Developing mathematics problems using local wisdom context of Maluku to improve students’ numeracy

Students' difficulty in solving numeracy problems is a problem that teachers must pay attention. Therefore, it needs media to help students improve their numeracy ability. The purpose of this study is to develop mathematics problems based on PISA (Programme for International Student Assessment) content and use local wisdom context in Maluku that obtained validity, practically, and effectively. Mathematics problems developed are expected to improve the numeration ability of junior high school students in Maluku. The study was research and development based on the Plomp model. The subject consists of 40 students (15 males and 25 females). The result showed that mathematics problems met validity criteria. The validity was tested using the Rasch model that consists of suitable, reliability, difficulty, and differentiation items. Then, it was validated based on experts in content, league, and media. All of them met validity criteria. Mathematics problems had met practically based on the user experience questionnaire (UEQ) that consisted of interest, clarity, efficiency, accuracy, stimulation, novelty, and prospered based on the system usability scale (SUS). Mathematics problems improve the student's numeration ability, as shown based on N-Gain 0.36. Thus, it effectively improves the numeracy ability of students in Maluku.


Introduction
Indonesia is one of the countries that participates every three years in PISA activities, starting from 2000 until the last PISA assessment in 2018. Based on participation data in 2018, Indonesia was ranked 70th for science skills, while literacy and math skills were ranked 72 out of 79 countries (Schleicher, 2019). The government has made various efforts to improve the ranking of Indonesian students on PISA assessments; one of the programs carried out by the government is the National Literacy Movement with Numeracy Literacy content. In addition, the Minister of Education has abolished the National Examination, which was turned into a Minimum Competency Assessment and character survey. The problems presented at the Minimum Competency Assessment consist of two types, namely literacy, and numeracy.
Literacy and numeracy are very important because they can activate the ability to analyze, build arguments, interpret, and communicate ideas or ideas effectively in a problem-solving process (Govorova et al., 2020;Stacey & Turner, 2015;Tai & Lin, 2015). The problems or problems presented at Minimum Competency Assessment have similarities with PISA problems .
The cause of the low numeracy of students in Indonesia is that students are still not used to working on PISA standard problems (Charmila et al., 2016;Dasaprawira, 2019;. In addition, the problems presented in the PISA problem have a general context because the problem presented is the same problem and then translated into the language of each country (Stacey & Turner, 2015), whereas students will be easier to solve problems in the context of the problem presented has been experienced or seen directly by students (Jonassen, 2010;Van den Heuvel-Panhuizen & Drijvers, 2020).
One of the contexts that can bring students closer to everyday life is the context of culture and local wisdom (Hasanah, 2017;Retnowati & UN, 2010;Supiyati & Hanum, 2019). In addition to bringing students closer to daily life problems, local culture is also very appropriate as a medium in instilling character education (Samo, 2018;Suradi, 2018). For this reason, it is essential to introduce children related to culture through learning mathematics in school, especially problems related to numeracy ability.
National Literacy Movement activities in 2021 get obstacles, especially implementation in the field because of the extraordinary events of the COVID-19 pandemic that forces all forms of learning activities to be diverted using online media based on the Ministry of Education Letter No. 4 of 2020. The results of initial observations in Ambon City showed that students' numeracy ability is still very low. From 40 students who were given random numeracy questions, only four students were able to answer the problem appropriately, and the rest had difficulty either understanding the problem, finding the right problem-solving strategy, or misusing the principle/concept in solving the given problem. Most students find it difficult to work on numeracy problems because they do not understand the context of the given problem.
In addition, some research shows that students' numeracy ability is still very low based on students' inability to understand problems, transform problems, and use appropriate concepts in solving problems (Aliifah, 2020;Ekowati et al., 2019;Puspita & Ahda, 2020).
The context of local wisdom is chosen based on the that students will easily solve a given problem if they understand the context of the problem presented (Mungmachon, 2012;Uge et al., 2019). In addition, the problem with PISA consists of a social context that contains the local cultural content of a society (Johar, 2012;Stacey, 2015). The local cultural context of Maluku is very diverse and numerous (Andaya, 2018;Latinis & Stark, 2005), thus allowing it to be used as a context for problems to be developed. Students are considered to need to have numeracy ability so that students can solve their life problems in the future; of course, the problem is not far from their lives (Tan, 2020) Using mathematics problem-based PISA content and the local wisdom context of Maluku, students can practice working on existing numeracy problems and being used as media for evaluating numeracy ability following the function used as a media of practice to solve problems and evaluation (Amtmann et al., 2020).
Several previous studies that used to develop PISA content mathematics problems, including using change and relationship content (Hasibuan & Fauzi, 2019;, using shape and space content (Efriani & Putri, 2019;Nasution et al., 2019), using quantity content (Murtiyasa et al., 2018;Nusantara, 2021), and uses uncertainty and data content (Nizar & Putri, 2018;Yansen & Putri, 2019). Of the few existing studies, no one has simultaneously developed a mathematical problem using four PISA contents. In addition, some studies develop mathematical problems using the context of local wisdom, including using the context of local Javanese wisdom (Machromah & Sari, 2021;Mulyatna et al., 2021), using the context of Jambi (Charmila et al., 2016), Palembang (Mayari, 2020), Malang (Pratiwi et al., 2020), Riau (Zulfah & Insani, 2020), and Bangka Belitung (Putra & Vebrian, 2019). Based on the research, there has been no development of mathematical problems that use Maluku context, so researcher are interested in developing mathematical problems using four PISA contents using Maluku context, namely Pela, Gandong, Famili, and Sasi. In addition, the development of mathematical problems is presented using Moodle-assisted websites.
Based on the above description, the researcher is interested in developing mathematics problems that accommodate the content of PISA and the context of Maluku content which includes validity, practicality, and effectiveness (Nuraeni et al., 2021;Rosli et al., 2020;van den Akker et al., 2012). So that, students are used to solving problems that require numeracy ability. The ultimate goal of this media development is to improve the students' numeracy in Maluku.

