The Orang Asli (Indigenous people) are part of the entity that makes up the population of Malaysia and is considered a minority race. Sources from the Orang Asli Development Department record that the total number of Malaysian Indigenous people is 206,777 (JAKOA, 2022) . It is estimated that 35,975 Indigenous Muslims in 2018, equivalent to 17.4% of the entire population of Indigenous people in Malaysia (Public Sector Open Data, 2022) . Based on these figures, it is clear that a continuous and planned initiative needs to be carried out in the context of preaching to this community. The process and approach of Islamic preaching need to be seen as something important that can improve the quality of life of the Asli people (Razaleigh, 2015) .

Orang Asli are early inhabitants often associated with indigenous people found in Peninsular Malaysia. The word “Orang Asli” is a term that refers to the indigenous people, who are now a minority race in Peninsular Malaysia. The Concise Oxford English Dictionary Eleventh Edition (2011) defines the vocabulary of “aboriginal,” “aboriginal,” and “indigenous” as follows: a) “aboriginal. adj. inhabiting or existing in a land from the earliest times or from before the arrival of colonists; b) indigenous.”, c) “aboriginal: an aboriginal person, animal or plant.”, d) “indigenous. adj. originating or occurring naturally in a particular place; native.” (Soanes & Stevenson, 2011) .

Based on Section 3 of the Aboriginal Peoples Act 1954 of Malaysia (Act 134), an orangasli is defined as a person whose father is a member of the orang asli ethnic group, who speaks an orangasli language and usually follows the orang asli way of life and orang asli beliefs. Includes a male descendant of any person of any race who was adopted as a child by a native and who has been brought up as a native. He usually speaks the natives’ language, following their way of life and beliefs. Also included in the orang asli category is a child from any union between an orangasli woman and a man from an orangasli tribe, provided that the child usually speaks the orang asli language and believes orang asli and is still a member of an orangasli community (JAKOA, 2022) .

Based on the Orang Asli (OA) population census issued by JAKOA, there are 78,636 Orang Asli household members throughout Pahang, Malaysia. Through the census, it was also identified that there are 262 Orang Asli Villages (KOA) around Pahang; Lipis District is the district that has the most number of KOAs, with 69 villages and has the most significant number of households, which is 14,310. On the other hand, Kuantan District has several KOAs; the fewest households are six KOAs, with only 465 total households (JAKOA, 2019) .

However, the religious data of the Indigenous community, updated in 2018, shows that only 7483 of the Indigenous community are Muslims (JAKOA, 2019) . This number represents only 9.5% of the indigenous people in Pahang.

Meanwhile, FardhuAin is an obligation that must be fulfilled by every mukallaf (Islamically accountable). The obligation is not exempted even if it has been done by others, such as performing obligatory prayers five times daily and at night. It must be known by every Muslim to free him from sin, such as knowing the things that must be known in the science of monotheism, jurisprudence, and Sufism (Joll, 2023) .

The life of the da’wah target is influenced by many factors, which hinder and delay the process of da’wah to the native people. These factors include inherited beliefs and customs, their way of life or living environment, their attitude towards preachers, and misconceptions or acceptance of the teachings of Islam (Syed et al., 2010) . Rahman & Mustapha (2020) identify twelve themes of challenges in implementing and succeeding in preaching efforts in Pahang: challenges related to the location of the indigenous community, not mastering the indigenous language, threats from the indigenous people, lack of preachers, characteristics and attitudes not suitable for residents, rival Christian missionaries, financial resources, internal problems of preachers, the attitude of the natives, attitude of the Malay race, not keeping promises, and no overlap between NGOs.

The laziness of the Orang Asli community in learning Islam is a significant reason why they fail to understand Islam, leading to their inability to meet the demands of Islamic teachings. Despite embracing Islam, many are still unable to perform required practices or abandon prohibited ones due to illiteracy. Consequently, some do not know how to pronounce the shahadah, pray, or uphold Islamic principles (Razaleigh, 2015) . Additionally, a lack of motivation within the indigenous community to learn Islam prevents them from fully understanding and fulfilling Islamic teachings, resulting in them being Muslims in name only, without practice (Razaleigh et al., 2012; Razaleigh, 2015) . This issue is compounded by their preference for activities that can increase their income, as many come from low-income backgrounds.

