‘Getting Stuck’: A Study of Indonesian EFL Learners’ Self-Efficacy, Emotional Intelligence, and Speaking Achievement

The problems of speaking mastery faced by EFL students in higher education are various. The frequency of making mistakes in completing specific tasks is because of poor communication skills. Students have minimum chances resulting lack of practice that affects their successful performance. Self-efficacy and emotional intelligence were determined to be two essential components to developing and improving learners’ speaking abilities. This study is purposed to investigate the correlation between self-efficacy, emotional intelligence, and the achievement of students’ English speaking. This study used a quantitative correlational design. The data are gathered by using a self-efficacy scale expanded from the General Self-Efficacy Skill (GSE), emotional intelligence was measured with the emotional intelligence scale adapted from the Emotional Regulation Questionnaire (ERQ), while the speaking achievement was administered to 100 students in public speaking class. The students performed a seven-minute speech and scored based on the content, proficiency, and delivery method. The data on self-efficacy and emotional intelligence were analyzed by using multiple correlations and multiple regressions. The finding shows a significant correlation between


INTRODUCTION
Speaking mastery has been regularly studied in relation to the effects of numerous issues encountered by the learners of English as a Foreign Language (EFL) in higher education.Teaching English speaking is quite challenging in English Language Teaching (ELT) regarding input-poor contexts.Learners are not exposed to English as the target language outside of the class, but they might practice it in classroom activities.Learners are anxious and quiet in speaking class due to the fear of making mistakes to do individual tasks with limited time (Anggraeni et al., 2020).If it is connected to self-criticizing related to mistakes in grammar or pronunciation (Aslan & Şahin, 2020), the students have difficulties in expressing their ideas and thoughts caused by minimum communication ability, cannot choose or use the appropriate words, idioms, structures and phrases (Ugla et al., 2012).Indeed, they also have few opportunities to practice speaking English as a foreign language (Alotumi, 2021;Uztosun, 2020).It causes a lack of practice, lowers their motivation, and impacts their achievement (Mandasari & Aminatun, 2020).Speaking achievement should be based on the classroom participation developed by the teachers to promote a supportive and non-threatening learning process (Aslan & Şahin, 2020).According to Celce-Murcia et al. (2014), speaking can be performed based on several activities; those are discussion, conversation, role play, audiotaped journal, drilling, and speech.As this study focuses on the speech performed by university students, they should create a way on how they could speak fluently by starting with an excellent one-minute opening speech (Santoso, 2017).Following today's trend that learning needs to involve technology, the speech is video-recorded by using information and communication technology (Djiwandono, 2019), then, it is posted on social media so that teachers and classmates can evaluate their speaking performance (Santoso & Mandarani, 2021), and students can evaluate their peers through virtual learning (Alberth et al., 2018).
However, speaking ability is a combination of cognitive, physical, sociocultural, and speaker inquiry information that occurs in a real-life time and context (Goh & Burns, 2012).Yet, even though much research is quite helpful in the teaching and learning process, especially in improving speaking achievement these decades (Gani et al., 2015;Kanaker & Ghani, 2018;Muslem & Yulianti, 2015;Sahardin et al., 2019), the learners' achievements are however still below expectation (Alsyouf & Al Kayed, 2021).According to an informal interview, we found a few learners who prepared the speaking performance confidently.They put some supporting materials for presenting a speech and added appropriate visual aids, for instance, texts, pictures, and short videos compiled altogether in the PowerPoint.This preparation helps them improve their performance better confidently.Furthermore, they also ensure and manage their emotion to finish the speech at the right time which impact on their achievement.It is a kind of belief influenced by the enthusiasm for improving initiative activities, so-called self-efficacy (Demir et al., 2015).However, although, this activity could influence their performance better, not many students do this.In addition to online learning sessions, students may require additional work to perfect their speaking skills as a result of their English exposure and practice which affects their lack of confidence (Ellis, 2008).According to our best practice of teaching experience to the students of the English Education study program at the biggest private university in Sidoarjo, Indonesia, to gain speaking achievement during the pandemic COVID-19 outbreak, a student has to perform in online meetings where other students can see their classmates' faces.In doing a speaking performance, the student looks at the camera or sometimes the screen while other classmates can see him or her on screen.From the performance, it appears that the eyes move left and right, some others look up or down without any specific reasons.
The students somehow directly stop speaking because they forget what to say next.After several seconds of pause, they resume the new things that may or may not have a relationship with the previous topic they presented.It seems that they plan the material and memorize every single word before presenting which makes it difficult to wake up after losing their memorization.In the end, the result of their speaking presentation on the screen shows problems with self-efficacy in speaking.Students may become tongue-tied or unable to find words, resulting in poor speaking performance.As a result, self-efficacy is viewed as a source of tension and despair that plays an important role in speaking because of its ability to relieve anxiety and negative ideas (Tahmassian & Moghadam, 2011) that correlate to the learners' English proficiency.
On the other hand, in the middle of delivering the presentation, many students tend to get stuck, for instance, they forget about the following materials to say in front of the audience.Most of them focus on memorizing the long script instead of improvising from the outline they have made through PowerPoint.So, they pause their performance due to the nervous feeling several times.These chronological problems affect their emotional intelligence.They have difficulty managing their emotion when delivering the presentation.Emotional intelligence in EFL students tends to boost such confidence related to the ability to plan and execute the steps; from opening speech, body of the speech, and closing; which are important for a successful performance.Learners who can recognize, control, and analyze their emotions may be more capable in their communication skills (Afshar & Rahimi, 2014).Concerning that matter, psychological issues such as lack of prior knowledge of the subject and low self-esteem are sources of students' inability to talk and fear of speaking in front of others.These fear and anxiety emerge when the students get their turns to deliver a certain topic, they worry about spelling and grammatical errors, mispronouncing specific words, and being evaluated by their peers or the lecturer (Aslan & Şahin, 2020).The theoretical D. R. Santoso,G. R. Affandi & Y. Basthomi,emotional intelligence,and speaking achievement | 387 framework revealed a significant lack of research on the correlation between selfefficacy, emotional intelligence, and students' speaking achievement.Many studies on self-efficacy and emotional intelligence in education are utilized to predict instructors' effectiveness in teaching based on our best understanding (Shead et al., 2016).
Although the literature in this area abounds with studies investigating the correlation between self-efficacy and emotional intelligence in EFL learning, limited studies in the Indonesian context examine the relationship between the two variables and the components of EFL learners' speaking achievement.As a result, to address a research gap, this study will look into the relationship between self-efficacy, emotional intelligence, and speaking ability.The research questions are phrased as follows: 1.What is the correlation between self-efficacy, emotional intelligence, and speaking ability among students?2. Which of the following factors, self-efficacy or emotional intelligence, is a better predictor of speaking ability?3. Which one is a stronger predictor of speaking ability, self-efficacy, or emotional intelligence?
According to the three research questions above, this research is expected to give significances.Theoretically, this research contributes new insight into the correlation between self-efficacy, emotional intelligence, and speaking achievement at higher education levels.Practically, this research can help English teachers easily predict the students' speaking achievement by knowing their self-efficacy and emotional intelligence.Therefore, the novelty of this research is the formula of the predictor of speaking achievement.

