population of
this study are concern on fourth-semester students of the English education
department in the academic year 2018-2019. The total target population of this
research is one hundred forty-eight (148) students. They are divided into five (5)
classes of speaking class. Further,
Since the number of the population is too big, the researcher did purposive
sampling after determining two criterions from the research subject are;
a. English education department’s student
b. Following the speaking 4 class
About research instrument the researcher used two kinds of research
instruments for each variable they are questionnaire and documentation. Among the
questionaire researcher used a questionnaire containing 13 (thirteen) of questions
are written in the form of statements which need the students’ responses. Besides, the
responses are written in 5 (five) choices aimed to collect the data on the students’ self
–confidence in speaking. Moreover, the questionnaire was adapted from ( Horwitz, E.
K, Horwitz, M. B, & Cope, J, 1986) which contain 10 (ten) following statements. In this
research, the researcher took statement number 1, 3, 5, 6, 7, and 9 which applied
researcher research questionnaire in item number 1, 6, 8 9,10 and 13. Meanwhile the
researcher also did adaptation the research’s questionnaire from (Griffee, 1997) which
contain 12 (twelve) items and took number 1, 6, 9,8,12 applied to number 2, 3,4,5,7.
For further information about the detail of about the outline of questionnaire content
and the focus of statements were explained with table 3.2.

Table. 1 The Outline of Questionnaire Content

No Indicators The number of question The number of items used
1. Self –capability 1,2,3,4,5 5
2. Personality 6 1
3 Self-confidence 7,8,9,10 4
4. Students ‘ opportunity 11 1
5 Students’ expectation 12, 13 2
Total 13 13
 LangEdu Journal ©2019 Universitas Islam Malang
Vol. - No. -, Januari-2020, pp.

In accepting the validity of questionnaire, according to (Sulistiyo, 2015) that

good research instruments has to be validated. To get the satisfactorily information
from research instruments, there are several requirements to check the quality of
research instrument by checking the reliability and validity. Furthermore, to get the
validity of questionnaire in this study, the researcher did some ways. The first,
administering the question that used and getting the revision, an English expert at the
University of Islam Malang validated it more it shown at appendix no 4 at page. The
last process, it is validated by Pearson Product moment in SPSS v.20. Thus the output
of validity test shown in table
From table 3.3 known that, the T-score is higher than T-table and the level of
significant is smallest that .005. It can be said that the questioner that applied by the
researcher is Valid.
In measuring about the reliability of the questionnaire the researcher used
SPSS v20 exactly with the Alpha Cronbach Coefficient (ACR) process. From the
statistical process found that the score of ACR from questionnaire was .832 and it was
higher than the significant level at 0.05. Based on the table of classification of
reliability, it shows that the reliability of the students’ degree of extroversion
questionnaire is moderate. For further information detailed shown by the output from
SPSS v20 as follows: shown above the questionnaire was reliable and able to apply be
the research instrument and also appropriate to be displayed to the research subject.
Moreover, regarding with limitation of time, and to figure out information
about students’ speaking performance, the researcher applied a document as the
source of information. The document was taken from students’ performance with
students’ speaking score, which was collected from all of speaking score from
lecturers’ assessment process. The researcher decided to use documentation caused,
there are too many research subject, the researcher do not have capacity to conduct
to test to each students and the researcher keep the neutrality of data.
To obtain the data, the researcher did some procedures. The data were
obtained from March 25 to May 8, 2019. In term of the students’ self- confidence, the
researcher distributed a questionnaire written in English in to 148 students of
speaking class, the questionnaire was consist of 13 (thirteen) item of statements and
5 item of answer choice. In following answer from the questionnaire, the researcher
provided five options of statements and it has different score in each. Thus, the detail
of the score realised in the table 3.5.
 Kutub
Do Students’ Self -Confidence affect to Their Speaking Performance? : The Correlation Study

Table 2. List of Questionnaire Scoring

Option Statement Score
A Strongly disagree 0
B Disagree 1
C Neutral 2
D Agree 3
E Strongly agree 4

However, to get information about students’ speaking performance, the

researcher used the document study as data fulfilment of students’ speaking
performance. Additionally, the score was taken from lecturers’ speaking assessment
which taken after students has already done their performance inside of class. After
the researcher has already got the list of students’ speaking performance score, the
researcher grades the students’ score from the deepest score was 0 and the
uppermost score was 100. Further information about scoring grade was explained
with a grading score on table 3.6 lists below:

