Discussion Task Model in EFL Classroom: EFL Learners’ Perception, Oral Proficiency, and Critical Thinking Achievements

This research was aimed to investigate the implementation of discussion task model toward the EFL learners’ oral proficiency achievement (OPA) and critical thinking achievement (CTA). Non-equivalent, pre-test, post-test control group design of comparison group design has been employed in this research. Oral proficiency rubric, critical thinking rubric, and a set of the questionnaire were used to assess the EFL learners’ OPA and CTA. The research findings showed that (1) the OPA and CTA levels in the experimental group were in the good category in which it was much better than that in the control group. (2) Statistically, the discussion task model gave a significant improvement and a significant mean difference between the EFL learners’ oral proficiency and critical thinking achievements after being taught by using discussion task model and those who were not. (3) There was a significant contribution to the aspect of oral proficiency and aspect of critical thinking toward the total of oral proficiency and total of critical thinking achievements where the highest contribution in the aspect of OPA and CTA was fluency and context. (4) The EFL learners gave positive perception of the discussion task model instruction in EFL classroom. In concluding, using discussion task model in EFL classroom significantly improved the EFL learners’ oral proficiency and critical thinking achievements..


Introduction
Speaking is pondered to be one of the most complex skills to be mastered by the English as Foreign Language (EFL) learners because it is used in their daily life particularly in the academic situation. In the sector of global trade, business, technology, economic, academic, and the like, the needs of English mastery of the international context is very crucial as this can be their medium of interaction to keep contact with other people. This will lead to the teachers' attention to seek the appropriate teaching strategies, media, and materials in order to meet the global need where the EFL learners are demanded to be a competent speaker and well-cultivated critical thinker so that they can compete with other global citizens.
Ironically, the EFL learners still come across with some barriers of interacting with others in the communicative sett. Ya-ni (2007) claims that the EFL learners feel frustrated when they cannot verbalize English in according way as being able to interact in English has given higher pride than being able to read and write in English. The frustrated feeling made by the EFL learners is mostly due to the avoidance of them to upgrade their communication skill because what they learn for ages just chiefly focuses on the vocabulary and grammar mastery. There are some complexities come across by the EFL learners when they learn how to speak English such as using mother language more frequently than using English in the classroom interaction and scarcity of English vocabulary mastery Al-Jamal, 2013, &Al Hosni 2014), have more feeling of anxiety and afraid to make mistakes when communicating (Dil, 2009, & Al Hosni, 2014. In line with that, it can also be caused by the unstructured strategies of listening, lack of English practices and time, as well as the resources in the teaching and learning environment (Basim 2007;Ahmed, 2012, &Al-Jamal, 2013).
Like speaking, the EFL learners are also demanded to be critical thinker, particularly on the understanding, analysing, and criticizing to what they learn in the academic teaching and learning environment. The teaching of critical thinking should, therefore, be exposed more in the classroom as this can be useful for EFL learners in the workplace after graduation. Nonetheless, to be the best in thinking, the EFL learners must cultivate their critical thinking skill as the skill of thinking is a natural process that the learners should experience in (Scriven and Paul, 2004). Critical thinking is a skill to be mastered by EFL learners in an effort to analyse and evaluate the information attained. This will make them be the critical thinkers as they can propose vital questions and problems, make a vivid formulation, gain and assess the related information, use the ideas, think broadly, and make effective communication with others. This is a vital skill to be earned in the workplace as this can assist the EFL learners to evaluate the colleagues, policies making, an institution where they work in as well as to cope with the social problems (Hatcher and Spencer, 2005). Pertaining to the rationale above, it can be drawn that the EFL learners, in general, still encounter some problems in terms of oral proficiency mastery and critical thinking cultivation.
To foster oral proficiency and critical thinking cultivation of EFL learners, various ways are used in the model of English teaching and learning. One of the potential models that can be applied is discussion task model (DTM) where this model is classified as one of the learner-centred classroom activities that can create interesting English teaching and learning environment in which the learners are required to be active, autonomous, and critical to what they discuss. It is supported by some English language scholars, in relation to oral proficiency, there were some students-oriented learning activities that can resolve the students with low communication skill such as group discussion, debates, storytelling, speeches, drama, and the like (Gathumbi & Masembe, 2005;Okech, 2005). These learning activities orientations can increase the learners' participation, motivation to utilize English in the academic context. Achmad and Yusuf (2014, p.151) stated that in the academic situation, the learners feel more convenient to keep working and interacting with their peers rather than with their teachers.
In addition, in relation to critical thinking cultivation, a discussion is a prominent teaching method that the teachers can employ as this can create active learning, developing understanding and cultivating critical thinking of the learners (Ikuenobe, 2002;Okolo, Ferretti, & MacArthur, 2007;Sautter, 2007). Gage and Berliner (2002) claim that -discussion based instruction can develop EFL learners' critical thinking, reasoning, and democratic participation. While having a discussion, the learners can state and formulate their viewpoints, accept an evaluation from others, and identify the unclear standpoints (Hadjioannou, 2007). Furthermore, discussion also can promote the EFL learners' tolerance to challenge others' perspective, increase their communication and critical thinking skill (Eggen & Kauchak, 2001). In addition, Brookfield (2012) points out that discussion is a method to sharpen the learners' critical thinking which involves the thinking process components such as identifying assumption, assessing assumption, seeing a different point of view, and taking informed action. By regarding the problem, strengths, and logical reasons described above, the researchers carried out the research to justify whether or not discussion task model gave a significant effect on the EFL learners' oral proficiency and critical thinking.
On the basis of the above mentioned description, the following research questions were formulated: 1. What were the EFL learners' oral proficiency and critical thinking achievements level before and after the applied intervention? 2. Did using discussion task model give a significant improvement on the EFL learners' oral proficiency and critical thinking achievements? 3. Did using discussion task model give a significant mean difference in the EFL learners' oral proficiency and critical thinking achievements? 4. How much did the oral proficiency (aspects) and critical thinking (aspects) contribute to oral proficiency (total) and critical thinking (total) ? 5. What was the EFL learners' perception of the implementation of discussion tasks model toward their oral proficiency and critical thinking achievements? In relation to the research questions formulated above, the research would like to: 1. Find out the EFL learner's oral proficiency and critical thinking achievements level before and after the intervention applied. 2. Find out whether or not the discussion task model gave a significant improvement on the EFL learners' oral proficiency and critical thinking achievements. 3. Find out whether or not the discussion task model gave a significant mean difference in the EFL learners' oral proficiency and critical thinking achievements. 4. Find out how much oral proficiency (aspect) and critical thinking (aspect) contributed to the oral proficiency (total) and critical thinking (total) after the intervention implemented 5. Find out the EFL learners' perception of the implementation of discussion task model toward their oral proficiency and critical thinking achievements. Based on the rationale explanation and the research questions formulation, the limitations of this research are chiefly focus on: 1. The implementation of discussion task model in EFL classroom on the EFL learners' oral proficiency and critical thinking achievements and their perception. 2. The aspects of oral proficiency (pronunciation, grammar, vocabulary, comprehension, and fluency) and the aspects of critical thinking (identifying and explaining issues, recognizing stakeholders and contexts, framing personal responses and acknowledging other perspectives, evaluating assumptions, evaluating evidence, evaluating implications, conclusions, and consequences). 3. The research site is in the University of Indo Global Mandiri, South Sumatera in which it included the university students, as the EFL learners, in any major having English subject.