Methods
This study used students at SMPN 14 Ambon and Muhammadiyah Ambon Junior High School in the development of Mathematics problems with details of 40 students (15 males and 25 females). The instruments used in this study are test question sheets, UEQ questionnaires, SUSquestionnaires, validation sheets, and test questions based on developed mathematics problems. The test question sheet consists of ten minimum competency assessment questions in 2020, including number material, algebra, geometry, and statistics. The UEQ questionnaire was adopted from Santoso et al. (2016), which contains 26 items with seven preferred scales that represented 1) attractiveness (fun, kind, exhilarating, comfortable, attractive, user friendly), 2) clarity (understandable, easy to learn, simple, clear), 3) efficient (fast, efficient, practical, organized), 4) accuracy (predictable, supportive, safe, as expected) 5) stimulus (useful, interesting, not boring, motivating), 6) novelty (creative, inventive, unorthodox, innovative). The SUS questionnaire consists of ten statements with five answer scales. SUS questionnaires are used to assess media systems developed related to the overall use of media that consist of reuse, complex, easy, hard to use, stable, inconsistency, easy to learn, confused, no problem, need time. The mathematics problem validation sheet consists of three instruments consisting of content (suitability of the problem with the curriculum, completeness of representation, suitability of PISA problems and content, proportionality of the problem, and level of difficulty of the problem), language (choice of words, easy-to-understand sentences, ambiguity, sentence structure, default words, proper writing, scientific terms, effective sentences, and informative), and media (layout, background, color, typeface, navigator, consistency, archival, completeness of information, image suitability). The validity of content is related to the conformity of content with the material taught in schools, language validity is related to the readability of the bank of the problem, and the validity of the media is related to the appearance of the problem. The final test question used questions that have been developed in Mathematics problems. It consists of 16 questions, consisting of four PISA contents that consist of change and relationship, shape and space, quantity, uncertainty, and data; and four contexts of local wisdom of Maluku: Pela, Gandong, Famili, and Sasi.
The test question sheet is used as an initial description of the numeracy ability of students used at the preliminary research stage. At the development or prototyping stage, the researcher creates and develops problem items based on PISA content: change and relationship, shape and space, quantity, and uncertainty and data in the context of Maluku local wisdom related to Pela, Gandong, Famili, and Sasi. Pela is a form of a fraternal bond of friendship that connects the indigenous peoples of two different countries (Wakano, 2019). In addition, Gandong is a fraternal relationship or friendship on the basis of a genological relationship (Malatuny & Ritiauw, 2018). Famili is the existence of fraternal relations based on the clan (Wakano, 2019). Sasi is a form of management and utilization of all existing sources, especially natural resources traditionally carried out for generations in Maluku (Lonthor, 2018). After the development is tested, the problem point is based on the analysis of the Rasch Model, which measures the suitability of the problem item with the model, the different power, and the difficulty level of the problem item. The criteria of the Rasch Model that analyzed used Winsteps 5.7.1.0 version Boone et al. (2013)  The measure used to determine the problem's difficulty level is to look at the item: measure using the Winsteps app. Measure values have a range between -3 to 3. However, some experts also limit the lower and upper limits of the possible values to between the range of -2 to 2, if they are outside that range can be expressed as extreme or deviant values (Alagumalai et al., 2005;Bond & Fox, 2013;Dutt et al., 2019). Sumintono and Widhiarso (2015) guide assessing the item into four categories: the following Table 2. Table 2. Difficulty level of problem Interval Category me < -1 Very easy Very hard At the assessment stage, construct validation is carried out to measure the validity of the bank based on expert assessment of content/ material, language, and media. The criteria for its validity test decision based on Widoyoko (2009) are presented in the following Table 3.
Highly invalid full revision The practicality and feasibility of the bank is measured using the UEQ (Schrepp et al., 2017), SUS (Martins et al., 2015). Here, in Table 4 is the EUQ decision criteria, and in Table 5 is the SUS decision criteria. To measure the effectiveness of mathematics problems, a comparison of initial test results (pretest) that uses Minimum Competency Assessment problems in 2020 and final test results (posttest) using the problem in mathematics problems. It is then analyzed using descriptive statistics with N-Gain parameters and inferential statistics with the Wilcoxon signed-rank test. The criteria for N-Gain (Nissen et al., 2018) are as follows in Table 6.