Table 3. List of students’ speaking score

No Students score Criteria
1 85-100 Very good
2 64-84 Good
3 40- 54 Moderate
4 0-39 Low

After the researcher got the data from research instruments to analyze
information which has been provided from the research instrument. There were
several techniques which researcher used.
After all of the questionnaire was collected in grading the result of
questionnaire the minimum score for each statement is 1 and the maximum score for
each statement is 5. In addition the researcher provided the table about
questionnaire scoring scale.
After the researcher found the mean, median and standard deviation score the
researcher decided the measure score from two instruments they are; questionnaire
of students’ self-confident and speaking score. After the students filled the
questionnaire, the researcher determined the students level of self-confident based
on the questionnaire score.
Moreover, the students’ self-confidence score data the questionnaire has been
analyzed then, to determine the students’ self-confidence are normal or not the
researcher did the Normality Test.
 LangEdu Journal ©2019 Universitas Islam Malang
Vol. - No. -, Januari-2020, pp.

In addition, the researcher did a normality test to find out whether the
distribution of the data about students’ self-confidence and students’ performance
are normally or not. The data was checked by using One Skewnes and Kurtosis in
SPSS. The probability value for the normality test of the data is 0.05. The variables are
normal if they:

a. accepted if the ratio score is between (-2) and (+2) it mean that the data
were normally distributed..
b. rejected when the ratio score is not between (-2) and (+2), it means the
data is norm

After the researcher found the data normality the researcher did Factor
Analysis, it was aimed to definite the significance of data structure which influenced
to the data. According Ghozali (2016) factor analysis. in Factor analysis there were
several stages, namely:

1) Testing the data normality

2) Testing the assumption of the value of Kayser Mayer Oilkin Measure of
Sampling Adequacy (KMOSA) and the score of the Barliet Test of Scarcity.
3) Determining the value of the MSA Measures of Sampling Adequacy
4) Determining the results of communal extraction
5) Determining Variance Explained
This analysis is used to examine whether or not the students self-confidence
influence students’ speaking performance.
Furthermore, students’ self-confidence represented with three factors which
taken from the factor analysis (see table 4.9) whether, if there was an influence, it
would said the best predictor of students’ self-confidence are to a positive classroom
environment and cooperative learning. Afterward, to answer why the correlation and
influence among the variables might occur, as a result, are subsequently interpreted.
The hypothesis of there is some correlation between students’ self-confidence
and students speaking performance. To test the hypothesis, the data is analyzed by
using the stastical formulations as follows:

H0= if rxly< r t- means there is correlation between students self-confidence

(X) and students speaking competence (Y) but has low significant.

Ha= if rxly > rt- means that there is corelation between students self-
confidence and students’ speaking competence
 Kutub
Do Students’ Self -Confidence affect to Their Speaking Performance? : The Correlation Study

RESULTS
As what has been the researcher presented in the previous, the research
population subject were fourth semester in one of private Islamic University in Malang
which concentrations on English Education Department which consist of 148
students.
In line with the findings of research, the researcher presented the description
statistic table which displayed the mean and standart deviation about variables. The
analysis of mean and standard deviation was conducted to understand direct
differencial score among variables
After deliberate the data, the researcher analyzed the normality and of data to
fulfill the assumption. The descriptive statistic was displayed as in the table below:

Tables 4. Descriptive Statistic

Variables N Mean Std. Deviation
Statistic Statistic Std. Error Statistic
Selfconfidence 148 35,60 ,571 6,941
Speakingcompetence 148 65,80 1,051 12,789
Valid N (listwise) 148

Table 4. Displayed (N) or the research participants are 148 students. Moreover,
the mean score of self-confidence was 35.60 and the mean score of students’
speaking performance was 65. 80. Furthermore, the standard devise’ score among
students ‘self-confidence there was 6.941. Then, speaking performance’ standard
deviation score there was 12.789. Before conducting hypothesis testing is the
normality of data. Normality used to investigate the data description.
To figure out the finding of this research, after founded the mean and standard
deviation score, the researcher did two processes to analyze the data and did
hypothesis testing. They are analysis factor and regression.
Before analyze the finding the researcher did normality test which aimed to see
the normality distribution between the speaking performance’ data and self-
confidence’ data. The normality test of the data was provided on table 4.2 and 4.3.

