To determine whether or not the article is suitable for use in research, the researcher sets the inclusion criteria and exclusion criteria as written in Table 1.

The researcher conducted the initial search procedure in the first stage (Identification). This initial search procedure aims to gather suitable articles to answer relevant additional research questions and references. The researcher searched for articles on the Scopus and Eric databases. In the Scopus database, researchers search for articles assisted by the Publish or Perish application. Researchers search for articles in the Eric database directly on the website. The keyword used during the article search is "Assessment in Mathematics Education." In this first stage (Identification), the

researcher obtained 400 articles on Scopus and 1133 on Eric, so the total number of articles obtained was 1,533.

After the article is obtained, the researcher's next step is the Screening process. First, the researcher checks for article duplication. Based on the researcher's findings, 11 duplicate articles and 39 articles not open access were found, so 50 articles had to be discarded in the first screening process, leaving 1483 articles.

Then, 1483 articles were entered for the second screening process: selection based on title and abstract. At the screening stage, based on title and abstract, 1193 articles were eliminated because 1) 6 articles were not journals or proceedings, 2) 6 articles were not written in English, and 3) 1174 articles were not related to the practice of mathematical assessment in the Southeast Asian region, leaving 290 articles to be included in the Eligibility Criteria stage.

After the screening process, the researcher continued with the screening process based on the Eligibility Criteria. At this stage, out of the remaining 290 articles, 251 were eliminated for the following reasons: 1) 146 articles were not conducted in Southeast Asia, and 2) 105 articles were not conducted in the Southeast Asian region. So, the final articles used for this research are 39 articles, which can be seen in more detail in Figure 1.

After a screening process that resulted in 39 final articles, data analysis was carried out systematically. The process of extracting and categorizing relevant data from each article follows the set of research questions (RQ) that have been determined, as follows: a) First, the articles were grouped by year of publication to identify the distribution of the research year of the article, the results of which are then presented in Figure 2, b) Second, articles are classified based on the country where the research was conducted to find out the map of the distribution of articles in the Southeast Asian region, the results of which are then presented in Figure 3, c) To answer RQ1, the articles were categorized based on the type of assessment discussed in the article. The data to answer RQ 1 are then presented in Table 2 and visualized in Figure 4, d) To answer RQ2, the research methodology used in each article was identified and grouped. The results of the grouping of research methods are presented in Table 3 and visualized in Figure 5, e) To answer RQ3, the research objectives of each article were analyzed and then classified into 9 categories as shown in Table 4. In addition, the researcher also analyzed the research objectives and research methodology used from each article whose results are presented in Table 5, f) To answer RQ4, the articles were grouped based on the level of education that became the focus of the research. After that, the distribution of education levels for each article is illustrated in Figure 6, g). To answer RQ5, the use of digital technology in assessment practices in each article was analyzed and recorded in detail. The findings are summarized in Table 6. This process of grouping and categorization is done manually with the help of Microsoft excel to ensure accuracy and ease in organizing the data, which then becomes the basis for the presentation of results and discussions.

In this study, as many as 39 articles were analyzed. The articles consist of 33 (85%) journal articles and 6 (15%) proceedings from 2015-2024, with the annual acquisition of articles listed in Figure 2. Figure 2 shows that the number of articles was relatively stagnant, namely two articles, from 2015 to 2017. Then, it increased significantly from 2018 to 2020. However, it dropped significantly in 2021, with the total data on the number of articles obtained by each country listed in Figure 3.

The overview of the distribution of articles in Figure 3 explains that Indonesia is the most significant contributor of articles with 18 articles, followed by Malaysia with nine articles, Thailand with four articles, the Philippines with three articles, Vietnam with two articles, Singapore and Brunei Darussalam with 1 article each and finally Timor Leste, Laos, Cambodia, and Myanmar with zero articles each. So, the total number of articles is 39. Furthermore, from the 39 articles, answers to research questions are obtained as follows:

Assessment based on its function can be divided into three types: Assessment of Learning, Assessment for Learning, and Assessment as Learning (Budiono & Hatip, 2023). Assessment of Learning (Summative Assessment) is an assessment that aims to validate learning and reports for parents and students about student progress in school (Earl, 2003); assessment for Learning (AfL) is an assessment designed to make each student's understanding and knowledge "visible" (Hattie, 2012), so that teachers can decide what they can do to help students progress (Magbeyi et al., XXXX). Examples of this Assessment for Learning are Formative and Diagnostic assessments (Asesmen, 2021). Meanwhile, Assessment as Learning (AaL) occurs when students act as assessors for themselves. There are two types of Assessment as Learning (AaL): Self-Assessment and Peer-Assessment (Asesmen, 2021).