Researches on Discussion Task Model towards Oral Proficiency and Critical Thinking
There are numerous empirical researches on the implementation of discussion task model toward the EFL learners' oral proficiency and critical thinking in the context of the academic teaching and learning environment. Adewuya (2003) revealed, that the activity of brainstorming in the discussion based teaching can be a step for the students to make opinions or viewpoint clarification, contribution, as well as feelings exploration. Stephen & Stephen (2005) also highlighted that discussion is interaction process in which it can assist the speakers and listeners to develop their knowledge and create mutual understanding. Having the same mind, Abdu-Raheem (2011) confirmed that teaching with discussion can activate the teachers and learners engagement in thinking process and promote the learners' social life skill. Yusuf & Al-Banawi (2013) also say that discussion and in class activities can activate the learners' involvement and promote their learning desire. Apart from that, conducting the research with the discussion tasks as the oral proficiency testing method to the Korean college level EFL learners (12 females and 5 males), Ryoo (2010) found out that when discussing the learners did not show much evidence of the positive value of collaborative interactional patterns in oral proficiency assessment.
In conjunction with the approach to critical thinking, there were an intellectual skill and emotional disposition approach that describe critical thinking. Fisher (2001) defines that critical thinking is categorized as intellectual skill and ability including assumptions identification and evaluation, claims acceptability, and explanations analysis, evaluation, and production. Having the same thought, Ennis (1990) also describes those critical thinking components consisting of elements and hidden assumptions identification and claims credibility evaluation.
Some researches related to the implementation of cooperative learning in the Iranian context, Zakeri (2010) did the research on the discussion based learning to the students of engineering where the findings revealed that cooperative learning can develop the learning process. Another researcher also recommended that cooperative learning can create the learning process for students in a meaningful way (Malakouti, 2009). Since this cooperative learning is not exposed regularly to the university classroom in terms of practice, the outputs are not vividly impactful to the students and professors (Noshadi, 2006). Besides, it is has become the crucial issue to seek the proper context to way up the critical thinking disposition as mostly of Iranian learners are not exposed not rehearsed to be a well-cultivated critical thinker (Fahim & Sa'eepour, 2011) Hence, the two discussion models done by Brookfield and Preskill (2005) were employed to see the learners' interpretation of those models. The research was carried out to see the critical thinking disposition components and social life interaction developed by Brookfield (2012) and Facione (2010) in relation to the two discussion types, the circle of voices and circular response. Additionally, Hacker and Niederhauser (200) claim that instructional design and strategies can lead to successful learning. They further say that there are four discussion tasks model namely inquiry-based task, guided discussion task, reflective discussion task, and exploratory discussion task.
From the researches mentioned above, this research highlighted on the development of learners' oral proficiency and critical thinking in university level and their perception through discussion task model based instruction. The researcher offered the guided discussion task model as the researcher, acting as the leader, can manage the situation of the discussion in order that every EFL learner, as discussants, can deliver their ideas and thoughts in turn.