Results
Mathematics problems are the result of the development that has been done. It before first use through the manufacturing and development process. It that has been developed can be utilized by students to improve numeracy ability. The study was developed based on three stages based on the following Plomp development model.

Preliminary research
At this stage, problem analysis is carried out to identify problems experienced by students related to numeracy ability. For this reason, the initial test is carried out using the Minimum Competency Assessment problem that has been used in 2020. Based on the results of tests given to students, information is obtained that students have low student numeracy ability. It is marked by an average student outcome of 38. For this reason, product development is needed in the website form that consists of mathematics problems that can improve students' numeracy ability.

Development or prototyping
At this stage, the creation and development of problem items and bank problems is carried out. The development of the problem item is based on the PISA content and context of local wisdom Maluku. The distribution of the mathematics problems made are as follows.

Assessment
After the problem item is made, the process of analysis of the item is carried out using the Rasch model. The results of the analysis are as follows.

Suitability model (Fit)
Based on the results of the model conformity analysis using the Winstep application, information was obtained from 35 points of the problem developed. There were 33 qualified questions received related to the suitability of the problem using two test indicators, namely MNSQ and ZSTD. The MNSQ test indicator is used to determine if the problem to be used can measure a student's numeracy ability. In contrast, the ZSTD test indicator is used to measure whether the grades obtained by students are likely rational. The MNSQ value in the item analysis conducted is in the range of 0.85 to 1.25. The ZSTD value in the item analysis conducted is in the range of -2.4 to 2.2. The ZSTD value on two problem items (number 17 and number 32) is not accepted because it has a value outside the specified range. Suitability model used to justify validity item based on Rasch model.

