LOCAL HALAL COSMETIC PRODUCTS PURCHASE INTENTION: KNOWLEDGE, RELIGIOSITY, ATTITUDE, AND ISLAMIC ADVERTISING FACTORS

The demand for halal cosmetic products in various parts of the world has increased, especially in Muslim-majority countries like Indonesia. The potential demand for halal cosmetic products will continue to increase in the future. Indonesia has the opportunity to become the world’s largest producer and consumer of halal cosmetic products. Therefore, Indonesia must be able to take advantage of these opportunities by maximizing its potential. This study aims to determine the effect of knowledge, religiosity, attitudes, and Islamic advertising factors on the purchase intention of local halal cosmetic products. In addition, it also aims to find out the reasons for the purchase intention of Muslim consumers towards local halal cosmetic products and to determine the purchase intentions of Muslim consumers towards imported cosmetic products in the future. This study used 400 respondents who are Muslim and domiciled in West Java. Then to test the model and hypothesis using Smart PLS Software. The results of this study indicate that all hypotheses are accepted. Knowledge and religiosity have a positive and significant influence on attitudes and purchase intentions. Then attitude and Islamic advertising have a positive and significant influence on the purchase intention of local halal cosmetic products. Thus, in building the intention to buy local halal cosmetic products, the company must know and understand what factors underlie consumers’ intentions to buy local halal cosmetic products.

of local halal cosmetic products. Thus, in building the intention to buy local halal cosmetic products, the company must know and understand what factors underlie consumers' intentions to buy local halal cosmetic products.

INTRODUCTION
Halal products are produced according to Islamic Sharia and are attractive to Muslim consumers because the products follow their needs guided by their religious teachings (Hussain-Gambles, 2020;Suhartanto et al., 2019;Wilson, 2014). The standard interpretation of halal products is that the ingredients do not contain pork, alcohol, and slaughtering animals must be following Islamic teachings (Abd Rahman, Asrarhaghighi, and Ab Rahman 2015). This statement is in accordance with the Law of the Republic of Indonesia Number 33 of 2014 concerning halal product guarantee, which states that halal products are products that have been declared halal in accordance with Islamic Shari'a and halal product processing is a series of activities to ensure the halalness of products including the provision of materials, processing, storage, packaging, distribution, sale and presentation of Products. One of the halal products that are currently trending is cosmetics, halal cosmetic products are considered as innovations and revolutions because halal cosmetic products offer high quality that follows halal compliance (Khan, Sarwar, and Tan 2020) and are considered healthier, safer products and can increase their power. Consumer attraction (Mohezar, Zailani, and Zainuddin 2016). Halal cosmetic products are one of the potential products in the future (Abd Rahman, Asrarhaghighi, and Ab Rahman 2015).
Indonesia is a country with the largest Muslim population in the world. Based on Data Centre released by Ministry of Religious Affairs on 2018, Muslim population in Indonesia reached 231 million people. Indonesia is expected to become one of the five potential markets for cosmetic products in the next ten to fifteen years (Briliana and Mursito 2017;Cheong et al. 2016). The Ministry of Industry of Republic of Indonesia recorded an increase in imported cosmetic products from 2017, amounting to US$ 631.66 million to US$ 850.15 million (Hamdani 2019). There will still be illegal imported cosmetics and cosmetics with dangerous raw materials with a fantastic amount of 10 billion rupiahs, in Indonesia during 2020. However, According to Databoks, Indonesian consumers prefer global cosmetic products to local cosmetic products. This is due to the loss of local cosmetic products, which are thought to be influenced by several factors, including the small scale of local producers' businesses, imported cosmetic products whose image is already known, and the lack of training and assistance from relevant agencies (Nursyamsiah 2018).
The previous research, consumers purchase intentions for halal cosmetic products can be identified through consumer attitudes towards halal cosmetic products (Abd Rahman, Asrarhaghighi, and Ab Rahman 2015;Briliana and Mursito 2017). Consumers who have a positive attitude towards halal cosmetic products will be more likely to buy halal cosmetic products (Briliana and Mursito 2017). Knowledge has a positive and significant relationship to attitudes (Adriani and Ma'ruf 2020;Briliana and Mursito 2017). Furthermore, consumer religiosity also has a positive and significant relationship with attitudes (Abd Rahman, Asrarhaghighi, and Ab Rahman 2015; Adriani and Ma'ruf 2020;Mukhtar and Butt 2012).
The purpose of this study is to determine and explain the influence of knowledge, religiosity, attitude, and Islamic advertising on the purchase intention of local halal cosmetic products. In addition. The results of this study are expected to help the development of the halal cosmetic industry in Indonesia to be able to compete nationally and globally and expected to provide understanding and education to Indonesian Muslim consumers in using local halal cosmetic products.