Research method
A quasi-experimental design, primarily concerned on the non-equivalent pre-test and post-test control group design or comparison group design, was used in this research. A quasi experimental design, non-equivalent groups pre-test post-test control group design or comparison group design, was used in this research. Two groups, experimental and control groups were required in this research methodology. The researchers administered a pre-test in the experimental group, intervention by using discussion task model, and the post-test. After the post-test given, a set of close-ended questionnaire was distributed to the experimental group in order to see the EFL learners' perceptions on the instruction with discussion task model toward their oral proficiency and critical thinking. On the other hand, the researchers only gave pre-test and post-test with no intervention in the control group. The intervention was conducted for one session of lesson comprising of 16 instructional sessions where each session lasted for 90 minutes including pre-test and post-test.

Research Participants
The research was conducted at the University of Indo Global Mandiri. The research population was all undergraduate students in the academic year of 2017/2018. The purposive sampling selection was employed to the forty university students. Two equal-sized groups, experimental and control group, consisted of twenty undergraduate students, respectively. The research sample was taken which required to have the same academic year and was taught by the same lecturer and not having an English course during the research was carried out.

Data Collection
Oral proficiency test and a set of the close-ended questionnaire were administered to collect the data of the EFL learners' oral proficiency and critical thinking. Oral proficiency test has been used to assess the EFL learners' English oral proficiency. The same type of speaking topics have been chosen, in terms of oral test, and provided to test the EFL learners in the pre-test and post-test. The oral proficiency scoring rubric and critical thinking rubric have been used to assess the oral proficiency achievement and critical thinking achievement. The analytical scoring rubrics of oral proficiency and critical thinking have been used a hand in hand where the two raters/experts assessed the EFL learners' speaking performance and their content of utterances. On one hand, to assess the EFL learners' oral proficiency achievement, the researchers employed SOLOM (Students Oral Language Observation Matrix). This rating scale consisted of comprehension, fluency, vocabulary, grammar, and pronunciation with a five-point score for each variable (see: http://www. cal.org/twi/EvalToolkit/solom.pdf). On the other hand, to assess the EFL learners' critical thinking achievement, NEIU's critical thinking rubric comprising of issue, context, own perspectives, assumptions, evidence, and the implication was used in this research (see: https:/web.uri.edu/assessment/files/CriticalThinkingRubric_NEIU.pdf). Also, a set of close-ended questionnaire was distributed to see the EFL learners' perception of the implementation of discussion task model toward their oral proficiency and critical thinking achievements. The instrument comprised of 16 items with a 5 point Likert Scale (Strongly disagree = 1 to strongly agree = 5). The questionnaire items were constructed based on the research objectives. It chiefly concerned on the learners' perception about the implementation of discussion task model in the teaching and learning environment.
In order to figure out the reliability of the oral proficiency and critical thinking test, the researchers used inter-rater reliability. The two raters/experts judgments rated the oral language based on the speaking topics chosen. Then, the scores rated were calculated using the Pearson product moment correlation coefficient analysis. Apart from that, to figure out the scale reliability of the questionnaire, the researchers employed the coefficient of Cronbach Alpha reliability. The above reliability calculation was done to justify whether the rated scores and questionnaire items are reliable. From the results of the Pearson Product Moment Correlation and Cronbach Alpha computation, the results revealed that there was a significant correlation between the two raters' assessment the EFL learners' oral proficiency and critical thinking achievement. Therefore, the results of inter-rater reliability both pre-test and post-test in the experimental and control group and the result of Cronbach alpha were reliable as both significance values were higher than 70.