Table 5. Table of Skewness dand Kurtosis of Speaking performance

N Skewness Kurtosis
Statistic Statistic Std. Error Statistic Std. Error
Speaking 148 -.474 .199 1.027 .396
Valid (listwise) 148

From table 2 showed the output of the skewness score was -.474 and output
kurtosis score was 1.027. To interpret the significant level of skewness and kurtosis it
 LangEdu Journal ©2019 Universitas Islam Malang
Vol. - No. -, Januari-2020, pp.

known from the ratio of skewness and kurtosis score which located between (-2) and
(+2). After counting process, the researcher found the ratio score of skewness was -
2.38191 and kurtosis score 2.593434 in between (-2) and (+2). Those ratio score
shows that data of speaking performance is normally distributed.

Table 6. Table of Normality Skewness and Kurtosis test of Self-confidence.

N Skewness Kurtosis
Statistic Statistic Std. Error Statistic Std. Error
Self-confidence 148 -.471 .199 .880 .396
Valid N (liswise) 148

Table 3 above showed that the output of skewness score was -.471 and output
kurtosis score was .880. To interpret the significant level of skewness and kurtosis,
known from the ratio of skewness and kurtosis score which located between (-2) and
(+2). After counting process, the researcher found the rasio score of skewness -
2.36683. The kurtosis score 2.222222 in between (-2) and (+2). Those ratio score
shows that data of speaking performance is normally distributed.
Furthermore, referring to the normality test as displayed in the table above, it
can be seen that the both data were in normal distribution.
This analysis aimed to ddiscovers the significance of data structure which
influenced to the data. The analysis of factor including eight stages, namely:
1) Testing the assumption of the value of Kayser Mayer Oilkin Measure of
Sampling Adequacy (KMOSA) and the score of the Barliet Test of Scarcity.
2) Determining the value of the MSA Measures of Sampling Adequacy
3) Determining Communalities
4) Determining total Variance Explained
5) Describing scree plot
6) Determining component matrix
7) Determining rotate matrix
8) Determining component matrix
For more information shown Table of Descriptive One-Sample Kolmogorov-
Smirnov Test (Part-2) can be seen on appendix 02. The analysis of the KMO (Kaiser-
Meyer-Olkin Measure of Sampling Adequacy) and BTS (Bartlett's Test of Sphericity)
constructed to investigates the appropriateness of variables.
Table 7. Table of KMO and BTS Test.
Kaiser-Meyer-Olkin Measure of Sampling Adequacy. .815
Bartlett's Test of Sphericity Approx. Chi- 577.917
Square
Df 78
Sig. .000
 Kutub
Do Students’ Self -Confidence affect to Their Speaking Performance? : The Correlation Study

After the normality of data was conducted, in factor analysis it was needed to
see the table of Output KMO and Bartelt Test is usefull to know variable
appropriateness that aimed to consider whether the variable can be measured in
factor analysis. The whole processes are considering to KMO (Kaiser-Meyer-Olkin
Measure of Sampling Adequacy) score.
Furthermore when the sore of KMO MSA bigger than 0.50, it shows that the
analytical process can be continued. In this occasion wherein the output scores of
KMO MSA known .815 > .50 and score Bartlett's Test of Sphericity (sig). 0.00 < 0.05.
From those score, it could be known that this analysis can be continued because the
variable has already attempt to the factor analysis assumption that if the score of
Kayser Mayer Oilkin Measure of Sampling Adequasy (KMOSA) higher than .050 and
the score of Barliet Test of Spcerecity has bigger than .50
Anti-Image Matrices is used to measure the appropriate variable which can be
processed on factor analysis. In addition, Anti-image Covariance provides some code
Measures of Sampling Adequacy MSA using (a) which shows the score of MSA inside
of the table. 5 for more display at appendix no 8 shows the data below,

Table 8. Table of MSA score

Item of Question MSA Score
No
1 Statement 1 .898
2 Statement 2 .833
3 Statement 3 .876
4 Statement 4 .800
5 Statement 5 .858
6 Statement 6 .847
7 Statement 7 .895
8 Statement 8 .813
9 Statement 19 .822
10 Statement 10 .779
11 Statement 11 .818
12 Statement 12 .667
13 Statement 13 .689

Table 8 displayed the MSA score comprises of the score of statement 1 was
.898 , statement 2 was .833a, statement 3 has score .876a, statement 4 the score was
a
 LangEdu Journal ©2019 Universitas Islam Malang
Vol. - No. -, Januari-2020, pp.