Based on this classification, the researcher divided the classification of types of assessments into six parts, as shown in Table 2. So, based on Table 2, the percentage of each type of assessment is obtained, as shown in Figure 4. Based on the percentage of the types of assessments in Figure 4, it can be known that Assessment for Learning (AfL) is the most researched assessment in the Southeast Asian region, with 15 articles (47%) of the total. Of the 15 Assessment for Learning articles, 10 discuss formative assessment, with five from Indonesia. These findings confirm the opinion of Sudakova et al. (2022), Which states that "The recent assessment trend has migrated from summative assessment to formative assessment."

Based on Table 3, the type of qualitative research gets as many as 16 (41%) articles, the type of quantitative research gets as many as 18 (46%), and the mixed method gets as many as 5 (13%) articles. So, the most widely used type of research is quantitative research, with as many as 18 (46%) articles, which can be seen more clearly inFigure 5.

Table 4 shows that articles that aim for evaluation are trending in assessment research in the Southeast Asian region, with as many as nine articles. In contrast, articles that discuss the readiness of mathematics teachers in carrying out assessments in the classroom, knowledge of HOTS question patterns, and analysis of the need for assessment tools are still very minimal, with 1 article each. Then, if the research objectives are related to the type of research, the results are obtained in Table 5.

Based on Table 5, we know that research that discusses the impact of assessment is widely used in the mixed method (as many as two articles), qualitative research is widely used to determine the implementation of assessment (7 articles), and in quantitative methods, the development of products/instruments for assessment is a widely used research objective (6 articles).

Figure 6 shows that Junior High Schools and Undergraduate education are the most researched educational levels in Southeast Asian countries.

Based on the systematic literature review, eight articles discuss the form of technology utilization for assessment purposes. Some of the technologies discussed in the article found by the researcher are shown in Table 6.

The study's findings show that Assessment for Learning, especially formative assessment, is the most researched in the Southeast Asian region. This is because, according to (Black & Wiliam, 2009), Formative assessment consists of five key strategies: a) Defining and communicating learning objectives and success criteria; b) Facilitating meaningful classroom discussions and tasks that showcase student comprehension; c) Offering feedback that helps students progress; d) Encouraging students to teach one another; and e) Empowering students to take ownership of their learning. Therefore, formative assessment should be thoroughly studied and implemented in education.

However, teachers still face many challenges in implementing formative assessments in Southeast Asia. An example is Timor Leste, a country. Research conducted by (Costa Akoyt, 2024) against 50 primary school teachers in Timor Leste shows that 92% of teachers agree with the statement, "I believe that formative assessment can promote learning, and I want to learn more about formative assessment practices." However, of the 50 teachers, 56% claimed that they rarely used formative assessments in their classrooms. In addition to formative assessments, another large percentage, 82% of teachers, also reported that they rarely use peer assessment and self-assessment in their classrooms.

In Myanmar, research conducted by (Oo et al., 2024) found a need to improve assessment literacy among Myanmar teachers. This is because teachers in Myanmar still have limited knowledge

related to assessment. They said the purpose of conducting assessments is to measure student learning outcomes, not to improve the teaching and learning process in the classroom (Lim, 2024).

Research in Vietnam by (Le, 2021) involving 579 junior and senior students majoring in basic education at 11 universities in Vietnam concluded that the capacity for formative assessment in teaching mathematics in elementary school departments is still limited. These limitations are mainly

limitations on the understanding of methods, tools, and forms of formative assessment in teaching mathematics in elementary schools. Based on some of these studies, the knowledge and skills of teachers and prospective teachers, especially in mathematics education in the Southeast Asian region, must be improved.

Meanwhile, research related to Assessment as learning is the least researched research in research related to assessment in mathematics learning in Southeast Asia. In fact, assessment as learning is an assessment that aims to enable students to become independent students and requires students to be aware of what is required of them and monitor and evaluate their own learning during the learning process (Yan & Boud, 2022). So that with the information obtained, they can organize their learning to achieve the goals they have set in advance (Yan & Boud, 2022). Therefore, because this assessment as learning is an important assessment to be carried out by teachers and students of prospective mathematics teachers. Therefore, mathematics teachers and prospective mathematics teacher students are expected to be able to have the knowledge and skills to carry out the assessment of learning in mathematics classes.