Instructional Procedures
The following is the instructional procedure of the discussion task model implementation in the experimental group as the English academic learning activity: 1. The lecturer explained the discussion task model objectives, rules, and its implementation. 2. The lecturer presented the topic of discussion. 3. The lecturer divided the class into a discussion group/team (Pros and Cons group). 4. The lecturer, as the facilitator, arranged the discussion setting and selected 3 speakers for each team. 5. The lecturer opened the discussion session and let the speaker from Pros group gave their opinion/viewpoints as well as the argument dealing with the topics given. During discussion, the speakers of both groups were allowed to propose objection and questions related to-the substantive delivery. 6. The lecturer let the speaker from the Cons group give a counter statement and their opinion/argument proposed from the Pros group. 7. The lecturer asked each discussion member to summarize their argument. 8. The lecturer added the summary from their argument. 9. The lecturer gave the topic of discussion for the next discussion session. 10. The lecturer dismissed the class meeting.

Data Analysis
The data analyses were taken from the test and a set of questionnaire. To analyse the EFL learners' oral proficiency and critical thinking achievements, analytical scoring rubrics were administered to gain their scores from the taken test. To interpret their individual score, the score range of oral proficiency was as follows: Very good (21-25), Good (16-20), Average (11-15), Poor (6-10), and Very poor (<6). On the other hand, the score range of critical thinking achievement was as follows: Very good (19-24), Good (13-18), Average (7-12), and Poor (19-24). The descriptive statistics, t-test statistical tools (Paired samples and independent samples t-test), regression analysis (multiple regression), and Likert scale analysis were used in this present research. The following are the simultaneous steps in terms of data analyses by using SPSS after having the intervention: (1) Descriptive statistics, (2) Paired samples t-test, (3) Independent samples t-test, (4) Multiple regression analysis, and (5) Likert scale analysis.

Research findings
In this research, four main findings were discussed in this research: (1) descriptive statistics analyses (mean score, standard deviation, frequency, and percentage, (2) statistical analyses (progressive and difference analyses) (3) multiple regression analyses (stepwise regressed percentage analyses), (4) Likert scale analyses (perception percentage analyses). Oral proficiency and critical thinking achievements, two major dependent variables, were analysed by using test and a set of close-ended questionnaire.

