The Correlation Between Students’ Speaking Self-Efficacy and Collocation Competence in Speaking at SMPN 1 Parengan

Speaking, among the four basic language skills – speaking, listening, writing, reading-, is of an exclusive importance in daily life and it is a vital complementary tool in communication skills. Unfortunately, it is not that easy to master English speaking skill and use it to communicate, especially for students who learn English as foreign language (EFL), where they rarely find the situation to use English. Meanwhile, self-efficacy is one of the factors that affect the success of EFL students speaking skill. On the other hand, collocation is a succession of two or more words that must be learned as an integral whole and not pieced together from its component parts which is also an important factor in mastering speaking. This research was carried out to know whether there is a correlation between students’ speaking self-efficacy and collocation competence in speaking. The researcher chose ninth grade students of SMPN 1 PARENGAN to conduct this study. The researcher used the students 9A, 9B, and 9C as the sample which involved 103 students. The data was obtained from a speaking self-efficacy questionnaire and a speaking test. The result of the data correlation coefficient is 0,760 with 0.00 Significant (ρ = 0.00). So, null hypothesis (Ho) can be rejected. It means there is a positive and high correlation between students’ speaking self-efficacy and collocation competence in speaking. It can be concluded that the higher the speaking self-efficacy, the higher the collocation competence in speaking.

Unfortunately, it is not that easy to master English speaking skill and use it to communicate, especially for students who learn English as foreign language (EFL), where they rarely find the situation to use English. Some students find difficulty in using English when they are trying to interact with others. As Shumin (in Richard and Renandya, 2002) stated that speaking a language is difficult for EFL students in which effective oral communication requires the ability to use the language approximately in social interaction. Some EFL students tend to feel the anxiety to speak English, which make them tongue-tied or lost for words.
It means that speaking mastery does not only involves cognitive factors, but also affective factors as well (Schunk, 2003). This statement supported by Brown (2000) which says that if teaching is only based on cognitive consideration, the fundamental side of human behavior will be erased. He claims that the best acquisition will happen in environments where anxiety is low and defensiveness is absent. It shows that affective factors play an important role which gives great effect in the process of language learning. The affective factors here are motivation, anxiety, attitude, personal traits, self-esteem, and self-efficacy.
As one of the factors that mentioned above, self-efficacy appears to play a great role in predicting learners' performance in educational contexts and it can predict performance even better than actual abilities, or aptitude and it is a strong predictor of academic achievement (Doordinejad & Afshar, 2014;Hsieh & Schallert, 2008;Rahemi, 2007;Bandura (2006) ;Bentz, (2010); Rahimpour & Nariman-Jahan, (2010). In contrast, students having knowledge and skill needed in language learning do not always succeed proficiently to perform it (Tilfarlioglu & Cinkara, 2009 ;Pajares, (2001); Wang, Kim, Bong & Ahan, (2009). Therefore, self-efficacy mediates the relationship between knowledge and action. It is because self-efficacy affects individual's behavior in four ways: selecting choice of behavior, determining how much and how long of the effort, affecting an individual's thought patterns and emotional reaction, and recognizing human as producers. For example, students with low self-efficacy creates fear and doubt that bring them away to pursuit the goals of learning; then they think that activities are tougher and more difficult than they really are; this emerges stress and failure on learning.
But it turns out that being good at speaking performance is not only a matter of having self-efficacy, but it is also a matter of mastery English vocabulary at collocation level or the vocabularies used. Thus, collocation competence is also important in order to reach communication goal. Therefore, to develop one of the hallmarks of an advanced language user; collocation competence which is the possession of a sufficiently large and sufficient phrasal mental lexicon that enables the student to produce language that is fluent, accurate and stylistically appropriate (Lewis, 2000).
It is important for English as foreign language (EFL) learners in Indonesia to produce their speaking well because English and Indonesian language has different structures, so there would be a possibility that there would be some ambiguities if they do not have good collocation competence. Therefore, collocation competence in this case holds an important role, that collocation which the combination of chunk of words makes different meaning when they do not be with its exact pairs. But what mostly happens to our English learners is that they directly translate Indonesian and make it sounds unnatural.
Next, EFL learners are often not even aware of the significance of collocation knowledge. In other words, the students often do not recognize the importance of collocation knowledge (Laufer, 2005). Students often underestimate the difference between receptive and productive vocabulary knowledge, and they assume that if they understand the words, they will be able to use them as well. As a result, many students tend to identify vocabulary learning with the acquisition of new words identify vocabulary learning or with an expansion of vocabulary size and they fail to pay sufficient attention to collocation relationships in the input. Also, they perceive knowledge of difficult or specialized words as an indicator of language progress. As a result, they often focus on memorizing long, decontextualized word lists, and they rarely pay attention on how these words are actually used. Those cases above happen because for intermediate students of a foreign or second language, vocabulary has always been a bottleneck impeding their progress. At this stage, the traditional way of memorizing the meaning and pronunciation of a new word is far from meeting the need of the students. With the expansion of the size of vocabulary, many problems arise concerning the proper use of a word in context, among which is collocation.
Jaen (2007) proposed that in order to speak fluently and accurately, non-native speakers of English have to acquire a vast knowledge of collocation. Another research done by Yazdandoost, AmalSaleh, and Kafipours in 2014 which analyzed the relationship between collocation knowledge and listening, speaking, reading, and writing proficiency of 50 Iranian EFL learners showed a significant correlation between collocation knowledge and listening, speaking, reading, and writing proficiency. It turned out that collocation knowledge has the biggest impact on speaking proficiency.
Finally, based on the speaking self-efficacy and collocation theories above, students have to have high speaking self-efficacy, so they will sustain in learning English, especially make great effort to master collocation as one of the ways to excel at speaking performance because when they enhance their collocation competence, their English skills in spoken will be improved.
As many research revealed the importance of self-efficacy in higher level of academic setting (e.g. Anggarini, 2014;Amwazir, 2013;Wahyuni, 2015), this study will be conducted for lower level which is junior high school. Regarding to see the effects which are caused by students' speaking self-efficacy and collocation competence on their speaking performance, the researcher carried out a study entitled "The Relationship Between Speaking Self-Efficacy and Collocation Competence in Speaking at SMPN 1 PARENGAN".