Self-Efficacy
Although studies on self-efficacy have been implemented as the predictor (Choi & Lee, 2016) in many lines of life such as for assessing perceived stress and mental health status (Moeini et al., 2008), conceptualize students' experience on science (Cartwright & Atwood, 2014), based on our best understanding, much education research is still at the top of google search.In education, self-efficacy is used to predict how effective teachers are in the teaching and learning process (Chukwudi, 2014;Clark & Newberry, 2019;Lee & van Vlack, 2018).Regarding speaking as a communication tool, self-efficacy is related to someone's belief to manage the situation and get a positive outcome dealing with the intuition to carry out and arrange the actions involved in the production of the projects.It is emphasized that selfefficacy is a crucial part of studying a language to obtain the English communication qualification (Ningias & Indriani, 2021).
The term 'self-efficacy' refers to someone's skill to coordinate and execute game plans to attain specific goals (Bandura et al., 1999).It affects the EFL students' performance in learning language skills, such as speaking.The self-efficacy in performing speaking will guide the students to gain better achievement, for it is strongly related (Alawiyah, 2018).In addition, Satyarthi and Malhotra (2021) describe self-efficacy as the belief in someone's capability to manage others' motivation, behavior, and social environment called cognitive self-evaluations.It affects experience which involves goals that people struggle with.The amount of energy expended to reach achievement and certain levels of performance.In other words, the studies on speaking anxiety and self-efficacy suggested that students might associate stress and depression in particular with the speaking task they were assigned (Fatimah, 2018).This is because speaking can be challenging for students learning a new language because it calls for the ability to interact with others productively.
Yet, the personal assessments of students' ability to plan and carry out actions in order to obtain specified educational results are referred to as perceived academic self-efficacy (Bandura et al., 1999).They created scales to assess perceived academic efficacy as a part of a micro-analytic procedure to determine its level, generality, and strength across activities and contexts.In terms of academic performance, self-efficacy level refers to variations across various levels of tasks.It involves the assessments of students' skills to carry out tasks.The students evaluate themselves based on their ability to complete a task, not on who they are as persons or how they feel about themselves generally.Second, rather than being a single trait, efficacy beliefs are multifaceted.Third, self-efficacy measures vary depending on the environment since a variety of non-ability effects can either improve or hinder the use of skills.Students may display less confidence in their ability to learn in competitive classroom environments than in cooperative ones, for instance.Self-efficacy measures' dependence on a mastery performance criterion rather than normative or other criteria is a fourth characteristic related to their strength component.Students are asked to rate their confidence in their ability to solve mathematical problems of various difficulty, not how well they think they will perform relative to other students.Finally, selfefficacy is evaluated prior to pupils engaging in the necessary actions.For assessing the function of self-efficacy beliefs in causal systems, this antecedent attribute offers temporal ordering.