LITERATURE REVIEW Knowledge, Religiosity, Islamic Advertising, Attitude and Purchase Intention
In general, knowledge is based on facts, feelings, or experiences that specific individuals or groups can know of people (Abd Rahman, Asrarhaghighi, and Ab Rahman 2015). Knowledge of attitudes and purchase intentions had a positive and significant relationship with halal cosmetic products consumers' purchase intentions (Abd Rahman, Asrarhaghighi, and Ab Rahman 2015;Briliana and Mursito 2017) Consumers who know halal products will be more selective in choosing the products to be used (Adriani and Ma'ruf 2020). However, previous research (Abd Rahman, Asrarhaghighi, and Ab Rahman 2015) showed that knowledge and attitudes did not significantly affect halal cosmetic products.
Religiosity indicates the level of individual commitment to religion (Suhartanto et al., 2019;Suhartanto et al., 2020). Based on previous research shows that religiosity has a positive and significant relationship with attitudes towards halal cosmetic products (Abd Rahman, Asrarhaghighi, and Ab Rahman 2015; Adriani and Ma'ruf 2020; Mukhtar and Butt 2012;Rohmatun and Dewi 2017). The religious belief plays the least important role in the purchase intention of halal cosmetic products (Khan, Sarwar, and Tan 2020). The better the level of consumer religiosity, the more positive attitudes towards Halal cosmetic products will be and the higher purchase intention for Halal cosmetic products (Adriani and Ma'ruf 2020) Consumers who have a more positive attitude towards halal cosmetics will decide to buy halal cosmetic products (Briliana and Mursito 2017). Based on previous research, attitudes positively influence interest in buying halal products, especially halal cosmetic products (Abd Rahman, Asrarhaghighi, and Ab Rahman 2015; Adriani and Ma'ruf 2020; Mukhtar and Butt 2012). So, the hypothesis of this research are as follows: The concept of Islamic advertising emerged. Islamic advertising is a new concept for the world (Shafiq et al. 2017). Islamic advertising is a form of communication message that contains Islamic values disseminated to consumers to introduce and offer products, services, and ideas (Saifudin 2019). Consumers accept Islamic advertising as a good advertising category (Safira 2017). Islamic advertising is sharia and "can be used for da'wah methods" (Mokhtar and Samsudin 2015). The presence of Islamic advertising brings positive emotions to Muslim consumers in Indonesia, making them feel more calm and peaceful so that it can influence the attitudes and behavior of Indonesian Muslim consumers (Safira 2017).
Islamic advertising has a significant and significant influence on purchases (Sari 2020). All Islamic advertising variables (Islamic story illustration variables and Muslim advertising stars) have a positive and significant influence on purchase intention, except that the Islamic advertising message variable does not affect purchase intention (Saifudin 2019). Research that uses Islamic advertising as a variable determining purchase intention of a product or service is still limited, especially in Halal cosmetic products. Therefore, this study was conducted to determine the effect of Islamic advertising on halal cosmetic products. The following is the research hypothesis: Hypothesis 1: There is a positive and significant relationship between knowledge and attitudes towards local halal cosmetic products. Hypothesis 2: There is a positive and significant relationship between knowledge and purchase intention of local halal cosmetic products. Hypothesis 3: There is a positive and significant relationship between religiosity and attitudes towards local halal cosmetic products. Hypothesis 4: There is a positive and significant relationship between religiosity and purchase intention of local halal cosmetic products Hypothesis 5: There is a positive and significant relationship between attitudes and buying interest towards local Halal cosmetic products. Hypothesis 6: There is a positive and significant relationship between Islamic advertising and purchase intention of local halal cosmetic products.