.800a, statement 5 was .858a, statement 6 847a, statement 7 has MSA score.895a
statement 8 .813a, 9 .822a in statement 9, statement 10 was .779a, statement 11 .818a,
12 was .667a, and .689a, in statement 13. From those table shown that the MSA score
is higher than .50 which means that all of the variables can be analyzed by using
factor analysis. Otherwise, when MSA score of some variables lower (< 0.50) then the
re-analyze process only for variables that have an MSA value > 0.50 is conducted.
Anti-Image Matrices is used to measure the appropriate variable which can be
processed on factor analysis. In addition, Anti-image Covariance provides some code
Measures of Sampling Adequacy MSA using (a) which shows the score of MSA inside
of the table. 4.5 for more display at appendix no 8 shows the data below,

Table 9. Table of MSA Score

No Item of Question MSA Score
1 Statement 1 .898
2 Statement 2 .833
3 Statement 3 .876
4 Statement 4 .800
5 Statement 5 .858
6 Statement 6 .847
7 Statement 7 .895
8 Statement 8 .813
9 Statement 19 .822
10 Statement 10 .779
11 Statement 11 .818
12 Statement 12 .667
13 Statement 13 .689

Table 9. displayed the MSA score comprises of the score of statement 1 was
.898 , statement 2 was .833a, statement 3 has score .876a, statement 4 the score was
a

.800a, statement 5 was .858a, statement 6 847a, statement 7 has MSA score.895a
statement 8 .813a, 9 .822a in statement 9, statement 10 was .779a, statement 11 .818a,
12 was .667a, and .689a, in statement 13. From those table shown that the MSA score
is higher than .50 which means that all of the variables can be analyzed by using
factor analysis. Otherwise, when MSA score of some variables lower (< 0.50) then the
re-analyze process only for variables that have an MSA value > 0.50 is conducted.
 Kutub
Do Students’ Self -Confidence affect to Their Speaking Performance? : The Correlation Study

Table 10. Table of Comunalities

Initial Extraction

Statement 1 1,000 .482

Statement 2 1,000 .628

Statement 3 1,000 .490

Statement 4 1,000 .506

Statement 5 1,000 .556

Statement 6 1,000 .395

Statement 7 1,000 .523

Statement 8 1,000 .590

Statement 9 1,000 .654

Statement 10 1,000 .623

Statement 11 1,000 .326

Statement 12 1,000 .712

Statement 13 1,000 .786

The table above showed the score of variable examined can explain the factor
or not. The variable is considered able to explain the factor if the extraction value is>
0.50. Based on data, it can be concluded that the extraction value for all variables was
greater than 0.50. Thus it can be concluded that all variables can be used to explain
factors.
Table 11. Table of Total Variance Explained
ST Initial Eigenvalues Extraction Sums of Squared Loadings Rotation Sums of Squared Loadings
% of Cumulati Cumulatie
Total Total % of Variance Total % of Variance Cumulative %
Variance ve % %
1 4.445 34.194 34.194 4.445 34.194 34.194 2.581 19.855 19.855
2 1.754 13.496 47.690 1.754 13.496 47.690 2.464 18.950 38.805
3 1.071 8.240 55.930 1.071 8.240 55.930 2.226 17.125 55.930
4 .894 6.874 62.804
5 .783 6.024 68.828
6 .746 5.738 74.566
7 .663 5.103 79.670
8 .595 4.580 84.249
9 .552 4.249 88.499
10 .465 3.577 92.076
11 .431 3.316 95.392
12 .324 2.492 97.884
13 .275 2.116 100.000
Extraction Method: Principal Component Analysis.
 LangEdu Journal ©2019 Universitas Islam Malang
Vol. - No. -, Januari-2020, pp.