Different power
To determine the consistency of the problem in determining students who can answer correctly and students who cannot answer correctly. Pt. Measure Correlation is used as an indicator of different power tests. Based on the data analysis obtained information, three points of questions do not meet. It is because of the value of Pt. Measure Correlation is negative. The problem that does not meet is the numbers 5, 17, and 32. Problem items that do not meet the different power criteria are omitted and not used.

Difficulty level
The problem of difficulty level, measure orders are used. Based on the analysis of the difficulty level of the item obtained information that the problem item is very difficult is the problem item number 21, 18, 26, and 30 because the value of me > 1. In contrast, problem item number 20,11,15,19,31,17,23,24, and 25 belongs to the category of difficult problems because they are at the interval of 0 < me ≤ 1. Easy problems are found in the item of the number 4, 10, 22, 5, 9, 14, 28, 35, 1, 29, 3, 6, 8, 12, 13, 2, 7, 16, 27, 32, and 33 because they are at an interval of -1 < me ≤ 0. While the number of very easy problems is only found in problem number 34 because it is at the interval me < -1.

Reliability
The reliability test is carried out to find out the problem item can show consistency in describing the numeracy ability of students. Based on the reliability test results, the obtained information is that the value of the reliability of the item is 0.74, and the reality of the person is 0.69 with the value of Alpha Cronbach of 0.76. Thus, the problem item has fulfilled the element of inconsistency (Schrepp, 2020). Based on the results of the analysis of the problem details, 32 questions are used to be further entered into the website, and three others are omitted. After all the problem items are entered into the website, an expert validation test is conducted based on content, language, and media. Expert validation results are presented in the following Table 8.

Content validity
Here in Figure 1 and Figure 2 are the results of validation and revision based on expert suggestion: The focus for revisions made by content experts is proportionality between the issues in each content and the context so that the content and context have the same contribution.

Language validity
Here are the results of validation and revision based on the linguists. The focus of the linguist's input is the effective use of sentences. The last sentence on the problem indicates an ineffective sentence because it has the potential to give rise to a double meaning (ambiguous). Should be If the ratio of the largest roof area and the roof area in the second part is 3:1, then how much roof is needed to cover the roof of the first and second parts, if the roof area used each sheet is 2 m 2 ...

Media validity
Here in Figure 3 and Figure 4 are the results of validation and revision based on expert media suggestion:  The focus of revision of media experts is the front page of the website. The front page view should be made more interesting and show information related to the naming of the website.
To find out the practicality and feasibility of Mathematics problems, students who have used Mathematics problems are given the opportunity to fill out the usage questionnaire based on UEQ and SUS. Here in Table 10 is the results of UEQ data analysis. Based on table 10 above, all components meet the criteria of practicality. So that students can use Mathematics problems. After knowing the practicalities, students can fill out the SUS questionnaire to determine the feasibility of Mathematics problems. Based on the results of the SUS questionnaire analysis, the average value of 75 or is on the criteria is proper. Here in Figure  5 is presented a picture of proper acceptance.

Figure 5. Acceptance diagram based on proper
To measure the effectiveness of mathematics problems, tests were conducted on students using the problems contained in Mathematics problems. Based on the test results obtained, the average student test result of 60.5 and N-Gain of 0.39, or if included in the category, the value can be declared to be in the moderate category. Thus, there is a difference in the numeracy of students who use minimum competency assessment problems and those who use mathematics problems. So it can be stated that Mathematics problems can improve students' numeracy. To find out the difference is statistically significant, the t-test with nonparametric criteria is obtained to obtain the following results in Table 11. Based on the Wilcoxon t-test obtained information that the value of significantly p is smaller than α = 0.05. So, it can be stated that there is a significant difference between the numeracy ability of students before and after using mathematics problems.

Discussion Preliminary Research
The development stage begins with identifying problems, one way to provide diagnostic tests to students. It is in line with the Huda et al. (2020); Kusumaningsih et al.(2020), which states that before the development of media, first conducted initial tests so that the student's abilities could be known. The study results showed that the numeracy ability of students is still low; this is in line with Megawati and Sutarto (2021); Saefurohman et al. (2021). Thus, The media is needed that can facilitate students in improving their numeracy ability.