Research Model
This research model is adapted from (Abd Rahman, Asrarhaghighi, and Ab Rahman 2015) and has been modified by adding an Islamic advertising variable. Islamic advertising has a significant and significant effect on purchasing decisions (Sari, 2020) and has a positive and significant effect on purchase intention, except that the Islamic advertising message variable has no effect (Saifudin, 2019). In addition, the theoretical basis of this adapted model is based on Theory Reasoned of Action (TRA) (Ajzen, I., & Fishbein 1980). This model is considered valid in predicting the intention to choose halal products (Mukhtar and Butt 2012).
In the Reasoned Theory of Action, individual intentions are a function of two essential determinants, including personal factors (attitudes) and social influences (Lada, Harvey Tanakinjal, and Amin 2009). The detail of the research model can be seen at

Research Design
Quantitative research methods are used in this study. Quantitative methods are used to determine and measure the extent to which consumers agree with some of the statements put forward in this questionnaire. In this study, the author uses a descriptive design to describe something (Malhotra et al. 2004) (Malhotra, 2004). The descriptive design used is a cross-sectional design, namely a research design in which data collection is carried out once in a while in answering questions from questionnaires (Suhartanto, 2014).

Population and Sample
In this study, the sample used is Muslim consumers with a minimum age of 17 years until 28 years old, and domiciled in West Java. The research was conducted on 2021. This research is included in marketing research with a minimum sample size requirement of 200 respondents (Malhotra, Nunan, and Birks 2017). However, the samples size used in this study was 400 samples.
Based on Table 1 shows that 400 respondents responded to this questionnaire. The majority of respondents who responded to this questionnaire were female Muslim respondents with 354 people or the equivalent of 88.5%. Then followed by male Muslim respondents as much as 46 or 11.5%. After the age of the respondents is categorized into several, the following is the order of age categories that make up the majority in filling out this questionnaire. First respondents aged 20-22 years old (63.5%) followed by 17-19 years old (28%), followed by 23-25 years old (7.5%), and the last category was 26-28 years old (1%). Half of the respondents in this study had a high school education background (58.5%) followed by the latest education D1/D2/D3 (25.5%), undergraduate (15.3%), less than junior high school or equivalent (0.5%), and Postgraduate (0.3 %). The majority of respondents in this study, almost half, were students (79%). Then some respondents work as private employees (12.3%) and entrepreneurs (3.5%), and respondents who have other jobs. In addition, the income of the majority of respondents ranged from <IDR 1,000,000 (65.3%), considering that the dominant respondents in filling out this questionnaire were still students.

Data Collection Instruments
The data used in this study is primary data. Primary data is obtained from questionnaires that have been distributed via Google Form to samples that meet the criteria. The method of data collection is done using a self-administered survey. The researcher used a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree) in measuring all items. There were open and closed questions to be used as material for qualitative analysis. The construct of this study adopted variables from previous studies (Abd Rahman et al., 2015;Mokhtar & Samsudin, 2015;Suhartanto et al., 2020). In addition, researchers used secondary data as supporting material for this research, such as books, articles, reports, and other sources.

