PROFILE OF STUDENTS’ COMPUTATIONAL THINKING BASED ON SELF- REGULATED LEARNING IN COMPLETING BEBRAS TASKS

Bebras task is a problem-solving problem that integrates computational thinking in it, which the stages in computational thinking consist of: decomposition, abstraction, algorithm, and pattern recognition. This study aims to describe the profile of students’ computational thinking based on the level of self-regulated learning in completing bebras task. This study is a qualitative-descriptive study with three research subjects based on the level of students’ self-regulated learning, namely high self-regulated learning, medium self-regulated learning, and low self-regulated learning. The results of this study indicate that students with different levels of selfregulated learning have different computational thinking ability in completing bebras task. Student with high level of self-regulated learning can reach the stages of decomposition, abstraction, algorithm, and pattern recognition. Student with medium level of self-regulated learning can reach the stages of decomposition, absraction, and algorithm. Student with low level of self-regulated learning can reach the stage of decomposition only. Student with low level of self-regulated learning do not yet reflect independence in learning.


INTRODUCTION
The Industrial Revolution 4.0 brings education into The Age of Knowledge, namely the acceleration of increasing knowledge marked by the application of media and technology (Mawardi, 2016: 65). So that requires humans to adapt to a mindset in accordance with the current developments and compete globally. So, ensuring students have the skills to think and innovate in solving problems becomes an urgency for education. As the considerations contained in the Law of the Republic of Indonesia No. 20 of 2003 concerning the National Education System, that education must be able to ensure equal opportunities for education,

◼ 41
Profile of students' computational thinking based on self-regulated learning in completing bebras tasks Nuraisa, Saleh, Raharjo increase the quality and relevance and efficiency of education management to face challenges in accordance with the changing demands of local, national, and global life. So it is necessary to do educational renewal in a planned, directed and sustainable manner. Based on this, education must be more responsive in developing quality in the midst of the times and preparing an appropriate educational framework.
Problem solving is an important component of the mathematics learning curriculum, both in activities and in the learning process to solve routine and non-routine problems (Telaumbanua, Sinaga, and Surya, 2017: 74). This is because the problem-solving process requires the use of knowledge and skills that are already owned in routine problem-solving processes to be applied in solving non-routine problems. According to Kusumawardani et al., problem solving does not only requires the ability to count for the solutions, but requires more ability such as to reason, so students can find out the meaning of the problem presented (Susanti and Taufik, 2021: 23). In addition, through the process of non-routine problem solving, aspects of mathematics learning can be developed, such as pattern recognition, generalization, and mathematical communication (Kusumaningtyas, 2017). But in fact, based on the value of daily math test, it shows that junior high school students still have difficulty solving non-routine problems, marked by students tend to be reluctant to solve questions that they think are rarely encountered and students have not been able to express creative ideas about the problems presented. In view of the importance of problem-solving abilities in non-routine problems, there are problem-solving techniques whose application is very broad and complex, namely through computational thinking.
Computational thinking is the new literacy of the 21st century. It enables you to bend computation to your needs (Wing, 2010: 3). Computational thinking is closely related to computational theory. According to Simonson, computational theory is an abstraction program about what can be calculated (Alfina, 2017:3) However, computational thinking is not only focused on solving problem, but more focused on how to solve it the problem (Nuraisa et al., 2019: 1). Computational thinking is the thought processes in formulating problems and solutions, so the solutions can be represented in a effectively form (Grover & Pea, 2013: 39). Computational thinking is the ability to think in solving problems with various levels of abstraction and based on indicators of computational thinking, including: decomposition, abstraction, algorithms, and pattern recognition. Although there are four indicators, computational thinking is synonymous with the use of decomposition and abstraction. In accordance with the characteristics of computational thinking that formulates problems through solving the information presented to be simpler and still structured. This is useful for focusing the algorithm in obtaining a solution. So, complex problems will be solved easily, efficiently, and creatively through computational thinking.
However, in reality the learning process that takes place in Indonesia has not integrated computational thinking (CT) into subjects, such as mathematics. Meanwhile, Indonesia itself already has problem solving problems that include computational thinking, namely Bebras Task. Bebras Task is a problem-solving problem related to informatics that focuses on logic and mathematics. According to Dagiene and Sentance (2016) tasks are the most important component for developing students' computational thinking. Bebras Task questions are presented along with pictures to attract attention and stimulate students to complete them.
In addition, Bebras Tasks are used in international standard competitions, namely "Bebras Challenge". The purpose of holding the "Bebras Challenge" is to promote and encourage the development of computational thinking (Tim Olimpiade Komputer Indonesia, 2018).
In addition, one thing that needs to be paid attention to in computational thinking skills is self-regulated learning. Self-regulated learning is an effort to direct self-initiative and motivation in the learning process to achieve optimal learning outcomes. Self-regulated learning has a significant effect on the learning process and learning achievement (Kristiyani, 2016: 11). According to Knain and Turmo, self-regulated learning is a dynamic process of building knowledge, skills, and attitudes when learning a specific context. To build knowledge in the process learning does not only require learning strategies, learning experiences, and applying the knowledge, but must be able to reflect/evaluate learning activities (Amir, Z., 2015: 168-169). Computational thinking is seen as a goal-directed process and uses heuristic reasoning to obtain solutions. Heuristic reasoning includes activities, such as planning, learning, dealing with uncertainty, and the search process (Wing, 2006: 34). Activities in the heuristic reasoning process are consistent with the components in self-regulated learning.
This suggests that the relationship between self-regulated learning and computational thinking processes allows the use of concepts, components, and strategies of self-regulated learning as a framework for improving computational thinking skills. This study describes in Prima: Jurnal Pendidikan Matematika