Emotional Intelligence
Learners with strong emotional intelligence engage in more speaking and brainbased tasks.In reality, their excellent self-esteem and social abilities enable them to readily collaborate with others.Learners with poor emotional intelligence, on the other hand, have less suitable social relationships, which isolates them from the classroom environment and prevents them from participating in speaking and brain-based tasks (Bora, 2012).
Emotional intelligence, according to Salovey and Grewal (2005), considers emotions as useful sources of information that help in making sense of and navigating the social world, bringing together the realms of emotions and intellect.They classify emotional intelligence into four categories.Perceiving emotions is the first branch, which refers to the capability of recognizing and decoding emotions through face, photograph, sound, and cultural artifacts.To put it another way, recognizing emotions entails the skill to recognize someone's emotions.This kind of skill that harnesses emotions to facilitate a variety of cognitive functions, such as thinking and problemsolving, is the second branch.Understanding emotions is the third branch, which allows you to decipher emotional language and appreciate complex emotional interactions.It includes the ability to be sensitive to minor changes in emotional states.The fourth branch is emotional management, which entails the ability to control our own and others' emotions.It demonstrates that everyone has had instances in their lives when they have temporarily lost control of their emotions, which can be embarrassing.
As quoted in Esmaeeli et al. (2018), the importance of emotional intelligence as a variable is connected to individual differences in language learning.It has been taken into account together with other elements that show individuals' particular language acquisition capacities.In light of this, emotional intelligence appears to be crucial for language learning.The most significant issue for EFL learners is that they are inadequate in oral communication, which is necessary for learning a foreign language.Speaking has thus always been difficult for the students.In short, the ability to recognize and express emotion, absorb emotion in cognition, understand, and reason with emotion, and manage emotion is referred to as emotional intelligence.They found that the techniques sought to identify the emotional characteristics of emotionally intelligent people, as well as the methods and procedures that cause the abilities to be used in everyday life.As a result, emotionally intelligent people are greater team players and more effective motivators (Chukwudi, 2014).Furthermore, emotional intelligence is described as the capability to control someone's and others' emotions, as well as the capability to use feelings and emotions to improve motivation, formulate and implement strategies, and achieve planned goals (Mayer & Salovey, 1995).So, emotional intelligence is then used to hypothesize and appeal the accurate emotional appraisal and expression among one person and another (Afshar & Rahimi, 2014).Our study tries to hypothesize the emotional appraisal of students' performance in presenting speaking, and whether their emotion influences the speaking performance that impacts their speaking achievement.Chukwudi (2014) showed a significant positive correlation between emotional intelligence, efficacy, and effectiveness in pre-service teachers.Emotional intelligence and teacher efficacy were found to be important predictors of teachers' effectiveness.In relation to emotional intelligence's nature, the study found that emotional intelligence is a strong predictor of teacher effectiveness.Abdolrezapour and Tavakoli (2012) experimented 63 students regarding with the possibility based on Goleman's concept and its relationship to reading comprehension achievement, to improve emotional intelligence through prepared literary response activities.They measured the result of emotional intelligence and reading comprehension achievement by using the Trait Emotional Intelligence Questionnaire-Adolescent Short Form and discovered that the tested group's reading achievement was higher.
In addition to the studies of emotional intelligence, they are used for predicting the students' nature of emotion toward the acceptance of the materials.In brief, selfefficacy is an important psychological capital so that students can achieve optimal performance.Apart from that, the ability to manage emotions can help students to remain calm when demands for high achievement, and stable emotions will help students get optimal information in learning, one of which is speaking achievement.Thus, according to the issues, we assume that self-efficacy and emotional intelligence are correlated to the students' speaking achievement Ateş (2019) measured emotional intelligence to reading comprehension skills and reading anxiety.He investigated 138 students and found that emotional intelligence affects reading comprehension skills and reading anxiety.He found that the students' emotional intelligence decreased their overall reading anxiety levels (ß=-6.69).The pupils' reading comprehension skills improved (ß=9.079) as a result of their general emotional intelligence level.It showed that female students (x=3.389) had higher general emotional intelligence ratings than male students (x=3.181)(t=2.944;p=0.004 <0.05).Learning a second language is a tedious and intricate process that needs to implement potential resources to reach optimal results (Esmaeeli et al., 2018).English teachers must consider either internal or exterior mechanisms to fully comprehend the complex language acquisition process (Santoso & Mandarani, 2021).In this regard, because emotional intelligence is a necessary internal component linked to individual differences in the classroom, English teachers need to be aware of the emotional intelligence concept and make efforts to help students develop their abilities.Students will be more engaged communicators and attain a high degree of proficiency in language lessons as a result of emotional intelligence (Barany, 2021).
Since the review of the relevant literature found that self-efficacy and emotional intelligence are two crucial elements in the development of EFL, cultivating these skills also enhances learners' speaking proficiency.It is the result of a deficiency in the development of self-efficacy and emotional intelligence, and it may negatively impact students' speaking proficiency in a foreign language learning course (Barany, 2021).In other words, learners with high self-efficacy and emotional intelligence are more likely to succeed than those who have low self-efficacy and emotional intelligence.

METHOD
The association between self-efficacy, emotional intelligence, and speaking achievement was investigated using a quantitative correlational methodology in this study.The students are from the Study Program of English Education at the Faculty of Psychology and Education, Universitas Muhammadiyah Sidoarjo.The sample employed is quota sampling, quota sampling is a technique of non-probability sampling in which the samples are chosen according to the probability proportionate to the distribution of a variable in the population (Yang & Banamah, 2014), which includes 100 students in a speaking class.

Data Collection
Self-efficacy scales, emotional intelligence scales, and speaking achievement criteria were used to collect data.To measure self-efficacy, a scale adapted from The General Self-Efficacy (GSE) is used.The GSE scale is a self-report self-efficacy evaluation (Schwarzer & Jerusalem, 1995).Ten items must be summed to get the total score.They added that the total score ranges from 10-40 which explained the higher the score indicated higher self-efficacy (see Table 1).I feel confident in my ability to deal with unforeseen situations well. 5.
I know how to deal with unforeseen situations thanks to my resourcefulness.6.
If I put in the effort, I can solve most challenges.7.
I can remain cool in the face of adversity because I have confidence in my coping abilities.8.
When I'm faced with a dilemma, I usually come up with various options.9.
When I'm in a pickle, I generally come up with a solution.10.I'm typically able to handle whatever that comes my way.
The emotional intelligence measure was adapted from the Emotional Regulation Questionnaire (ERQ) as a start.A ten-item scale was developed to assess participants' ability to regulate their emotions in two ways: cognitive reappraisal and expressive suppression (Gross & John, 2003).Each issue is answered on a 7-point Likert-type scale ranging from 1 to 7 (strongly disagree to strongly agree) (see Table 2).
Most of the time, I have a good understanding of why I experience various feelings.