Measurement Model
The outer model in this study was measured by conducting validity tests and reliability tests. To determine the validity, convergent validity can be used. If the Average Variance Extracted (AVE) value is above 0.5, it can be categorized as valid (Joe F. Hair et al. 2014;Joseph F. Hair et al. 2019). In addition, discriminatory validity testing can use the Fornell-Lacker Criterion (Henseler, Ringle, and Sarstedt 2015). Next, reliability measurement can be done by looking at Cronbach's alpha and composite reliability values. If Cronbach's alpha value is more than 0.6 (Joe F Hair et al. 2018;Haque et al. 2018) and the composite reliability value is more than 0.6, then it can be categorized as reliable (Joe F. Hair et al. 2014;Haque et al. 2018).

Structural Model
After measuring the outer model and the results that meet the criteria, the next step is to test the structural model (inner model). There are several testing steps in this stage, including testing Goodness of Fit (GoF), Coefficient of Determination (R 2 ), Predictive Relevance (Q 2 ), and testing hypotheses. Goodness-of-Fit (GoF) is used to evaluate the suitability of the research model made.
The results of the calculation of Goodness-of-Fit (GoF) can be categorized as follows: 0.1 (small), 0.25 (Medium), and 0.38 (large) (Tenenhaus, Amato, and Vinzi 2004). The R² test was conducted to measure the effect of the independent variable on the dependent variable. R2 values can be categorized as follows 0.75 (substantial), 0.50 (moderate), and 0.25 (weak) (Joseph F. Hair et al. 2019). Q2 testing was conducted to test the relevance of the model in this study. The research model can be categorized as having a high level of relevance if the resulting value is greater than 0 (Hair et al., 2014).
Hypothesis testing in this research is through path coefficient analysis. Path coefficient analysis was carried out to calculate the strength of the relationship using a matrix or covariance as the input (Malhotra, Nunan, and Birks 2017). The test method used in this hypothesis is the bootstrap method. The value of each variable will be entered. The test results can later be used as a benchmark for acceptance or rejection of the proposed hypothesis.

Validity and Reliability
The purpose of testing the validity and reliability is to determine which research indicators can explain and reflect latent variables through the evaluation of the outer model. The outer model is one part of the measurement model used to determine the validity and reliability of each construct in a study (Malhotra, Nunan, and Birks 2017). This is done to determine whether the analysis will be carried out further or not. Table 2 shows that all variables in this study can be categorized as reliable because of the value of Cronbach's alpha 0.60 (Haque et al. 2018). In addition, the composite reliability value is also 0.60 (Joe F. Hair et al. 2014;Haque et al. 2018). In this study, several variables have indicators that do not match the criteria. Thus, the author does not include this indicator in the following analysis because it can affect the Average Variance Explained (AVE) value. The AVE value is acceptable if the value is ≥ 0.50 (Joe F Hair et al. 2018;Joseph F. Hair et al. 2019). There are three indicators on the knowledge variable that were not included in the subsequent analysis. Among them are the understanding of Islamic law about halal and haram food and drinks, knowledge about what Islamic teachings prohibit foods and drinks, and knowledge to distinguish between food ingredients allowed and not allowed by Islam.
Furthermore, there are four indicators on the religiosity variable that are not included in the subsequent analysis, namely the implementation of the five daily prayers, paying zakat, avoiding illicit income, and knowledge of the pillars of faith and Islam. The factor loading value for each indicator is less than 0.708. The condition for the factor loading value is ≥ 0.708 (Joe F Hair et al. 2018;Joseph F. Hair et al. 2019). Thus, the author does not involve indicators on these variables.  Table 3 shows that all variables in this study can be categorized as valid because they have an AVE value of ≥ 0.50 (Joe F. Hair et al. 2014;Joseph F. Hair et al. 2019). Discriminant validity in this study can be determined by the Fornell-Larcker Criteria shown in table 3. It can be seen that all latent variables have greater values than other constructs in the same model. Therefore, the construct in this study meets the requirements of discriminant validity.