Table 7 of variance showed the value of each variable analyzed. There were 13
components analyzed and were two kinds of analysis to explain a variant, namely
Initial Eigenvalues and Extraction Sums of Squared Loadings. The Initial Eigenvalues
variant showed the factor formed wherein all factors showed the number of variables
(4.445 + 1.754 + 1.071+. 894 +. 783 + .746 +.663 +. 595 +. 552 + .465 + .431 + .324 +
.275 = 12.998 whereas in the Extraction Sums section of Squared Loadings showed
the number of variations in the number of factors that can be formed which was
including 3 variations of factors 34,194 + 13,602 = 49,930 and 36.328 + 13.496 +
8.240 = 55.930.
Based on the results of Initial Eigenvalues, there are three factors formed from
13 variables analyzed which has the requirement to be a factor which was higher than
1 follows in factor 1 = 4.445>1and able to explain 34.194, factor value 2 = 1.754>
1that explain 13.496 and factor 3= 1.071> 1 which was able to explain 8.240. Further,
all of the variations from component 1, 2 and 3 were summed, and then explain
55.930 variations.
On the hand, the components 4 until 13 were not calculated because the value
of Equivalent was lowest (<) 1. In sum, from the whole processes it could be founded
that there were three factors from those components that build in speaking
confidence data which located on component number 1 till 3.
Picture1. Picture of Scree Plot

This plot showed the number of factors formed wherein there was one point

with the value of Eigerrvalue >1. In this case, there were 3 factors that is in factor 1, 2

and 3. Moreover, the component 1 stayed at 4.5, component 2 at 1.5, and component

3 stayed at 1.20.
 Kutub
Do Students’ Self -Confidence affect to Their Speaking Performance? : The Correlation Study

Table 12. Table of Component Matrix

CONTENT Component
1 2 3
Question1 .621 .308 -.044
Question2 .637 .142 -.449
Question3 .619 -.149 -.291
Question4 .602 -.132 -.355
Question5 .628 .261 -.305
Question6 .550 -.234 -.195
Question7 .635 -.283 .199
Question8 .634 -.393 .182
Question9 .634 -.402 .301
Question10 .596 -.348 .383
Question11 .451 .257 .238
Question12 .426 .630 .366
Question13 .513 .705 .159
Extraction Method: Principal Component Analysis.
a. 3 components extracted.
Table 12. Exhibited that the component matrix that displayed the correlation
between components and factors. In addition, the table showed that the correlation
between 13 with 3 factors. Moreover, the question with factor 1 was .621, with factor
2 was .308 and factor 3 was -.044. Question 2 has correlation with factor 1 was .637,
factor 2 was .142 and factor 3 was -.044. Question 3 has correlation with factor 1 was
.619, question 2 was -.149 and factor 3 -.291. Question 4 has correlation with factor
1was .602, factor 2 was -.132 and factor 3 -.355 and question 5 has correlation score
with factor 1 was .628 factor 2 was 261 and factor 3 with -.305. In question 6 has
correlation with factor 1 been .550, factor 2was -.234 and factor 3 with -.195. Question
7 has correlation with the factor such factor 1 with .635, factor 2 with -.283 and factor
3 with .199.
Table 13. Table of Rotated Component Matrix
Component
1 2 3
Question1 .172 .432 .515
Question2 .074 .753 .235
Question3 .319 .621 .055
Question4 .265 .659 .035
Question5 .071 .637 .381
Question6 .377 .503 -.008
Question7 .661 .256 .144
Question8 .718 .268 .049
Question9 .784 .177 .088
Question10 .771 .089 .145
Question11 .244 .107 .506
Question12 .070 -.007 .841
Question13 -.029 .207 .861
Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization.
a. Rotation converged in 6 iterations.
 LangEdu Journal ©2019 Universitas Islam Malang
Vol. - No. -, Januari-2020, pp.

To make sure some variables into factorial group, it can be measured by

finding and persuiting the high correlation score of correction between variables
(Question) with research factor (Component). Furthermore, the following paragraph is
the finding of Rotation converged as;
a. In item 1, there was correlation score between factor 1 =.172, factor 2=.432
and factor 3 =.515. From the finding known that correlation score of factor 3
is higher than factor 1 and factor 2 that means item 1 was included in factor 3
group.

b. Item 2, it has correlation score in factor 1 = .074 in factor 2 = .753 and factor 3
= .235. From the score description found that score in factor 2 is higher than
factor 1 and 3. It can be said that variable 2 were included into factor 2 group.

c. Item 3 has correlation score in factor 1 there were =.319, factor 2 =.621 and
factor k 3 there were = .055. From the data description known that correlation
score of factor 3 is higher than factor 1 and 2. It can be said that, item 3 were
included in to factor 3 group.

d. Item 4 has correlation score in factor 1 for about = .265, in factor 2 with = .637,
and factor 3 = .035. After score descriptions found that the correlation score
from factor 2 is higher than factor 1 and 3. So that item 4 is companied to
factor 2 groups.