Descriptive Statistics Analyses
To answer into research question (RQ1), descriptive statistical analyses were used to see the mean score, standard deviation, frequency, and percentage of the EFL learners' oral proficiency achievement (OPA) and critical thinking achievement (CTA). This was meant to determine the EFL learners' OPA and CTA level before and after the applied intervention.
In descriptive statistics analyses, the results of pre-test administration to see the OPA level showed that in the experimental group there were 6 (30%) students in the poor category with the mean score and standard deviation (8.83 and 1.169), 10 (50%) students were in the average category with the mean score and standard deviation (12.10 and 1.449), and 4 (20%) students were in the good category with the mean score and standard deviation (16.50 and 0.577). Meanwhile, in the control group there were 10 (50%) students in the poor category with the mean score and standard deviation (8.50 and 1.354), 7 (35%) students were in the average category with the mean score and standard deviation (12.29 and 1.113), and 3 students (15%) were in the good category with the mean score and standard deviation (16.33 and 0.577). On the other hand, the results of pre-test administration to see the CTA level showed that in the experimental group there were 12 (60%) students in the average category with the mean score and standard deviation (8.83 and 1.267), and 8 (40%) students were in the good category with the mean score and standard deviation (14.25 and 1.035). Meanwhile, in the control group, there were 14 (70%) students in the average category with the mean score and standard deviation (9.14 and 1.657), and 6 (50%) students were in the good category with the mean score and standard deviation (8.83 and 1.267). From the results of the descriptive statistics, it could be stated that statistically the experimental group and control group had the same start, in terms of OPA and CTA, in which it could be seen from each category had the same category distribution and a close total mean score.
In addition, the results post-test administration to see the OPA level showed that in the experimental group there were 5 (25%) students in the average category with the mean score and standard deviation (13.60 and 0.894), 10 (50%) students were in the good category with the mean score and standard deviation (17.80 and 1.135), and 5 (25%) students were in the very good category with the mean score and standard deviation (21.80 and 0.837). Meanwhile, in the control group there were 6 (30%) students in the poor category with the mean score and standard deviation (9.67 and 0.516), 9 (45%) students were in the average category with the mean score and standard deviation (13.33 and 1.500), and 5 (25%) students were in the good category with the mean score and standard deviation (17.40 and 1.517). On the other hand, the results post-test administration to see the CTA level showed that in the experimental group there were 6 (30%) students in the average category with the mean score and standard deviation (10.50 and 1.049), and 14 (70%) students were in the good category with the mean score and standard deviation (15.50 and 1.787). Meanwhile, in the control group, there were 13 (65%) students in the average category with the mean score and standard deviation (9.85 and 1.463), and 7 (35%) students were in the good category with the mean score and standard deviation (15.14 and 1.345).
By concerning the results of descriptive statistics of post-test in both experimental and control group, it could be revealed that the EFL learners' OPA level in the experimental group was in a good category while EFL learners' OPA level in the control group was only in the average category. In relation to the CTA level, comparatively EFL learners' CTA level in the experimental group was in the good category while EFL learners' CTA level in the control group was in the average category. Hence, it could be concluded that the EFL learners' OPA and CTA level in the experimental group was higher than those who were in the control group which means the EFL learners in the experimental group were more competent speakers and more proficient thinkers than those who were in the control group after the intervention applied.

Normality of the Test and Homogeneity of Variances Test
In determining the normality of the test, the Shapiro-Wilk test of normality was employed in this research as the numerical means of assessing normality. To see the normal distribution of the test used, the significance value of the Shapiro-Wilk test must be higher than 0.05. The data is then said normal. However, if the significance value of the Shapiro-Wilk test is lower than 0.05. Therefore, the data significantly deviate from normal data distribution. The results of the Shapiro-Wilk test of normality concerning OPA and CTA in the experimental group were presented. It showed that the significance value of pre-test and post-test of OPA was .293 and .302, and the significance value of pre-test and post-test of CTA was .062 and .221. Meanwhile, the results of the Shapiro-Wilk test of normality concerning OPA and CTA in the control group revealed that significance value of pre-test and post-test of OPA was .510 and .176, and the significance value of pre-test and post-test of CTA was .115 and .083. From the results of the Shapiro-Wilk test of normality, in terms of OPA and CTA, in the experimental group and control group, it could be concluded that the data of pre-test and post-test of both groups were normally distributed as each significance value of both tests was higher than 0.05. To see the homogeneity of variances test of OPA and CTA in the experimental and control group, the Levene statistics (F test) was used in this research. There were 3 major points measured: (a) pre-test and post-test in the experimental group, (b) pre-test and post-test in the control group, and (c) post-test and post-test between both groups. The results of the Levene test for equality of variances showed that the significance level of the F test of pre-test and post-test, both OPA and CTA in the experimental group was .893 and .509. Meanwhile, the results of the Levene test for equality of variances showed that the significance level of the F test of pre-test and post-test, both OPA and CTA in the control group was .714 and .675. Since the all significance level of the F test was higher than 0.05, it could be stated that the data of the 3 major points were homogenous or equal.

Statistical Analyses
Two statistical analyses, progressive analysis, and difference analysis, were employed in this research: Paired sample t-test was used to see whether there was a significant improvement on EFL learners' oral proficiency and critical thinking achievements after being taught by using discussion task model, and independent samples t-test was used to see whether there was a significant mean difference in EFL learners' oral proficiency and critical thinking achievements after being taught by using discussion task model and those who were not.