Goodness-of-Fit
The next stage is to evaluate by assessing the structural model with calculations that have been done previously manually based on the existing formula. There are three criteria for GoF scores with the following conditions 0.1 (small), 0.25 (Medium), and 0.38 (large) (Tenenhaus, Amato, and Vinzi 2004). Based on Table  4 shows that the GoF value is 0.501 and can be categorized as large. Thus, this research model is accurate for testing hypotheses.

Coefficient of Determination (R 2 ) and Predictive Relevance (Q 2 )
The next step is to measure the effect of the independent variable on the dependent variable. R 2 values can be categorized as follows 0.75 (substantial), 0.50 (moderate), and 0.25 (weak) (Joseph F. Hair et al. 2019). Based on the results of the calculation of R 2 shown in table IV.5, attitudes are influenced by 30% of knowledge and religiosity. In contrast, the other 70% are influenced by other factors that are not included in this study. So, it can be categorized as weak. Furthermore, purchase intention is influenced by 44.7% of Islamic knowledge, religiosity, attitude, and advertising. In addition, 55.3% are influenced by other factors not included in this study. Therefore, based on the calculation results, it can be categorized as weak. In table 4, there is a relevant predictive value (Q 2 ). The value of the variables in this study has a relevance value above 0. This means that this research model has high relevance. The research model is categorized as having a high level of relevance if the resulting value is greater than 0 (Joe F. Hair et al. 2014).

Path Analysis
Hypothesis testing is done through path coefficient analysis. The purpose of path coefficient analysis is to calculate the strength of the relationship using a matrix or covariance as the input (Malhotra, Nunan, and Birks 2017). The bootstrap method is used in testing this hypothesis. Each value of each variable will be entered. The results of this analysis test can be used as a benchmark for the acceptance or rejection of the hypothesis that has been made.  Table 5 displays the results of the empirical evaluation of the hypothesis. Based on the path coefficient calculation results, it shows that knowledge and attitudes have a coefficient value of 0.22 with t-value of 4.45. This means that it can be assumed that knowledge has a positive and significant influence on attitudes. Thus, H1 in this study can be accepted. Knowledge of purchase intention has a coefficient value of 0.109 with t value of 3.888. Therefore, it can be concluded that knowledge has a positive and significant indirect effect on purchase intention. Thus, H2 in this study can be accepted. Religiosity to attitude and religiosity to purchase intention have coefficient values of 0.427 and 0.211. Then each obtained t values as follows 10.314 and 6.732. It can be assumed that religiosity has a positive and significant direct and indirect influence on attitudes and purchase intentions. Thus, H3 and H4 in this study can be accepted. The results of the path coefficient calculation show that the attitude towards purchase intention has a coefficient value of 0.494 and t value of 10.013. Based on the coefficient value and t value results, attitudes towards purchase intentions have a positive and significant influence on purchase intentions. Thus, H5 is accepted. In addition, based on the calculation of the coefficient of the Islamic advertising path, it has a coefficient value of 0.276 and t value of 5.788. It can be concluded that Islamic advertising has a positive and significant effect on purchase intention. That is, H6 is accepted.
From the results of the path analysis, it can be concluded that all hypotheses (H1, H2, H3, H4, H5, and H6) have a positive and significant effect. So, all hypotheses can be accepted. Interestingly, the findings from this analysis show that the religiosity variable towards attitude and the attitude variable towards purchase intention has the highest significance value (t: 10.314 and 10.013). In addition, the indirect effect of the religiosity variable on purchase intention has a high significance value (t: 6.732) compared to the knowledge variable on purchase intention (3.888).