e. In item 5 known that, the correlation score in factor 1 is about = .071, then in
factor 2 = .626 and factor 3 = .681. From the data description known that
correlation score from factor 3 is higher than factor 1 and 2. It can be said that,
variable 5 was included into factor 2 groups.

f. Item 6 has correlation score for factor 1 is =.377, factor 2 = .503 and factor 3=
-.008. From the score explanation known that factor 2 is higher than factor 1
and 3. It means that item 6 is included into factor 2 group.

g. In item 7, it has correlation score which came from three factors. The
correlation score on factor 1=.611, factor 2=.256 and factor 3 it has = .144.
From the whole data explanation above, revealed that factor 1 has highest
score that factor 2 and 3. It can be said that, item 7 were included into factor 1
groups.
 Kutub
Do Students’ Self -Confidence affect to Their Speaking Performance? : The Correlation Study

h. Item 8 has correlation score about in there factors, in factor 1 =.718, factor
2=.268 and factor 3= .049. From the data, can be known that factor 1 has
highest score than other variables which included into factor 1 group.

i. Item 9 has correlation score on factor 1 = .784, factor 2=.177, and factor 3 has
= .088. After the score was explained, known that factor 1 is higher than
another factor. It can be said that item 9 was included into factor 1 groups.

j. Item 10 has correlation score in factor 1 with =.771, on factor 2 = .089 and
factor 3 with = .145. From the description known than factor 1 has higher score
than another variable and it reflected that item 10 was joining factor 1 groups.

k. Item 11 has correlation score in factor 1=.244, factor 2 =.-007 and factor 3 =
.506. From the factor analysis found that, on factor 3 has highest score than
another factor. It can be said that, item 11 is including factoring 3 groups.

l. Item 12 has correlation score likewise, factor 1 with = .070, factor 2= .014 and
= .841 on factor 3. From the score description known that factor 03 has higher
correlation score that factor 1 and 2. Means that, item 12 were included on 3
factor group.

m. Item 13 has correlation score on factor 01 about =.-029, factor 2 has sore
about =.207 and factor ke 3= .861. Because of factor 3 has the high score it can
be said that, item 13 was on factor 3 groups.
For more information about the group classification likewise;

Table 14. Table Factorial group of questionnaire’s item.

No Factor Number of questionnaire’s question item
1 1 7, 8,9,10
2 2 2,3,4,5,6
3 3 1, 11,12,13

After the researcher explained the data description in previous phase, the
researcher decided to change the factors’ name became group and the question
become an item. The classification is used to simplify the researcher and reader in
classifying the group that aims to give clear information. In designing group
classification, the researcher was considered to the most topics which concern to
discuss in the whole of group items. The classification likewise;
 LangEdu Journal ©2019 Universitas Islam Malang
Vol. - No. -, Januari-2020, pp.

A. Factor 1 changed became Awareness group

B. Factor 2 changed became Capability group
C. Factor 3 changed become Believe group

Table 15. Table of Component Transformation Matrix.

Component 1 2 3
1 .612 .642 .462
2 -.602 -.001 .799
3 .514 -.766 .386

Component transformation matrix shown that, on 01 component has

correlation score .583 > 0.5.02 component has correlation score about = .769 and
component 3 has score .504. From the explanation, it could be stated that correlation
score of whole components were higher than > 0.5. It can be said that, those three
components were appropriate to resume all of variables which examined.
Regression used to know the simulate correlation between two or more X
variables (predictor) and Y variable (criterion), the researcher used regression to know
the simulate correlation between students’ speaking confidence and their speaking
performance. In this passage the researcher displayed the finding of calculation
process in regression.

Table 16. Table of Model Summary

Model R R Square Adjusted R Std. Error of the Estimate
Square
a
1 .188 .035 .015 12.691
a. Predictors: (Constant), believe, capability, awarnes

It can be seen from the table 4.13 that R value is .188. Additionally R square is
.015which showed that the contribution of the dependent variable students’ self-
confidence(X) to students’ speaking performance (Y) was 1.5%.