◼ 43
Profile of students' computational thinking based on self-regulated learning in completing bebras tasks Nuraisa, Saleh, Raharjo detail the relationship between students' self-regulated learning and their computational thinking ability which is shown through problem solving skills in the form of bebras task.

METHODS
This research is a qualitative-descriptive study. This study aims to describe the profile of the 8th grade students' computational thinking of SMP Negeri 17 Tangerang based on selfregulated learning, from all of the students, there are 3 students only who had met the criteria of subject. The data were collected by self-regulated learning questionnaires, bebras task as a computational thinking test, and unstructured interviews.
Self-regulated learning questionnaires was adopted by Saepulloh (Hendriana, H., Rohaeti, E. E., Sumarmo, 2018: 244-245). The questionnaire was used to obtain scores and determine the categories of students' self-regulated learning. The questionnaire consists of 28 statements with 4 answer choices and using Likert scale. The research subjects can be seen in Table 1 below. Determining the level of self-regulated learning is to get specific difference that will be seen from how students solve problems, including planning to evaluating/re-checking the solution. The number of bebras task questions in this study were 4 and were in the form of essays. Each question contains four indicators of computational thinking, namely decomposition, abstraction, algorithms, and pattern recognition. The indicators of bebras task questions in this study can be seen in Table 2 below. Students can build and combine information into structured networks.

C5 (Synthesis) Difficult
In addition, the interview in this study is unstructured interview conducted with the aim of obtain deeper data students' computational thinking ability in completing Bebras Task. The questions in the interview are in the form of questions that clarify the indicators of computational thinking achieved by students that can not be seen from the results of the test they do. So, to find out how students can solve problems, it needs to be found through interviews.

RESULTS AND DISCUSSION
Based on the results of the self-regulated learning questionnaire, there are 3 levels of

The Computational Thinking Profile of Student with Low Self-Regulated Learning Computational
Thinking's Indicators

Number of Conclusions Question-1
Question-2 Question-3 Question-4  represent the process of regulating their learning by demonstrating their ability to diagnose needs (referring to students understanding what is needed in solving problems), have persistance, and performing cognitive strategies, especially rehearsal and elaboration in the completion process, so as to create and identify patterns. It is following the results of research by Yanti and Surya (2017) which states that self-regulated learning (independent learning) affects the quality of learning itself, which is shown at the level of achievement/student learning outcomes. The better process of regulating the learning process, the better the learning outcomes obtained.
SRL2 can do planning and implementation quite well in the completion process.
However, the behavior is not careful, both in the process and in make conclusions (evaluation/reflection phase). Lack of activities to evaluate the process affects the making of conclusion and the results obtained by the settlement. It is following the results of research by Yanti and Surya (2017) which states that self-regulated learning (independent learning) affects the quality of learning itself, which is shown at the level of achievement/student learning outcomes. SRL2 shows that the lack of evaluation activities carried out also affects the learning outcomes that are owned.
SRL3 achieved the decomposition indicator only. At the abstraction, SRL3 cannot determine the correct representation of the solution (planning phase), this is because SRL3 is unable to diagnose what informations are needed in completing bebras task. This affects the automation of the solution that is carried out is also incorrect. SRL3 do a solution based on trial and error, but only once, then do not re-checking. So that if the answer is not found, SRL3 think the problem solving has been completed. This shows that SRL3 do not see learning Profile of students' computational thinking based on self-regulated learning in completing bebras tasks Nuraisa, Saleh, Raharjo difficulties as challenges, so they can easily give up when they experience difficulties in learning. It is following the results of research by Hamundu, Sudia, and Samparadja (2017: 157) which states that students with low self-regulated learning have a feeling of boredom, give up easily, prefer to choose a more instant way and use less careful thinking, take a long time, lack willingness to examine problems and feel complicated to identify. SRL3 do not yet reflect independence in learning.

CONCLUSION
Besides being applicable to various problem contexts, computational thinking is useful for practice logic and pattern recognition for students in solve non-routine problems that require deeper analysis and thinking. Computational thinking is important to be included in mathematics learning. The recommendations for further research are the need for research in the form of appropriate learning methods to teach computational thinking to students and the development of computational learning instruments, especially in mathematics subject and learning.

ACKNOWLEDGMENTS
I would like to express my special thank of gratitude to my lecturers who gave me the opportunity to explore the topics and helped me in doing this research. Secondly, I also would like to thank my parents and brother who helped me a lot in finishing this research. I am really thankful to them.