Progressive analysis
To answer into research question 2 (RQ2), paired samples t-test was employed in this research where the data used were taken from the pre-test and post-test of the students' oral proficiency and critical thinking achievements within each group either experimental or control group. Then, the significant improvement of pre-test and post-test could be seen from the mean scores gained. In association with the results of progressive analysis of the pre-test and post-test of the students' oral proficiency achievement (OPA) within the experimental group, it showed that the mean scores, in terms of OPA, in the pretest and post-test of the experimental group were 12.00 and 17.75 with standard deviation (2.974 and 3.127). The mean difference gained within pre-test and post-test of the experimental group was 5.750 with standard deviation (0.910), the t-obtained gained was 28.244, and the significance level of OPA was .000 which it was lower than 0.05. Apart from that, the results of the progressive analysis of the pre-test and post-test of the students' critical thinking achievement (CTA) within the experimental group revealed that the mean scores, in terms of CTA, in the pre-test and post-test of experimental group were 11.00 and 14.00 with standard deviation (2.956 and 2.828). The mean difference gained within pre-test and post-test of the experimental group was 3.000 with standard deviation (0.973), the t-obtained was 13.784, and the significance level of CTA was .000 which it was lower than 0.05.
On the other group, the results of progressive analysis of the pre-test and post-test of the students' oral proficiency within the control group, it showed that the mean scores, in terms of OPA, in the pre-test and post-test of control group were 11.00 and 13.25 with standard deviation (3.112 and 3.177). The mean difference gained within pre-test and post-test of control group was 2.250 with standard deviation (1.118), the t-obtained gained was 9.000, and the significance level of OPA was .000 which was lower than 0.05. Additionally, the results of the progressive analysis of the pre-test and post-test of the students' critical thinking achievement (CTA) within the control group revealed that the mean scores, in terms of CTA, in the pre-test and post-test of control group were 10.75 and 11.70 with standard deviation (2.918 and 2.940). The mean difference gained within pre-test and post-test of the experimental group was 0.950 with standard deviation (0.510), the t-obtained was 8.324, and a significance level of CTA was .000 which it was lower than 0.05. From the results of paired samples t-test (Progressive analyses), it could be concluded that there were significant improvements made by the EFL learners, both OPA and CTA, in the experimental and control group after being taught by using discussion task model and those who are not as this could be seen from the t-obtained of OPA and CTA were higher than t-table.

Difference analyses
To answer into research question 3 (RQ3), independent samples t-test was employed in this research where the data used were taken from post-test of the students' oral profi-ciency and critical thinking achievements between experimental group and control group. To see the mean difference between the two groups, mean scores gained were used in this analysis. In accordance with the results of difference analysis of the post-test of the students' oral proficiency achievement (OPA) and critical thinking achievement (CTA) between the experimental group and control group, it showed that mean difference gained of OPA was 4.500, t-obtained was 4.515, and the sig. 2tailed was .000 (p<0.05). On the other point, mean difference gained of CTA was 2.300, t-obtained was 2.521, and the sig. 2tailed was .001 (p<0.05). By pondering the results of independent samples t-test (difference analyses), it could then be concluded that there were significant mean difference on EFL learners' oral proficiency and critical thinking achievements after being taught by using discussion task model and those who are not as the t-obtained was higher than t-table and the p values were lower than 0.05 (.000<0.05 and .001<0.05).

Stepwise Regression Analyses
To answer into research question 4 (RQ4), Multiple regression analyses were employed in this research in which the percentage analysis of the contribution of oral proficiency (aspects) and critical thinking (aspects) toward the oral proficiency (total) and critical thinking (total) was stepwise regressed. In Table 3, the results of stepwise regression analysis of oral proficiency achievement (OPA) and critical thinking achievement (CTA) were presented accordingly. Comprehension, vocabulary, pronunciation grammar, and fluency were categorized in oral proficiency achievement. Each oral proficiency (aspects) was stepwise regressed toward oral proficiency (total). The results revealed that the aspect of OPA that gave the highest contributions. They were fluency (82.9%) and comprehension (7.4%). The smallest contributions were vocabulary (3.6%), pronunciation (3.2%), and grammar (2.9%). Meanwhile, the highest contributions among the aspect of CTA were context (73.7%) and assumption (14%). The smallest contributions among them were evidence (4.1%), implication (3.3%), issues (2%), and perspective (2.8%). Since the sig. 2tailed of each oral proficiency (aspects) and critical thinking (aspects) was less than 0.05, it could be concluded that oral proficiency (aspects) and critical thinking (aspects) had a significant contribution toward the oral proficiency (total) and critical thinking (total) . Table 3 Stepwise Regression Analyses of oral proficiency (aspects) and critical thinking (aspects) toward the oral proficiency (total) and critical thinking (total)