a. Research Design
The researcher used quantitative research design to conduct this research. The researcher also used a correlation research design to know the correlation between students' speaking self-efficacy and collocation competence in speaking.
There two variables in this study. The first is speaking self-efficacy is symbolized by "X" as independent variable and the second is collocation competence in speaking is symbolized (Y) as dependant variable.

Independent Variable
Dependant Variable Speaking Self-Efficacy (X) Collocation Competence in Speaking (Y)

Population
In this research, the researcher chose the students of ninth grade at SMPN 1 PARENGAN in the first academic year 2021/2022. The population of this research is ninth grade students at SMPN 1 PARENGAN 2021/2022 academic year which consist of 208 students.

Sample
To determine the total sample of this research, the researcher adopted formula from Ary et al (1990). They stated, the smaller level of standard error (1%, 5% or 10 %) the larger sample should be taken close to the amount of population. Sevilla et. al (2007) demonstrated formula to determining sample: = 1 + 2 n = Sample N = Population e = Level of error tolerance After applying this formula, the amount of the sample which should be taken at 5% level of error tolerance in this study is 103. It is almost close to the amount of the population itself. Nasution (2006) explained that the higher amount of sample the higher strength of generalization can be obtained. In line with Widi (2010), the larger sample the larger certainty and accuracy can be obtained. So, the total of the sample is 103 students of ninth grade at SMPN 1 PARENGAN.

Data Collection
In this research, the researcher collects the data from a speaking self-efficacy questionnaire to the students. After that, the researcher gives a collocation competence speaking test. Then, the researcher begins to correlate the data by using SPSS program.

Research Instrument
There are two instruments used in this research. They are questionnaire and speaking test. The researcher administered the questionnaire to know the students' speaking self-efficacy. The questionnaire is adapted from Asakereh, A. and Dehghannezhad, M. (2015). The speaking self-efficacy questionnaire comprised of 28 items, based on a Likert scale ranging from strongly disagree to strongly agree. The score were arranged gradually into 5-4-3-2-1.
On the other hand, the collocation speaking test will use a rubric adopted from Xu (2015). The theme will be about "Asking and Giving Information" since they already got the material of "Asking and Giving Information" from their teacher.

Validity and Reliability
In order to design the speaking skills self-efficacy beliefs questionnaire items were adopted from Asakereh, A. and Dehghannezhad, M. (2015). According to them, the questionnaire enjoyed acceptable validity and the reliability was 0, 84. It means the questionnaire has very high correlation and reliable. While the collocation speaking test will use a rubric adopted from Xu (2015). The rubric was valid and the reliability was 0, 96. It means that the rubric has very high correlation and reliable

Procedure of Collecting Data
In order to reach the research objective of this research, the researcher attempts to do these procedures for collecting data to support the data analysis, they are: a. Then the researcher will conduct an online meeting with the students and tell them the detail of the research and what they need to do.
b. The researcher share the link of the questionnaire and ask the students to fill the it.
c. Do a speaking test online.
d. After the questionnaire and students' speaking score obtained, the researcher will analyse the data using SPSS 16.