New converts, including the Orang Asli, require protection and guidance from authorities through appropriate education modules. Education on the fundamentals of fardhuain is crucial to ensure they clearly understand Islam. Nurayuni et al. (2019) and Siti Fathimatul and A’thirohMasyaa’il (2015) emphasize the importance of specific education programs for new converts as a platform for social education support. These programs help increase their understanding of Islam and enable them to practice the Islamic way of life. The curriculum for these study programs must be carefully planned to achieve optimal learning outcomes (Siti Afifah et al., 2022) .

Nazatul Akmar (2023) identifies the challenges teachers and students face in mastering the Al-Quran and Fardhu Ain among the native people in Pahang. Teachers face challenges in teaching methods, dealing with specific groups, and managing student attitudes while students encounter their own problems.

A study by Halim et al. (2023) on the Fardhu Ain module used by the Council of Islamic Religion and Malay Customs of Pahang (MUIP, 2023) reveals that practices like ablution, recitation of prayer intentions and recitation while iktidal are well-mastered by the Asli people. However, recitations of the qunut prayer, iftitah prayer, and tahiyat are less mastered. The study concludes that while the Orang Asli community has a basic understanding of Fardhu Ain, continuous efforts are needed to improve teaching methods to enhance learning accessibility and effectiveness.

In conclusion, various external and internal problems exist in the appreciation of Islam among the Indigenous people in Pahang that must be addressed. A comprehensive teaching module and an instrument for measuring the Indigenous community’s mastery of Fardhu Ain programs are required. Therefore, a mastery module on Fardhu Ain for the Indigenous community in Pahang needs to be developed and regularly updated.

This study was conducted to design and develop the criteria and content of the ‘Fardhu Ain Mastery Module of the Indigenous People in Pahang State’ based on the expert verification process using Fuzzy Delphi analysis (Chang, Huang, & Lin, 2000; Mustapha & Darussalam, 2018) . This is Phase 2 of the previous research process, which, as a whole, used the ADDIE instructional model (Dick et al., 2014) and the DDR (design and development research) method (Richey & Klein, 2007; Saedah et al., 2013) .

This study is a quantitative study that uses the Fuzzy Delphi Method (FDM) application. The use of FDM is to obtain expert consensus on the items from the elements of the teaching process that make up the Differentiated Pedagogy Model. The researcher chose the FDM method because it allows the researcher to process the ambiguity with predictive and non-predictive items, and the attributes of the participants can be explained (Chang et al., 2000) . Researchers also chose FDM as a method of analyzing data because FDM can overcome the weaknesses of the Traditional Delphi Method, among which 1) the traditional Delphi Method has the potential to get answers at a low rate and 2) the Traditional Delphi also involves a long time (Mustapha & Darussalam, 2018) .

a) Study Implementation Steps

Step 1: Selection of experts. Step 2 in the data collection process shows the researcher selecting experts to answer the constructed questionnaire. The researcher selected nine experts according to the researcher’s field of study, which is the field of al-Quran recitation and pedagogy, as well as experts in the fuzzy Delphi method (FDM). The researcher chose nine experts because, according to Twiss and Harry (1978) . The expert information collected by the researcher in this study is related to the position and place of duty, experience, and area of expertise. Expert experience is more than five years of experience. This coincides with Akbari and Yazdanmehr (2014) , who stated that experts have more than five years of experience in a specific field.

Step 2: Distribution of questionnaires to experts After the researcher identified relevant experts in the field, the researcher contacted the experts via email. After receiving the experts’ feedback, the researcher emailed the questionnaire. Before that, the researcher informed the researcher of the information related to the study to facilitate the experts’ answering of the questionnaire.