2.
My personal emotions are well understood by me.

3.
I completely understand how I'm feeling.4.
I can always tell if I'm happy or not. 5.
I am always able to deduce my friends' feelings based on their actions.6.
I am an excellent observer of other people's emotions.7.
I am sensitive to other people's sentiments and emotions.8.
I have a solid grasp of the feelings of others around me. 9.
I make it a point to create objectives for myself and then work hard to attain them.10.I always tell myself that I am capable.11.I am a self-motivated individual.12.I would always tell myself to give it my all.13.I can manage my emotions so that I can deal with problems rationally.Next, the speaking achievement criteria are administered from the students' scores when performing a speech.Since the speakers are the students of advanced-mid level, Luoma (2009) stated that the learners can handle a significant range of communication activities with ease and confidence, as well as participate actively in most casual and formal interactions on a variety of concrete topics.They also display the capacity to narrate and describe in all major time frames (past, present, and future) by presenting a complete account with good aspect control while adapting flexibly to the conversation's demands.As a result, in terms of English competence and delivery mode, the National Certificate descriptive scale, as specified in Luoma (2009), and presented in Table 3 below.Table 3. Scoring Criteria of English Proficiency and Delivery Method (adapted from The National Certificate Descriptive Scale National Board of Education in Luoma, 2009).

English proficiency
Delivery method 6 With few if any non-native traits, such as a foreign accent, speaks fluently.Is capable of conveying even small subtleties of meaning with clarity, as well as using idiomatic terms in a variety of ways.

6
Can explain even the most difficult issues and incorporate sub-themes, develop several points of view, and bring the presentation to a satisfactory finish.
5 Speaks fluently without the need to look for an expression frequently.Naturalness, coherence, and adequate length characterize this delivery.

5
Can give a clear and complete description of even the most difficult subject.Idiomatic and daily idioms can be used, and nuances can be expressed rather well.4 Even in less common speech settings, he seems to cope rather effectively.At least to some extent, distinguishes formal and informal registers.Is capable of clearly presenting and justifying an argument.

4
Can explain and discuss sights, sounds, and feelings.Due to a lack of linguistic proficiency, he is only seldom forced to employ circumlocutions in ordinary communication. 3 Deals with the most common speech scenarios and is capable of taking the lead in ordinary language situations.Although the speech is sluggish, there are seldom unnecessary pauses.
3 Despite translating local or foreign language structures and terminology to the target language, it is understandable.Pronunciation may diverge significantly from target language norms. 2 Deals with everyday speaking circumstances that call for a quick exchange of information.However, the speaker's language proficiency severely limits the range of topics that can be discussed.