The type of quantitative research is also the most researched by researchers in the Southeast Asian region, with 19 articles, of which seven articles aim to develop products and use assessment tools for assessment purposes. The development and assessment tools used include: a). Development of statistics textbooks supported by ICT and portfolio assessment (Hendikawati, 2016); b). Leveraging Google Forms with IATA software (Hau, 2020); c). Development of Assessment for Learning Humanistic model (AfL-H) (Winarno et al., 2019); d). Development of assessment instruments using polytome responses (Sutiarso et al., 2022); e). Development of numeracy test instruments for Minimum Competency Assessment (MCA) (Purnomo et al., 2022); f). Development of a Diagnostic Cognitive Assessment Tool to assess students' mastery of the concept of "Parallel and Straight Tegal" lines (Chin et al., 2022); and finally, g). Development of statistical thinking frameworks and assessment tools that involve statistical thinking (Hooi Lian & Yew, 2023).

Regarding the level of education, we all know that junior high school and undergraduate are the levels of education that are widely studied about the assessment of mathematics classes in the Southeast Asian region, with 10 articles each. At the junior high school level, assessment for evaluation is the most studied research objective, consisting of four articles. Undergraduate research to find out the perception of students and teachers related to assessment is the most studied research objective, with as many as four articles.

An interesting study from Indonesia is worth discussing, according to the researcher. This research was conducted by (Khaira, 2020). This study was conducted at a Junior High School to explore how teachers support students, the role of teachers in assisting slow learners in mathematics, and the challenges faced by these students. One key issue slow learner encounter is the lack of opportunities to solve problems in front of the class. Additionally, they struggle to keep up with their peers due to slower comprehension and learning processes and face time constraints during class lessons. A potential solution for educators or future educators is the implementation of clinical teaching. This approach involves a tailored learning assessment to assist students with learning difficulties (K.T.S.P., 2009). Clinical teaching aims to tailor students' learning experiences to the unique needs of students who experience learning difficulties. In other words, this clinical teaching ensures that all learners, including those needing more time, can achieve the necessary clinical competencies and feel supported throughout their educational journey.

Another interesting thing that the researcher found is, based on PISA data in Table 7. The countries of Indonesia, Malaysia, Thailand, the Philippines, Vietnam, and Brunei Darussalam, which have more articles related to assessments in mathematics classes, still have a lower average PISA score compared to Singapore, which only has 1 article. This shows that many or few articles related to assessments have not fully demonstrated the level of assessment quality in the country. There may be other factors that have an impact as well.

If we want to compare Southeast Asian countries with China, which has a lower OECD average than Singapore (with an average OECD score of 552). The average score of OECD countries such as Vietnam, Brunei Darussalam, Malaysia, Thailand, Indonesia, the Philippines and Cambodia is still quite far behind. This is because in China, the use of technology for assessment purposes is also quite good, for example in research conducted by (Chen et al., 2020).They have used interactive whiteboards as one of the technologies to carry out formative assessments.

Of the 39 articles studied in this study, there are only 8 articles that discuss digital technology used to support the practice of mathematics assessment in the Southeast Asian region. Countries that discuss technology to support the practice of mathematical assessment are the Philippines, Indonesia, Malaysia, Vietnam and the most widely used digital media to support assessment practice is Quizizz which is discussed in 2 articles out of 39 articles namely research conducted by (Saleh & Sulaiman, 2019) and (Rahman et al., 2019-06-25). Therefore, more research discusses digital technology used to support the practice of mathematics assessment in the Southeast Asian region and more effort is needed to make prospective teachers and teachers of mathematics to utilize technology in assessment in mathematics learning.

Based on the results of the above research, the research gap for researchers in the future is as follows:

Based on the research and discussion results, several things can be concluded in this study: 1) Assessment for Learning (AfL) is the most studied assessment topic in Southeast Asian countries, with 15 articles. Of the 15 articles, 10 discuss formative assessments, and of the 15 articles, five are from Indonesia. Assessment for Learning (AfL) is one of the most fundamental assessments in the classroom. Through this Assessment for Learning (AfL), teachers can discover things that must be improved during learning. 2) Quantitative research is a research methodology that dominates articles related to mathematical assessment in the Southeast Asian region, with a total of 18 articles and seven articles discussing the development of products/instruments for assessment. 3) Assessment for evaluation is the most dominating research objective in articles related to mathematics assessment in the Southeast Asian region, with as many as nine articles. 4) Junior high school and undergraduate education levels are the most studied levels of education in research related to mathematics assessment in the Southeast Asian region, with each level discussed in 10 articles. 5) The Quizizz application is the most widely used assessment practice in mathematics classes in the Southeast Asian region, with two articles that discuss the application. The practical implication of these findings highlights the need for study about topic of assessment as learning in mathematics education. More effort is needed to make prospective teachers and teachers of mathematics to utilize technology in assessment in mathematics learning.

This research was translated using a translation tool on Microsoft Word by checking grammar using the Grammarly application.