Step 3: Analyzing the data after the expert sent the questionnaire, the researcher analyzed the questionnaire using Fudelo software. FDM data analysis must meet the three main conditions of FDM, namely 1) the threshold value, (d) does not exceed or equal to 0.2 (Chen, 2000) , 2) the percentage of expert agreement must be greater than or equal to 75% (Murry & Hammons, 1995) and 3) the defuzzification value (alpha cut) should be greater than or equal to 0.5 (Bodjanova, 2006) .

Step 4: Determination of items by experts Through the Fuzzy Delphi Method, the results of data analysis are through expert consensus. Expert consensus will determine the ranking of the most essential items. The results of the expert consensus will be discussed in the study findings.

b) Delphi Fuzzy Analysis Process

1) Selection of experts: This study used nine experts. Several experts were invited to determine the importance of the evaluation criteria for the variables that will be measured using linguistic variables. The researcher met face-to-face with the selected and identified experts to facilitate the discussion and explanation of issues in items, etc.

2) Step 2: Determination of linguistic variables (determining linguistic scale). This process involves converting all linguistic variables into triangular fuzzy numbering (triangular fuzzy numbers). This step also involves converting linguistic variables by adding fuzzy numbers (Hsieh et al., 2004) . Triangular Fuzzy Number represents the values of m₁, m₂, and m₃, and it is written like this (m₁, m₂, m₃). The m₁ value represents the minimum value, the m₂ value represents the reasonable value, and the m₃ value represents the maximum value. Triangular Fuzzy Number produces a Fuzzy scale to translate linguistic variables into fuzzy numbers. The number of levels for the fuzzy scale is odd.

This study collected and analyzed data using the Fuzzy Delphi technique after experts were given a questionnaire, a Likert scale, and a space for expert comments and suggestions for each instrument. The Likert scale data obtained was analyzed using the Microsoft Excel program. All data is converted into Triangular Fuzzy Number form. A fuzzy scale (7) of seven points is used in this study.

Step 3: After the researcher obtains a response from the selected expert, the researcher needs to convert all Likert scales to Fuzzy scales. This process is also known as identifying the average response of each fuzzy number (Benítez et al., 2007) .

It takes place based on the formula $M=\frac{{\displaystyle {\sum }_{i=1}^n{m}_i}}{n}$ as illustrated in Figure 1 .

Step 4: Identifying the value of Threshold “d”. The threshold value is significant in identifying the level of agreement between experts (Thomaidis et al., 2006) . The distance for each fuzzy number m = (m₁, m₂, m₃) and n = (m₁, m₂, m₃) is calculated using the formula explained in Figure 2 .

The threshold value is significant in determining agreement between experts. According to Cheng & Lin (2002) , expert agreement has been reached if the threshold value is less than or equal to 0.2. The overall agreement (group consensus) must exceed 75% agreement for each item; otherwise, the second round must be implemented.

Step 5: Identifying the aggregate alpha level of the fuzzy evaluation. After the expert agreement is obtained, fuzzy numbers for each item are added (Mohd Ridhuan et al., 2013) , explained in Figure 3 . The fuzzy value is calculated using A_max = 1⁄4(m₁ + 2m₂ + m₃).

Step 6: The next step is the diffusion process phase. This process uses the formula A_max =1⁄4(a₁ + 2a_m + a₃). If the researcher uses Average Fuzzy Numbers or average response, the score is 0 to 1 (Mohd Ridhuan et al., 2013) . In this process, there are three formulas, namely: 1) A = 1/3*(m₁ + m₂ + m₃), or; 2) A = 1/4*(m₁ + 2m₂ + m₃), or; 3) A = 1/6*(m₁ + 4m₂ + m₃). α-cut value = median

value of “0” and “1”, where α-cut = (0 + 1)/2 = 0.5. If the resulting A value is less than the α-cut value = 0.5, the item will be rejected because it does not show expert agreement. According to Bodjanova (2006) , the alpha cut value should exceed 0.5. It is supported by Tang & Wu (2010) , who stated that the α-cut value should exceed 0.5.