Likert Scale Analyses
To answer into research question 5 (RQ5), the Likert Scale analysis was used to see the percentage of the EFL learners' perception of Likert items on the implementation of discussion task model toward their oral proficiency and critical thinking achievements. A set of close-ended questionnaire was constructed to find out whether the EFL learners have positive or negative responses toward the discussion task model instruction on their oral proficiency and critical thinking achievements. Therefore, strongly disagree and disagree belong to the negative responses and strongly agree and agree belong to the positive responses.
The results of the Likert scale analyses showed that 18 (90%) students gave positive responses that they enjoy having discussion in the classroom, 14 (70%) students gave positive responses that they are motivated to learn speaking with discussion task model, that 8 (40%) students gave positive responses that they find difficulties in learning speaking before using discussion task model, 14 (70%) students gave positive responses that they feel more confident in speaking in the classroom, 11 students (55%) gave positive responses that they can understand at ease about the issue in the discussion, 16 (80%) students gave positive responses that they earn many new words to use when speaking, 17 (85%) students gave positive responses that their pronunciation is getting better, 8 (40%) students gave positive responses that they speak more grammatically when having discussion, 8 (40%) students gave positive responses that they speak. English more fluently and bravely, 13 (65%) students gave positive responses that they feel more critical after having discussion, 8 (40%) students gave positive responses that they can identify and explain the issue well, 13 (65%) students gave positive responses that they recognize the context of the issue spoken in the discussion, 8 (40%) students gave positive responses that they formulate a clear and precise personal and others 'perspectives, 7 (35%) students gave positive responses that they can evaluate implicit and explicit assumptions during discussion, 11 (55%) students gave positive responses that they identify and evaluate thoroughly the evidence of arguments proposed, 11 (55%) students gave positive responses that they can draw and evaluate the implications, conclusions, and consequences. From the results of the descriptive analyses of EFL learners' perception of the implementation of discussion task model toward their oral proficiency and critical thinking achievements, it showed that the EFL learners gave positive responses upon the statement proposed in which it was much greater than those who gave negative responses.