Development
The development of mathematical problems is based on PISA content and the context of local wisdom of Maluku, based on the results of research showing that there are 35 problems developed proportionally. Problem development refers to the context of local wisdom that can increase students' attractiveness in solving problems because they can easily understand the context of a given problem. (Hasanah, 2017;Hasibuan & Fauzi, 2019).

Assessment
The development of the mathematics problem has fulfilled the Rasch model consisting of reliability, conformity of the model, difficulty level, and other power. It must meet the criteria that have been determined so that the problem item used is valid. Some studies show that the problem item that has been tried and validated first is better than the problem item that is not validated first (Alavi et al., 2020;Erfan et al., 2020;Lu et al., 2021). Invalid items are negligible and unfit for use (Petersen et al., 2020;Robitzsch, 2021). In addition to item validation, the mathematics problems website has gone through an expert validation process for the website, including content, language, and media. Validation results show that the content is following the issue of PISA standards, has been following the rules of good Indonesian, and the media has been following the purposes of media creation. Thus, the item on the website has been validated to meet the element of validity and is worth to use. The item that must be developed has fulfilled the validity of content with proportionally diverse issues. It is in line with Barrada et al. (2008), Koneputugodage et al. (2019), Widana (2014), who stated that a good item accommodates the diversity of types of questions presented and it can represent all materials proportionally. In addition, the developed item has met the validity of language characterized by the use of effective sentences in presenting problems. It is in line with Fadhilah (2021), Van Anh (2021), Winarti et al. (2021), which states that sentences on the problem must be effective and easily understood by students so as not to cause ambiguity. The item banks must also show an attractive display to increase students' attractiveness in using the item banks. It is in line with Mutmainah (2017), Nurpala and Munawaroh (2020) which states that the item must be presented attractively so that students are enthusiastic in working on the problems presented.
The website of mathematics problems on the results of development shows practicality based on indicators of attractiveness, clarity, efficiency, accuracy, stimulation, and novelty in line with the opinion of Schrepp et al. (2017), which states that the development media is worth using when it has met the six indicators with a value above 1.20. In addition, the item banks are also worth using because the value of SUS is above 0.75 (Lewis & Sauro, 2018), so the mathematics problems are proper to use. Media development results are said to be effective if there are differences before and after using media (Chanda et al., 2019;Widodo, 2018). It supports the mathematics problems based on website development research on this issue because it improves students' ability to solve numeracy problems. It is characterized by increasing students' N-Gain grades before and after being given a problem using PISA content and local wisdom context with average category. It indicates that there is an increase in the numeracy of students. The average category simply reveals the difference before and after using the website.
The novelty in the research is in the products produced in the form of a collection of problems that contain problems related to the PISA content of the context of local wisdom of Maluku. The item's content and context are presented systematically, making it easier for students to choose and work on the type of problem they want. In addition, Mathematics problems can integrate student knowledge in the context of Maluku and numeracy issues so that students can solve numeracy problems through mathematics problems which have implications for improving students' numeracy ability in solving contextual problems.

Conclusion
Mathematics problems meet the elements of validity, practicality, and effectiveness. The validity consists of questions and item banks. The problem item's validity and reliability are based on the model's adjustable test with 32 valid question points with an Alpha Crombach value of 0.76 or can be said to be reliable. In addition, the validity of mathematics problems website based on experts shows valid criteria in content, language, and media. Mathematics problems that use the website effectively improve the numeracy ability of Maluku students because the N-Gain of students is 0.39 or is in the average category. In addition, the increase was accompanied by significant differences in students' numeracy abilities before and after using mathematics problems based on the website. The research is limited to the PISA content and socio-cultural context domains.

Conflicts of Interest
The author declares that no conflict of interest regarding the publication of this manuscript. In addition, the ethical issues, including plagiarism, misconduct, data fabrication and/or falsification, double publication and/or submission, and redundancies have been completely by the author.