Step 7: Ranking process. The ranking process involves selecting elements based on defuzzification values based on expert agreement, in which the element with the highest value is the most important (Hierro et al., 2021; Fortemps & Roubens, 1996) . The ranking process follows the Ai formula (Cheng et al., 2011) , as shown in Figure 4 .

c) Research Respondents

This study uses purposive sampling. This method is the most appropriate since the researcher wants to get a consensus view and consensus on a matter. According to Hasson, Keeney, and McKenna (2000) , purposive sampling is the most appropriate method in FDM. Meanwhile, a total of nine experts were involved in this study. These experts are selected based on their experience and expertise in their respective fields. In this study, the number of experts used by the researcher is based on the recommendations of Clayton (1997) , who states that if the experts involved are homogeneous, then the number of experts required is 5 - 10. Adler & Ziglio (1996) , the appropriate number of experts in the Delphi method is between 10 to 15 people if there is uniformity (homogenous). Cavalli et al. (1984) stated that the sample for FDM is between 8 and 12 if the sample is homogeneous and sufficient, and the opinion of Philip (2000) stated that the expert sample is between 7 and 12. Therefore, the researcher used a total of nine experts.

d) Module Components Analyzed

There are four components of the module teaching curriculum based on Fred (2011) , which are analyzed in this expert consensus study which are:

a) Objectives of the module

b) The content of the topics in the module

c) Module teaching activities

d) Module assessment activities.

The findings of this study are as explained and shown below:

a) Module Objectives

The findings of the Fuzzy Delphi analysis of the module objectives are shown in Table 1 below:

Threshold value: **“d” value ≤ 0.2, Average of consensus ≥ 75%, Defuzzification ≥ 0.5 (Alpha cut).

Based on Table 1 , the threshold value blacked out exceeds the threshold value of 0.2 (>0.2). This means there are uneven expert opinions and no consensus on certain items. However, the d value of the overall construct shows 0.067 (<0.2). According to Cheng and Lin (2002) and Cheng, Hsu, and Chang (2011) , if the average threshold value (d) obtained is less than <0.2, then the item has reached expert agreement. Meanwhile, the overall percentage of expert agreement is at a value of 97% agreement, which is more than (75%) meaning that the expert agreement on the item is met. According to Cheng, Hsu, and Chang (2011) , the agreement percentage should exceed 75%. In addition, all the Alpha-Cut defuzzication values (average of fuzzy response) exceed 0.5. According to Tang and Wu (2010) and Bodjanova (2006) , the alpha cut value should exceed >0.5. This shows that the experts have agreed upon the module’s objective items. Overall, all items have been agreed by experts with a good agreement value and meet the specified conditions.

b) Module Content

The findings of the Fuzzy Delphi analysis of the content of the topics in the module are shown in Tables 2-5 below:

Threshold value: **“d” value ≤ 0.2, Average of consensus ≥ 75%, Defuzzification ≥ 0.5 (Alpha cut).

Based on Table 2 , no threshold value is blacked out above the threshold value of 0.2 (>0.2). This means there is an even expert opinion and consensus on certain items. However, the d value of the overall construct shows 0.050 (<0.2). According to Cheng and Lin (2002) and Cheng, Hsu, and Chang (2011) , if the average threshold value (d) obtained is less than <0.2, then the item has reached expert agreement. Meanwhile, the overall percentage of expert agreement is at a value of 100% agreement, which is more than (75%) meaning that the expert agreement on the item is met. According to Cheng, Hsu, and Chang (2011) , the agreement percentage should exceed 75%. In addition, all the Alpha-Cut defuzzication values (average of fuzzy response) exceed 0.5. According to Tang and Wu (2010) and Bodjanova (2006) , the alpha cut value should exceed >0.5. This shows that the experts have agreed upon the items in the construction of the prophets and their apostles and their people. Experts have agreed upon all items with a good agreement value and meet the specified conditions.

Threshold value: **“d” value ≤ 0.2, Average of consensus ≥ 75%, Defuzzification ≥ 0.5 (Alpha cut).