2
The interlocutor's willingness to assist the speaker in developing the message is a prerequisite for successful message communication.Pronunciation may diverge significantly from the norm in the target language, necessitating more effort on the part of the interlocutor and inhibiting successful communication. 1 Although it can handle the most basic speaking duties, communication is slow and scattered.To be understood, they are frequently forced to use nonverbal ways.
1 Is competent to ask and answer simple questions about immediate daily requirements.Simple polite forms can be used.
Furthermore, the scoring criteria also require the content or the material that reflects the speaking skill.Rybold (2006) developed three main speech organizations: opening, body, and closing.The opening consists of the introduction of the topic, thesis statement, and preview; the body is arranged with sub-topics, reasons for choosing the topic, and some evidential support; and the closing tells the reinforcement of the central idea, followed by summarizing or concluding the topic.Rybold (2006) only provided a long description regarding the criteria of good speech content in each aspect so we classified them into several points.Next, to simplify the scores, it requires 1 and 0. Score 1 means that the speaker follows the content criteria, while score 0 does not (see Table 4).D. R. Santoso,G. R. Affandi & Y. Basthomi,emotional intelligence,and speaking achievement | 393 Table 4. Speech Content Evaluation Guideline, adapted from Rybold (2006).
No Content Score 1.
Opening speech the speaker gained attention and interest by giving initial questions/ story/illustrations/ jokes the speaker introduced the subject and thesis statement the speaker mentioned the preview body of the speech 2.
Body the speaker stated the main points clearly the speaker tells the reason why choosing the main points the speech was fully supported by related evidence the speech was well organized and provided transitions 3.
Closing speech the speaker prepared the audience for an ending by reinforcing the central idea The speaker summarized the whole speech The speaker presents a vivid closing and asks the audience to do something The steps to collect all the data in this study are gaining speaking performance scores and measuring self-efficacy and emotional intelligence.First, the speaking performance is gained from the students to present an informative speech.The theme is an analysis of small businesses in town.Before performing, the students are assigned to look for a small business around their hometown.Next, they interview the business owner, asking how the business can grow up, then, analyze if it grows quickly or slowly.Furthermore, the students present in front of their classmates.The score of English proficiency and delivery method is done as presented in Table 3, while the score of the speech content is taken from Rybold (2006), as presented in Table 4.The second step is measuring self-efficacy by using GSE as presented in Table 1.The third step is measuring emotional intelligence by using ERQ as shown in Table 2.

Data Analysis
Before the data are analyzed by using statistical parametric, the data were tested to find out the assumption by using the Assumption Test with Shapiro-Wilk Test for Multivariate Normality to see if the data were distributed normally or not.The data is categorized as normal if the level of significance is p>0.05.To evaluate the correlation and prediction of the variables, the data were analyzed using multiple correlation and multiple regression.It creates an influence, which then creates a statistical model that can predict whether someone has strong or poor self-efficacy and emotional intelligence.If students have a high level of self-efficacy, they will have higher scores in speaking.Then, if they have a high level of emotional intelligence, will also have higher scores in speaking.