DISCUSSION
This study aims to determine and explain the influence of knowledge, religiosity, attitudes, and Islamic advertising on the purchase intention of local halal cosmetic products. In addition, this study also aims to find out and confirm the respondents' reasons for buying local halal cosmetic products and buying imported halal cosmetic products in the future. This study indicates that knowledge has a positive and significant influence on attitudes and purchase intentions of local halal cosmetic products. This means that consumers who know local halal cosmetic products will positively influence attitudes and purchase intentions. If consumers know a product, it will affect them in using or not using the product (Adriani and Ma'ruf 2020). The results of this study are in line with several previous studies (Adriani and Ma'ruf 2020;Haro 2018;Rohmatun and Dewi 2017). Furthermore, knowledge has a positive and significant influence on attitudes (Putri, Daryanti, and Ningtias 2019). However, the results of this study contradict the results of the previous studied (Abd Rahman, Asrarhaghighi, and Ab Rahman 2015).
The latest findings from this study show that the effect of Islamic advertising on the purchase intention of local halal cosmetic products has a positive and significant effect on the purchase intention of local halal cosmetic products. The addition of Islamic advertising variables in this study can differentiate indicators from previous studies. The results of this study are in line with several previous studies (Saifudin 2019;Sari 2020). In general, Islamic advertising is accepted by consumers as good advertising (Safira 2017). Advertisements found to have inserted elements of religious symbols influence the value obtained by consumers (Jayanegara and Najib 2020). Thus, it can be concluded that Islamic advertising is a good advertisement and can significantly influence consumers to buy local halal cosmetic products.
In addition, based on the analysis of Islamic advertising, it can be categorized as informative ads and has its value. Informative advertising can help increase consumer purchase intention (Dehghani et al. 2016). This proves that product promotions are intensely carried out during the months associated with the holy month, such as; Ramadan, Shawwal, and Dzulhijjah. Suppose consumers in some markets have a high level of religiosity. In that case, there needs to be adjustments and preparations in promoting these products according to the beliefs held by consumers themselves (Ateeq-ur-Rehman and Shabbir 2010). Then as a promotional strategy, companies need to display models that have a high level of religiosity in advertising halal cosmetic products (Abd Rahman, Asrarhaghighi, and Ab Rahman 2015).

CONCLUSION
The findings of this study indicate that all hypotheses are accepted. Knowledge and religiosity variables have a positive and significant effect on attitudes. The attitude variable and Islamic advertising have a positive and significant influence on the purchase intention of local halal cosmetic products. Then, the variables of knowledge and religiosity have a positive and significant indirect effect on the purchase intention of local halal cosmetic products. Interestingly, the findings of this study show that religiosity has the greatest significant effect on attitudes. Religious consumers will have a positive attitude in acting or making decisions. In addition, the attitude variable towards the purchase intention of local halal cosmetic products is the variable that has the second-largest significant effect. Suppose consumers have a good and positive attitude towards a product, the greater their potential to buy it. Islamic advertising variables in this study can be one of the differentiators and expand previous research. Unexpectedly, the research findings show that Islamic advertising has a positive and significant influence on the purchase intention of local halal cosmetic products. Islamic advertising can influence and attract consumers to use local halal cosmetic products. Then based on the results of qualitative analysis, respondents intend to buy local halal cosmetic products due to several reasons, such as quality products. The purpose of quality here is to have guaranteed ingredients to be safe, comfortable, and halal for sure. In addition, consumers also follow Islamic teachings and help local cosmetic products to be able to compete locally and globally and for other reasons. In addition, some consumers intend to buy imported halal cosmetic products for reasons of quality, interested in making comparisons between local and imported products.

LIMITATIONS OF RESEARCH AND FUTURE RESEARCH
The limitation of this research is the lack of coverage of the research area, considering that the respondents of this study came from only one area. I hope for the next research, the scope of the area can be expanded again. In addition, this research has not been specific to local halal cosmetic products. The following research can be focused on local halal cosmetic products. Especially local halal cosmetic products produced by Small and Medium Industries (SMEs). So, later it can help SMEs in the field of local halal cosmetic products. Next, the advice given is to use other variables that can influence the purchase intention of Muslim consumers and non-Muslim consumers towards local halal cosmetic products. Furthermore, subsequent research can determine the loyalty of Muslim consumers and non-Muslim consumers to a local halal cosmetic product brand. In addition, there are shortcomings in conducting qualitative analysis because there are manual calculations and subjective statement code analysis.