Based on Table 3 , one threshold value is blacked out above the threshold value of 0.2 (>0.2). This means there are uneven expert opinions and no consensus on certain items. However, the d value of the overall construct shows 0.05987 (<0.2). According to Cheng and Lin (2002) and Cheng, Hsu, and Chang (2011) , if the average threshold value (d) obtained is less than <0.2, then the item has reached expert agreement. Meanwhile, the overall percentage of expert agreement is 98% agreement, which is more than (75%) meaning that the expert agreement on the item is met. According to Cheng, Hsu, and Chang (2011) , the agreement percentage should exceed 75%. In addition, all the Alpha-Cut defuzzication values (average of fuzzy response) exceed 0.5. According to Tang and Wu (2010) and Bodjanova (2006) , the alpha cut value should exceed > 0.5. This shows that the experts have agreed upon exemplary moral constructs in life. Experts have agreed upon all items with a good agreement value and meet the specified conditions.

Table 4. Tafsir and tadabbur. (Adapted table from Cheng, Hsu, and Chang, 2011 )

Threshold value: **“d” value ≤ 0.2, Average of consensus ≥ 75%, Defuzzification ≥ 0.5 (Alpha cut).

Based on Table 4 , no threshold value is blacked out above the threshold value of 0.2 (>0.2). This means there is an even expert opinion and consensus on certain items. However, the d value of the overall construct shows 0.051 (<0.2). According to Cheng and Lin (2002) and Cheng, Hsu, and Chang (2011) , if the average threshold value (d) obtained is less than <0.2, then the item has reached expert agreement. Meanwhile, the overall percentage of expert agreement is at a value of 100% agreement, which is more than (75%) meaning that the expert agreement on the item is met. According to Cheng, Hsu, and Chang (2011), the agreement percentage should exceed 75%. In addition, all the Alpha-Cut defuzzication values (average of fuzzy response) exceed 0.5. According to Tang and Wu (2010) and Bodjanova (2006) , the alpha cut value should exceed > 0.5. This shows that the experts have agreed upon the items in constructing tafsir and tadabbur. Overall, all items have been agreed by experts with a good agreement value and meet the specified conditions.

Table 5. Selected hadith. (Adapted table from Cheng, Hsu, and Chang, 2011 )

Threshold value: **“d” value ≤ 0.2, Average of consensus ≥ 75%, Defuzzification ≥ 0.5 (Alpha cut).

Based on Table 5 , no threshold value is blacked out above the threshold value of 0.2 (>0.2). This means there is an even expert opinion and consensus on certain items. However, the d value of the overall construct shows 0.0513 (<0.2). According to Cheng and Lin (2002) and Cheng, Hsu, and Chang (2011) , if the average threshold value (d) obtained is less than <0.2, then the item has reached expert agreement. Meanwhile, the overall percentage of expert agreement is at a value of 100% agreement, which is more than (75%) meaning that the expert agreement on the item is met. According to Cheng, Hsu, and Chang (2011) , the agreement percentage should exceed 100%. In addition, all the Alpha-Cut defuzzication values (average of fuzzy response) exceed 0.5. According to Tang and Wu (2010) and Bodjanova (2006) , the alpha cut value should exceed > 0.5. This shows that the experts have agreed upon the items in constructing Tafsir and Tadabbur. Experts have agreed upon all items with a good agreement value and meet the specified conditions.

c) Module Teaching Activities

The findings of the Fuzzy Delphi analysis of the module teaching activities are shown in Table 6 below:

Based on Table 6 , no threshold value is blacked out above the threshold value of 0.2 (>0.2). This means there is an even expert opinion and consensus on certain items. However, the d value of the overall construct shows 0.0482 (<0.2). According to Cheng and Lin (2002) and Cheng, Hsu, and Chang (2011) , if the average threshold value (d) obtained is less than <0.2, then the item has reached expert agreement. Meanwhile, the overall percentage of expert agreement is at a value of 100% agreement, which is more than (75%) meaning that the expert agreement on the item is met. According to Cheng, Hsu, and Chang (2011) , the agreement percentage should exceed 100%. In addition, all the Alpha-Cut defuzzication values (average of fuzzy response) exceed 0.5. According to Tang and Wu (2010) and Bodjanova (2006) , the alpha cut value should exceed >0.5. This shows that the experts agreed upon the items in the module’s Teaching Activity construct. Experts have agreed upon all items with a good agreement value and meet the specified conditions.

e) Module Evaluation Activities

The findings of the Fuzzy Delphi analysis of the module evaluation activities are shown in Table 7 below:

Threshold value: **“d” value ≤ 0.2, Average of consensus ≥ 75%, Defuzzification ≥ 0.5 (Alpha cut).

Based on Table 7 , no threshold value is blacked out above the threshold value of 0.2 (>0.2). This means there is an even expert opinion and consensus on certain items. However, the d value of the overall construct shows 0.0482 (<0.2). According to Cheng and Lin (2002) and Cheng, Hsu, and Chang (2011) , if the average threshold value (d) obtained is less than <0.2, then the item has reached expert agreement. Meanwhile, the overall percentage of expert agreement is at a value of 100% agreement, which is more than (75%) meaning that the expert agreement on the item is met. According to Cheng, Hsu, and Chang (2011) , the agreement percentage should exceed 100%. In addition, all the Alpha-Cut defuzzication values (average of fuzzy response) exceed 0.5. According to Tang and Wu (2010) and Bodjanova (2006) , the alpha cut value should exceed >0.5. This shows that the experts agreed upon the Module Evaluation Activity construct items. Experts have agreed upon all items with a good agreement value and meet the specified conditions.

The consensus among experts underscores the critical need for these updates to ensure the module’s relevance and effectiveness in delivering essential religious education to the Pahang Indigenous community. By addressing the unique educational needs of the Orang Asli, this updated module aims to provide a more robust and comprehensive religious education framework.

The implications of this study are profound. It not only identifies specific areas for curriculum enhancement but also provides a clear pathway for the implementation of these changes. The findings will inform the upcoming implementation stage of the Fardhu Ain strengthening program, where the updated module will be applied and its effectiveness rigorously evaluated. This study sets a precedent for future curriculum development initiatives, emphasizing the importance of continuous improvement and expert validation in educational program design.

This quantitative study has substantially enhanced the Fardhu Ain module for the Pahang Indigenous community by updating its content and structure through a meticulous expert consensus process using Fuzzy Delphi analysis. The experts identified four essential components to be included in the module: a) Objectives of the Module, b) Content of the topics, c) Teaching activities, and d) Module assessment activities. These updates have led to adding four new topics, expanding the module from seven to eleven components. The new topics—basic reading skills of the Quran and Tajwid, interpretation of selected Quranic verses, selected hadiths, and stories of selected Prophets—address critical areas necessary for a comprehensive religious education. The consensus reached among experts highlights the importance and relevance of these updates in enhancing the educational framework tailored for the Orang Asli community in Pahang. By incorporating these new topics, the module offers a more robust and inclusive curriculum better suited to meet this community’s educational needs.

The findings of this study provide a clear roadmap for implementing the updated module in the upcoming Fardhu Ain strengthening program. This implementation phase will allow for a thorough evaluation of the module’s effectiveness in delivering enhanced religious education. The study contributes to immediate curriculum improvements and sets a foundation for ongoing curriculum development efforts, emphasizing the value of expert validation and continuous enhancement in educational program design. The anticipated impact of this study extends beyond immediate educational outcomes, aiming to foster long-term educational and spiritual growth within the Pahang Indigenous community.

This study expresses its highest appreciation to the Pahang Council of Islamic Religion and Malay Customs (MUIP) for funding a grant amounting to RM 30,000 for the year ending 30 June 2024 and to the Research Management Center (RMC), IIUM, for cooperating in managing this grant with MUIP for one year from 2023.