Discussion
After the research findings were descriptively and statistically discussed, the interpretation could be drawn in relation to the implementation of discussion task model toward oral proficiency and critical thinking achievements in the EFL classroom. First research findings showed that the EFL learners' OPA and CTA level in the experimental group were in a good category while the EFL learners' OPA and CTA level in the control group were in the average category. Second research findings showed that there was a significant improvement on the EFL learners' oral proficiency and critical thinking achievements in both groups. However, the t-obtained in the experimental was much greater than that in the control group. Third research findings showed that there was a significant mean difference in the EFL learners' oral proficiency and critical thinking achievements after being taught by using discussion task model and those who were not. Forth research findings showed that the oral proficiency (aspects) and critical thinking (aspects) had a significant contribution toward oral proficiency (total) and critical thinking (total) after the applied intervention. Fifth research findings showed that EFL learners gave positive perception of the implementation of discussion task model toward their oral proficiency and critical thinking achievements.
There were some reasons why discussion task model gave positive improvement toward oral proficiency and critical thinking achievements in the EFL classroom. First reason, this is caused by the students' activity before the discussion started that they should have the brainstorming with the teammates related to the discussion topic given in the classroom where they should find the arguments and data to support their personal substantives. Therefore, the students in person had arguments and opinion to speak actively and bravely and this activity could allow the students to interact and to exchange their ideas and thoughts among their peers. This statement is highlighted by Adewuya (2003) stated that the activity of brainstorming in the discussion based teaching can be a step for students to make opinions or viewpoint clarification, contribution, as well as feelings exploration. Apart from that, Stephen & Stephen (2005) claimed that discussion is the process of interaction in which it can assist the speakers and listeners to develop their knowledge and create mutual understanding.
In addition, the second reason, when having a discussion the students had a chance to ask for help to the lecturer about the case, unfamiliar vocabulary, pronunciation and the like because in the classroom the function of the lecturer becomes the facilitator. Hence, lecturer-students interaction exists during the discussion. This is line with Ornstein and Lasley (2000), getting the EFL learners into small group allowed them to create active engagement among their counterpart and lecturer as well as develop their cooperation and social life skills.
The third reason, the students gave counter opinions and proper responses actively when they answer the objection and the question during the on-going substantive delivery. This situation really had the students attempt to speak their arguments accordingly and to listen to question and to think the way out to find the answer at once. This statement is supported by Abdu-Raheem (2011) said that teaching with discussion can activate the teachers and learners engagement in the thinking process and promote the learners' social life skill. In addition, a discussion also can promote the EFL learners' tolerance to challenge others' perspective, increase their communication and critical thinking skill (Eggen & Kauchak, 2001).
The forth reason, since the highest contributions of OPA and CTA were Fluency and Context and the lowest contributions of OPA and CTA were grammar and perspective. On one hand, it is believed that the student was well-prepared before the discussion started so that they were very fluent when they had a chance to speak out before the public. On the other hand, when the students spoke their arguments fluently and confidently, consequently, they ignored to speak grammatically correct English as they just wanted to convince their ideas and thoughts to the opponent team. Regarding the CTA, it is believed that when the students saw the discussion topic, they tried hard to find the context of the issue related to the opinions to be outspoken and divide the context with their teammates. With respect to the perspective as the lowest contribution, when the students wanted to bring the perspectives pertaining to the topic give, they gave vague and unclear perspectives while their opponent team could give a better perspective which closer to the case brought in the discussion.
The fifth reason, the students gave positive responses upon the implementation of discussion task model on their oral proficiency and critical thinking achievements. This is caused by the student -student interaction and lecturer -student interaction before, while and after the discussion in which the students were assigned to have good preparation in terms of the understanding of the discussion topic in order that they could participate actively and confidently in the discussion. These also allowed the students to speak naturally and bravely as well as to think critically what to say before having substantives delivery. Having the same thought, concerning on the oral proficiency and critical thinking, Abdu-Raheem (2011) pointed out that the teaching with discussion can activate the teachers and learners engagement in the thinking process and promote the learners' social life skill.Yusuf & Al-Banawi also added that discussion and in-class activities can activate the learners' involvement and promote their learning desire. Having the same perspective, a discussion is a prominent teaching method that the teachers can employ as this can create active learning, developing, understanding, and cultivating critical thinking of the learners (Ikuenobe, 2002;Okolo, Ferretti, & Mac Arthur, 2007;Sautter, 2007). Gage and Berliner (2002) claim that-a discussion based instruction can develop EFL learners' critical thinking, reasoning, and democratic participation. While discussion session, the learners can state and formulate their viewpoints, accept an evaluation from others, and identify the unclear standpoints (Hadjioannou, 2007). Ultimately, research findings showed that discussion task model statistically gave significant improvement on the EFL learners' oral proficiency and critical thinking achievements. This is proven by the high results of the descriptive statistics analyses, statistical analyses, stepwise regression analyses, and positive perception made by the EFL learners after the intervention employed.

Conclusion
Based on the research findings and discussion, the conclusion and suggestion were drawn as follows: First, the OPA and CTA level in the experimental group were in the good category in which it was much greater than that in the control group. Second, statistically discussion task model gave significant improvement on the EFL learners' oral proficiency and critical thinking achievements in both groups. Third, there was a significant mean difference in the EFL learners' oral proficiency and critical thinking achievements after being taught by using discussion task model and those who were not. Forth, there was a significant contribution on oral proficiency (aspects) and critical thinking (aspects) toward the oral proficiency (total) and critical thinking (total) where the highest contribution in the aspect of OPA and CTA was fluency and context. Fifth, the EFL learners gave a positive perception on the implementation of discussion task model toward their oral proficiency and critical thinking achievements. Hence, it could be concluded that using a discussion task model in EFL classroom significantly improved the EFL learners' oral proficiency and critical thinking achievements.
In relation to the conclusion above, some pedagogical implications are recommended to the English lecturers, students, stakeholder, and future researchers. First, the English lecturer should include the discussion task model, as the instructional strategy, in teaching oral skill and critical thinking as this is successfully proven that it can promote the EFL learners' oral proficiency and hone their critical thinking. Second, the students should be well guided in EFL classroom in terms of the implementation of discussion task model as this can practice the lecturer-student interaction and student-student interaction in constructing the case and arguments and build the social skills. Third, the stakeholder should include this learning model in the teaching curriculum as the alternative, if not primary, so that there will be a variety of the instructional model in developing the EFL learners' oral proficiency and critical thinking cultivation. Forth, the future researchers are recommended to conduct the similar research with different skills focus and to use the interview as the additional instrument so that the triangulation analyses generate much deeper research findings pertaining to the implementation of discussion task model in the EFL classroom.