RESULTS
The findings provide answers to the three questions mentioned earlier.Tables 5  and 6 describe the correlation of self-efficacy towards the students' speaking achievement.The analysis data of Table 5 above shows a significant correlation between selfefficacy and students' speaking achievement (F=9.510;p=0.003<0.05),while Table 6 shows the effect size of self-efficacy on students' speaking achievement is 0.102 or 10,2 %.The following Tables 7 and 8 describe the correlation between emotional intelligence and speaking achievement.The data analyzed from Table 7 shows that there is no correlation between emotional intelligence and speaking achievement (F=2.861;p=0.094>0.05).Due to the absence of a significant relationship between emotional intelligence and speaking achievement then this makes the effect size of emotional intelligence on speaking achievement is very small which is 3.3% (see Table 8).
After partially analyzing the data, we find that there is a substantial association between self-efficacy and speaking achievement; however, there is no significant relationship between emotional intelligence and speaking achievement.The outcome of the Assumption Test with the Shapiro-Wilk Test for Multivariate Normality is shown in Table 9 to demonstrate normality.The result of the assumption test above showed p=0.288>0.05which means the data is normally distributed so that it can be continued with the parametric test.Whereas, the following The analysis result of multiple regression with the entry backward method is used to test the correlation among self-efficacy and emotional intelligence toward speaking achievement of English Education students.It revealed that model 1 (the influence from 2 variables) scored F=4.825 and Sig.0.010<0.05.Whereas, model 2 which puts the variable of self-efficacy showed a bigger score F=9.510; Sig.0.003.
In addition, the big correlation of both variables (self-efficacy and emotional intelligence) towards speaking achievement is Adjusted R 2 =0.083 (8.3%).If we only put self-efficacy into the model, it has a bigger correlation Adjusted R 2 =0.091 (9.1%) presented in Table 7. Table 12 shows that according to the backward entry method, it can be seen that the best model for describing students' English-speaking skills is by using self-efficacy variables and eliminating emotional intelligence variables (t=3.084;Sig.0.003<0.05).So that the resulting regression equation to predict students' English-speaking ability with the following formula: Speaking Achievement=51.182+(0.776x self-efficacy) The intercept score as α=51.182 was used because based on the analysis which showed that there was only a significant influence between self-efficacy on English speaking achievement, the model included was only a regression model of the influence of self-efficacy on English speaking achievement.This formula is then used to predict the students' English-speaking achievement, whether this student is able or not to complete a task.The failure of emotional intelligence in predicting students' English-speaking skills, on the other hand, can be demonstrated by the results of the correlation test in Table 13.The results of the product-moment correlation demonstrate that self-efficacy and students' ability to speak English have a significant positive relationship (r=0.319;sig=0.003<0.05).The higher a student's self-efficacy, the better his or her capacity to communicate.However, while the significance value is more than 0.05 (r=0.181;sig.=0.094>0.05), the association between emotional intelligence and students' ability to speak English is not significant.
The second finding has to do with a novel model of self-efficacy as a mediator in the relationship between emotional intelligence and speaking ability.Using mediator analysis, a novel model was developed for the relationship between selfefficacy, emotional intelligence, and speaking ability.As seen in Table 5, emotional intelligence and self-efficacy have a significant link.The level of self-efficacy has a strong link to English proficiency.Emotional intelligence, on the other hand, has no bearing on how well you speak English.This suggests that self-efficacy is a moderator of the association between emotional intelligence and English language proficiency.Tables 14 and 15 show the outcome of the mediator analysis.Table 14 shows that emotional intelligence and speaking ability have no direct effect (z score=0.485;p=0.628>0.05).However, Table 11, after self-efficacy is put as the mediator variable, shows a significant result with z score=2.248;p=0,025<0.05.The self-efficacy picture as the mediator towards the correlation between emotional intelligence and speaking achievement can be seen in Figure 1.