Data Analysis
After collecting the data from sources through some procedures mentioned above, the researcher analyses the data by doing some steps to find the answer of the research question. So, the researcher inputs the data from the students' selfesteem and speaking ability of ninth grade at SMPN 1 PARENGAN into SPSS 16 program for doing statistical analysis.

a. Descriptive Statistics
Since the statistical analysis in this study is done both descriptive and inferential statistically, the descriptive analysis is the first analysis in this study. Descriptive analysis is important to help a researcher make sense of a large amount of data (i.e. variables) by reducing it to a more interpretable form so it can be understood easily. The description of the data can be seen from mean, standard deviation, variance, modus, etc. (Zawawi, 2012). The description of the data statistically can be explored by using Descriptive Statistics in SPSS.

b. Normality test
Normality test is one of important requirements in the procedure of this research. It is to check whether or not the data is distributed normal by using Test of Normality. To test the normality, the researcher will insert the data into SPSS and explored its normality by using Shapiro-Wilk and Liliefors (Kolmogorov-Smirnov).
The result of this Test of Normality is very important to determine which inferential analysis statistic that will be used to examine the correlation of the variables. From the result of Normality Test, if the data is distributed normal, the researcher uses parametric statistics analysis to find correlation coefficient, in this case is Pearson Product Moment. Nevertheless, if the data is not from normal distribution, the researcher uses non-parametric statistics analysis to find correlation coefficient, in this case is Spearman Rank.

c. Coefficient of Correlation
Correlation Coefficient is number which shows the strength of correlation between two variables (Zawawi, 2012). The result of the Correlation Coefficient will determine the strength of the correlation between self-esteem and speaking ability of ninth grade students at SMPN 1 PARENGAN. In this research, the researcher adopted classification of level correlation from Arikunto (2008).

d. Hypothesis Testing
The researcher uses hypothesis testing to find statistical significance of the correlation coefficient. Since this is an educational study, the level of significance is on 5% level (0.05). According to Best (1981), in educational circles, the 5% (0.05) alpha level (significance level) is often used as standard for rejection. So, the principle is Null hypothesis (H0) cannot be rejected if P value (Sig.) is bigger than 5% (0.05). Meanwhile, the Null hypothesis of this study is there is no significant correlation between Self-esteem and Speaking ability.
According to Best (1981), the test of significance of correlation coefficient can be computed with formula: -tr= the statistical significance of the correlation coefficient -r = the correlation coefficient -N = number of paired ranks Finally, the researcher will test the hypothesis whether the null hypothesis is rejected or not.
This part explains the finding by analysing and reflecting of the previous study to get detail explanation toward the findings. This part the researcher focuses on the correlation between students' speaking self-efficacy and their collocation competence in speaking.

Students' Speaking Self-Efficacy
Based on the result of data analysis on students' speaking self-efficacy obtained from the questionnaire, it was found that most of the students agreed with statements such as Meanwhile, for item number 25 which said "I can introduce my teacher to someone else in English" only 3,9% of the students who disagree that they can introduce their teacher to someone else. It means that only a few students who have low speaking self-efficacy.
The result showed that the questionnaire was dominantly answered "agree" by the students rather than "disagree". It means that the ninth grade of SMPN 1 PARENGAN have high speaking self-efficacy which is good for them when they learn a foreign language because as Bandura said, how people behave can often be predicted by the beliefs they hold about their own capabilities (more so than by what they are actually capable of accomplishing). That is, individuals' perception about how they themselves will perform at specific tasks and their confidence in their ability to complete them successfully has a large impact on how these individuals actually behave in situations. Besides that, self-efficacy is what an individual believes he or she can accomplish using his or her skills under certain circumstances. Said differently, self-efficacy is a judgment people make about their own future potential and ability (Margolis & McCabe, 2006).

Students' Collocation Competence in Speaking
Based on the result of the data analysis on collocation competence in speaking, the researcher found that the students of ninth grade at SMPN 1 PARENGAN shows that the highest mean was 1,43 for Semantic Accuracy with the SD 0,497. While the lowest mean was 0,50 for grammatical accuracy with the SD 0,502. It shows that the students have good collocation competence in speaking.

The Correlation Between Students' Students' Speaking Self-Efficacy and Collocation Competence in Speaking at SMPN 1 PARENGAN
Based on the other result especially of Pearson r correlation analysis, it was found there is relationship between students' speaking self-efficacy and speaking competence in speaking of ninth grade students at SMPN 1 PARENGAN. The result of the data correlation coefficient is 0,760 with 0.00 Significant (ρ = 0.00). It means there is a positive and high correlation between students' speaking self-efficacy and collocation competence in speaking. It can be concluded that the higher the speaking self-efficacy, the higher the collocation competence in speaking.
Students with high speaking self-efficacy will sustain in learning English, especially make great effort to master collocation as one of the ways to excel at speaking performance because when they enhance their collocation competence, their English skills in spoken will be improved. Students with higher speaking self-efficacy will manage to have higher collocation competence in speaking.

Conclusion
Based on the research findings and discussions in the previous chapter, the researcher comes to the following conclusion: a. The students' speaking self-efficacy of ninth grade at SMPN 1 PARENGAN is high. It indicates that almost all of students have high speaking self-efficacy.
b. Students' collocation competence in speaking of ninth grade at SMPN 1 PARENGAN is high. The mean score was 4,59 compared to the maximum score (6).
c. The result of the data correlation coefficient is 0,760 with 0.00 Significant (ρ = 0.00). It means there is a positive and high correlation between students' speaking self-efficacy and collocation competence in speaking. It can be concluded that the higher the speaking self-efficacy, the higher the collocation competence in speaking.