Table 17. Table of F value contribution

a
ANOVA
Model Sum of Df Mean Square F Sig.
Squares
b
1 Regression 850.647 3 283.549 1.760 .157

Residual 23193.272 144 161.064

Total 24043.919 147
a. Dependent Variable: speaking
b. Predictors: (Constant), believe, capability, awarnes
 Kutub
Do Students’ Self -Confidence affect to Their Speaking Performance? : The Correlation Study

Table17. Presented the result of regression between Self-Confidence which was

comprised of awareness, capability, believe and Speaking Performance. Referring to
the result of regression analysis above, it could be seen that the sig. value is 0.157
which means it was bigger than 0.05 and F value is 1.760 lower than 3.91. In addition,
based on the Sig. and F value above, it can be stated that the regression was not
significant statistically between independent and dependent variable. It means that
there is no correlation among self-confidence and speaking performance. In the other
words, self-confidence is no affect towards speaking performance.

Table 18. Table of Contribution of Each Variable.

Coefficientsa

Model Unstandardized Standardized T Sig.

Coefficients Coefficients
B Std. Error Beta
1 (Constant) 51.634 6.401 8.066 .000
Awarnes .003 .365 .001 .008 .994
Capability .212 .397 .048 .533 .595
Believe .849 .460 .166 1.845 .067
a. Dependent Variable: speaking

Table 15 provides us with the necessary information to predict the contribution

of each independent variable by looking at the significant value column and
unstandardized coefficients B column.
T test between X (awareness) and Y shows that the t value is 0.008<1.976 and
the significance is 0.994, which bigger than .05. it means that self-confidence do not
contribute to speaking performance significantly by showing the contribution X (
awareness) to Y (speaking performance) is .003.
T test between X (capability) and Y has t value 0.533<1.976 and the
significance value is 0.595 which bigger than 0.05. It means that there is no significant
contribution between self-confidence to speaking performance by score 0.212. In
other words, when self-confidence decrease the value, speaking performance
decreases the value as the speaking performance as .212.
T test between X (believe) and Y shows that the value of t is 1.845<1.976 and
0.067 in the significance value that lower than 0.05. It means that statistically self-
confidence do not contribute to speaking performance significantly. It shows that
the contribution X (awareness) to Y (speaking performance) is .849.
For more information about the significance of correlation, it explained with
the index of correlation.
 LangEdu Journal ©2019 Universitas Islam Malang
Vol. - No. -, Januari-2020, pp.

Table 19. Table of index of correlation

No Scale Interpretation
1 0.800-1.000 High Correlation

2 0.600-0.800 Sufficient Correlation

3 0.400-0.600 Fair Correlation

4 0.200-0.400 Low Correlation

5 0.000-0.200 Very low Correlation

Equations
Afterwards, the theoretical hypothesis that assured by regression analysis
formulated as follows;
1. H1 = r×y>rt. It means that there is correlation between students’ self-
confidence (X) and speaking performance (Y).
2. H0 = r×y<rt. It means that there is no correlation between students’
self-confidence (X) and speaking performance (Y).
Based on the table of analysis above, founded that the value of coefficient
correlation (R) 0.188<177.389, F 1.760<3.91, t <1.976 and sig. value>0. Hence,
referring to those values, it could be assumed that statistically students’ self-
confidence do not correlate significantly toward their speaking performance.
Therefore, based on the result of analysis, the aforementioned null hypotheses (H1)
was rejected while the alternative hypothesis (H0) there is negative correlation
between self-confidence in speaking performance was accepted.

DISCUSSION
The main purpose of this study was to investigate the correlation between
students' self-confidence and their speaking performance. The strongest finding of
this correlation study is that the students’ self-confidence did not work on their
speaking capability. In addition, the result of the data analysis in the previous chapter
showed that the correlation between students’ self-confidence and their speaking
performance is low. The finding appeared that Linier Regression correlation result
shows that the F score 1.760 > 3.91 with the Sig. value rxly=.157b > 0.05 and the T
values for the three factors (awareness 0.008, capability 0.533, believe 1.845) < 1.976.
Based on the statistical point of view, the research finding indicated that
statistically there was no significant correlation between students’ confidence and
their speaking performance. It means that the alternative hypothesis which stated (H1)
the higher self-confident the students have the better they achieve in speaking
performance was rejected and the null hypothesis (H0) was accepted. Because the
 Kutub
Do Students’ Self -Confidence affect to Their Speaking Performance? : The Correlation Study