Figure1.The Model of Self-Efficacy is the mediator in the relationship between emotional intelligence and speaking achievement.
Self-efficacy perfectly mediates the relationship between emotional intelligence and English-speaking achievement, as shown in Figure 1.This shows that to improve English-speaking achievement, what must be considered is students' confidence in their ability to speak English.Even though students have good emotional intelligence, if their self-efficacy is low, it will have an impact on achieving inadequate Englishspeaking achievement.Vice versa, even though students have emotional intelligence that is not too high, their self-confidence that they can speak English will have an impact on higher English-speaking achievement.

DISCUSSION
This section describes the findings of a study into the correlation between selfefficacy, emotional intelligence, and speaking achievement among Indonesian EFL learners.The two research questions are addressed.In response to the first question about whether there is a link between students' self-efficacy, emotional intelligence, and speaking ability, the results demonstrated that self-efficacy and emotional intelligence had a significant effect on ELF learners' speaking achievement.This is supported by previous research which states that self-efficacy in education is used to predict teacher effectiveness in the teaching and learning process (Chukwudi, 2014;Clark & Newberry, 2019;Lee & van Vlack, 2018).
This shows that students' self-efficacy and initial ability are important in the application of goal-setting techniques to students' English performance.Students who think that they are unable to achieve goals because they lack confidence, lack a strong conviction in their abilities, and do not have sufficient knowledge tend to refuse to set goals, which will have an impact on the performance they produce (Toshi, 1991).Depending on the intensity of their efficacy beliefs, according to Nguyen et al. (2022), self-efficacy and achievement were positively correlated.The research concludes with a series of suggestions for educators regarding the significance of self-efficacy in learning English achievement.
The sensation of self-efficacy is directly correlated to the feeling of competence being able to learn a foreign language because many students who learn a foreign language make previous assumptions about the success or failure of a person in learning depending on their learning experience, and this belief can affect when learning a foreign language (Hsieh & Kang, 2010).Students with high self-efficacy tend to more commit to achieving their objectives, obtain and use better strategies to achieve goals, and show a more positive response to negative rewards than those who have low levels of self-efficacy who in the past ultimately impact the resulting performance (Seijts & Latham, 2001in Locke & Latham, 2002), the performance in this case is English foreign language.
However, concerning the second research question, when comparing selfefficacy with emotional intelligence, self-efficacy is a better predictor of speaking achievement.Even when we tried to eliminate emotional intelligence in the statistics, meaning only included self-efficacy in the developed model, it made the influence model bigger.This result illustrates that the possibility of students getting 'stuck' is due to their lack of confidence in the abilities and experiences being pursued.
Self-efficacy becomes a powerful predictor of performance in various language skills and tasks related to mastery of foreign language proficiency (Raoofi et al., 2012).Research conducted by Raoofi et al. (2012) also determines the correlation of selfefficacy with English performance showing predictive power becomes higher at 24.3% by including covariables into the model rather than using only the determination model.The purpose of predicting English performance is only 7.3%.Otherwise, emotional intelligence is not correlated with the achievement of English speaking.This determines that self-efficacy is the variable mediator of the correlation between emotional intelligence and English-speaking achievement.Concerning emotional intelligence, as stated earlier, the correlation between emotional intelligence and students' achievement in speaking does not have a significant relationship.It contradicts the findings of Abdolrezapour and Tavakoli (2012) who found that students who were taught using traditional methods that did not lay any focus on emotional content or language scored higher on the Emotional Intelligence Test.

CONCLUSION
This study aimed to see three research questions: the correlation between selfefficacy, emotional intelligence, and speaking ability among students, which of the self-efficacy or emotional intelligence is a better predictor of speaking ability, and which of the self-efficacy or emotional intelligence can be the stronger predictor of the speaking ability.The results indicate a significant relationship (F=9.510;p=0.003<0.05) between students' speaking achievement and self-efficacy, with a 0.102 or 10,2% impact size for self-efficacy on speaking achievement.Emotional intelligence and speaking achievement, however, do not correlate (F=2.861;p=0.094>0.05).The lack of a meaningful correlation between speaking achievement and emotional intelligence means that the effect size of emotional intelligence on speaking achievement is just 3.3%.Therefore, utilizing self-efficacy factors and excluding emotional intelligence variables yields the best model (t=3.084;Sig.0.003<0.05) to describe students' achievement.
The result reveals a significant lack of research on the correlation between selfefficacy, emotional intelligence, and students' speaking achievement.The results of this research are not universal since it is limited to speaking achievement taken from the score gained from a hundred students taking the speaking performance.So, the result might not be helpful for other skills of English language learning.Since it had no significant effect on speaking achievement, further research might increase the number of samples to be tested and deepen the category of being tested.Lastly, further researchers can also develop self-efficacy and emotional intelligence toward other language skills.
This study contributes theoretically to all speaking teachers being able to reveal the higher education students' self-efficacy and emotional intelligence toward learning speaking.The higher the self-efficacy of the learners, the higher the score they could achieve in speaking.Otherwise, emotional intelligence does not influence the capability of the students' speaking ability.Thus, the contribution to the pedagogical development is that the teacher can use the formula: Speaking Achievement=51.182+(0.776x self-efficacy score) to predict the students' achievement to pass the standard made by the teacher in any subject.Therefore, this study conclude that the positive correlation between self-efficacy, and emotional intelligence can promote revealing students' achievement in learning speaking.
14. I'm really good at managing my own emotions.15.When I'm angry, I can always cool down quickly.16.My emotions are under my control.Note: Average items = Total Emotional Intelligence 1-16 Appraisal of Total Self-Emotions = Average Items 1-4 Total Emotional Regulation = Average Items 5-8 Total Emotion Use = Average Items 9-12 Appraisal of Others' Emotions as a Whole = Average Items 13-16

Table 5 .
The result of the ANOVA Test of Self-Efficacy and speaking achievement.

Table 6 .
The effect size of self-efficacy towards speaking achievement.

Table 7 .
The result of ANOVA tests emotional intelligence and speaking achievement.The intercept model is not used since it does not provide any useful information.

Table 8 .
The effect size emotional intelligence and speaking achievement.

Table 9 .
Assumption test with Shapiro-Wilk Test for multivariate normality.
Table 10 presents the result of the hypothesis test with the Multiple Regression Test with the entry backward method.Tabel 10.ANOVA-Multiple Regression Test.

Table 14 .
The direct effects of emotional intelligence and speaking achievement.

Table 15 .
Self-indirect efficacy's impacts as a mediator in the link between emotional intelligence and speaking achievement.