significance score of the correlation was lower than .005, it can be said that students’
self-confidence uncorrelated with their speaking performance. Referring to the
finding which showed there is no significant correlation between students’ self-
confidence and their speaking performance, the researcher believes there is a
possibility that the result of this study have affected by factor such as the students’
have insufficient understanding including vocabulary mastery and so forth. Sufficient
vocabulary, grammatical and having accurate memories of the features will be
beneficial toward speaking skill. On the other hand, it also could be influenced by the
duration and the students’ condition as well since working in hurry and feeling
uncomfortable have a biggest effect.
On the other hand, the researcher perhaps needs more time as well as the
duration in researching to assure the correlation between self-confidence and
speaking skill. Additionally, during the research process, the researcher has to aware
of the students’ conditions which can influence the result of research. In sort, in
running up research, all of the terms should be fulfilled.
Concerning with the finding wherein there is not significant correlation such as
displayed in the previous chapter, it could be noticed that the result of this study was
the opposite finding and contradicted with the theory of self-confidence stated by
Anderson (2013) wherein proposed that confidence is students' beliefs about their
capability in doing something well and Ibadurrohman (2016) that noted confidence is
one’s positive attitude in the learning process.

CONCLUSION
The result of this research was to know the finding from research question "Is
there any correlation between students' self-confidence and their speaking
performance?. The result indicated that there was no significant correlation between
three factors of students’ self-confidence (awareness, capability, believe) and
students’ speaking performance. It was proven by the result of Regression between
two variables which showed the F value 1.760 < 3.91, about the Significant score of F
value .157b > 0.05 and T value (awareness 0.008, capability 0.533, believe 1.845) <
1.976. In other words, it can be assumed that the students who have high self-
confidence do not affect the speaking performance. In short, the researcher
concluded that self-confidence was not correlated with speaking performance, and
otherwise.
 LangEdu Journal ©2019 Universitas Islam Malang
Vol. - No. -, Januari-2020, pp.

REFERENCES
Al-Hebaish, S. M. (2012). The Correlation between General Self-Sonfidence and
Academic Achievement in the Oral Presentation Course. Theory and Practice in
Language Studies, 2(1), 60–65. [Link]
Creswell, Jhon W. (2012). Education Research: Planning, conducting, and Evaluating,
Quantitative and Qualitative Research. New York: Pearson Education.
Doqaruni, V. R. (2014). A Quantitative Action Research on romoting Confidence in a
Foreign Language Classroom: Implications for Second Language Teachers. Inquiry in
Education: Vol. 5: Iss. 1. Retrieved from: [Link]
Ghozali, I. (2016). Aplikasi Analisis Multivariat Dengan program IBM SPSS 21.
Semarang: Badan Penerbit Universitas Diponegoro.
Hayes, A. (2016). Careers Using Languages: how to set up a conference for schools. In
E. Corradini, K. Borthwick and A. Gallagher-Brett (Eds), Employability for
languages: a handbook (pp. 51-55). Dublin:[Link].
[Link]
Ibadurrohman, M. B. (2016). Study on Students’ Lack of Confidence In Speaking Class
At English Teacher Education Department. (Unpublished Undergraduate Thesis).
UIN Sunan Ampel, Surabaya.
Keefe, K., & Shi, L. (2017). An EAP Program and Students’ Success at a Canadian
University. TESL Canada Journal, 34(2). [Link]
Latief, M. A., (2016), Research Methods On Language Learning: An Introduction.
Malang : Universitas Negeri Malang.
Pamungkas, Z. B. (2018). The Correlation between Students’ Vocabulary, Mastery,
Self-Confidence and Their Speaking Ability. Unpublished Postgraduate Thesis.
Pulford, B. D. & Woodward, B. &Taylor, E. (2018). Do Social Comparisons in Academic
Settings Relate to Gender and Academic Self-Confidence. Socio Psychol Educ,
21:677–690. [Link]
Tello, L .N.C. (2012). On Air : Participation in an Online Radio Show to Foster Speaking
Confidence. A Cooperative Learning-Based Strategies Study. Profile Vol. 14, No.
1, April 2012. ISSN 1657-0790. Bogotá, Colombia. Pages 91-112.
Tridinanti, G. (2018). The Correlation between Speaking Anxiety, Self-Confidence, and
Speaking Achievement of Undergraduate EFL Students of Private University in
Palembang. International Journal of Education & Literacy Studies, IJELS 6(4):35-
39.
Tunçel, H. (2015). The Relationship between Self-confidence and Learning Turkish as
a Foreign Language. Educational Research and Reviews 2015. Vol. 10(18). pp.
2575-2589, 09, 2015. [Link]
Vazquez, A. (2014). Building Academic Confidence in English Language Learners in
Elementary School. Unpublished Bachelor Thesis