Awareness and intention to register halal certification of micro and small-scale food enterprises

Literature review

Studies on halal awareness and intention to register halal certification have been carried out in the last decade. Giyanti and Indriastiningsih¹⁶ hypothesized that awareness/intention to register a halal certificate is influenced by knowledge of halal, perception of benefits, and perception of procedures. The study results show that most SMEs food in Surakarta City, Central Java Province, Indonesia have a good knowledge of halal and agree that halal certification benefits their business. However, knowledge of halal does not significantly affect halal certification. Only the perception of the benefits of halal certification significantly influences the intention to obtain halal certification. They do not register halal certification due to a lack of understanding of the procedure for obtaining a halal certificate.

Abdul et al.,¹⁹ investigated the perception of halal certification among SME food entrepreneurs in Yogyakarta City, Indonesia. Food entrepreneurs who already have halal certificates report that the halal certification process is not tedious or stringent. They learn a lot about halal while going through the certification process. They also believe that halal certificates can increase market share and expand their business by instilling consumer trust in their products, i.e., providing a sense of security. Furthermore, halal-certified products are more competitive. On the other hand, SME food entrepreneurs who do not have halal certificates, suppose that the certification process is complicated and time-consuming. This was also stated by Viverita and Kusumastuti,¹⁴ Giyanti and Indriastiningsih,¹⁶ and Santosa et al.²⁰ It could be due to insufficient information about the process of obtaining a halal certificate and the benefits of having halal-certified products.^14,19 These studies indicate that intention to obtain halal certification is affected by the perception of the procedure.

According to Liba et al.,⁹ Elias et al.,²¹ and Masithoh et al.,²² awareness of halal is positively correlated with the intention to obtain halal certification. Meanwhile, Dinev and Hu,^23–26 Bachok et al.,²⁴ Lee and Shin,²⁵ Rezai et al.,²⁶ state that customer awareness is a strong predictor of a customer’s intent to buy or select a product. These indicate that awareness influences intention.

Waluyo²⁷ presumes that religious understanding, profit motivation, and level of education influence the awareness of halal-certified food producers in Sleman and Bantul, Yogyakarta Special Region, Indonesia. The significance test results show that the variables of religious understanding and profit motivation have a significant effect on awareness of being certified halal.

Halal awareness

The Law of Republic Indonesia No. 33/2014 describes halal products as those that conform to Islamic Sharia (principles). Carrions, blood, pigs, and/or halal animals (e.g., chicken, cow, goat, etc.) slaughtered in a manner inconsistent with Islamic Sharia are all considered non-halal (haram) materials. Furthermore, non-halal materials also include intoxicating plants or drinks, material that is harmful to one’s health, and microbes contaminated with non-halal materials. Halal encompasses substances (dzatihi), the nature of the substances, processes, processing areas, processing instruments, product storage, product distribution, and serving.²⁷ Based on these explanations, this study defines halal as what is permissible for Muslims to eat, drink, and use under Islamic law.

Awareness is defined as the state of being aware: knowledge and understanding that something is happening.²⁸ According to the definition of halal used in this study and the definition of the word awareness in the dictionary, halal awareness is then conceptualized as a process of being aware of what is allowed for Muslims to eat, drink, and use. The level of halal awareness is influenced by religious beliefs, exposure, the role of halal certification through the halal logo/label, health-related reasons,²⁹ genders, marital status,³⁰ religious knowledge, and motivation to gain profit.²⁷

The religious knowledge of halal considers knowledge of the laws relating to what allowed Muslims to eat, drink, and use as described in the Quran and Hadith. All foods are generally permitted except for those derived from prohibited animals such as pigs, dogs, and carrion, as well as foods and beverages containing alcohol and other toxic or dangerous substances. Slaughter must be carried out following Sharia, to do so in the name of God.³¹ Allah says in the Quran Surah (chapter) 2 (Al Baqarah) ayah (verse) 173 as follows³²:

Meanwhile, motivation to gain profit is defined as the food entrepreneur’s perception of the effect to be gained by producing halal food or having halal certificates and labels. The benefits of producing halal food include increasing market share and competitiveness,³³ business growth,¹ or business development.²

Attitude and intention to register halal certification

Several previous studies show that there is a positive relationship between attitude and intention to buy or choose a product.^34–36 According to the Planned Behavior Theory, the intention is determined by three independent factors, i.e., attitude toward behaviour, subjective norm, and perceived behavioural control. Supposing that there is a positive attitude that is supported by people around (as a subjective norm) and there is a perception of ease to perform the behaviour under consideration (as a behavioural control), then an individual’s intention to behave will be higher.³⁷

From an Islamic perspective, every Muslim must have an attitude to like and want to do a good job. In a broader context, attitude means to do good deeds due to Allah (God) loves those who do good as Allah (God) commands in Quran surah (chapter) Al-Baqarah ayah (verse) 195:³²

Attitude is also associated with two conditions i.e. good (‘mahmudah’) and bad (‘mazmumah’).³⁸ Hence, from an Islamic perspective and according to the Planned Behaviour Theory, if a Muslim produces halal foods as a do good (as an attitude toward a behaviour), supported by an awareness that Allah promises to love the good-doers (as a subjective norm), and there is a perception of ease of the procedure to obtain halal certification (as a behavioural control), then the individual’s intention to register halal certification will be higher.

The procedure to obtain the halal certificate is conceptualized as the food entrepreneur’s perception of the steps that must be taken to obtain a halal certificate and label. Standards for gaining halal certification in Indonesia are explained in the Law of Republic Indonesia No. 33/2014 concerning the guarantees of halal products.¹⁰

Identification of variables

According to Waluyo,²⁷ motivation to obtain halal-certified is significantly influenced by religious understanding and motivation to gain profit, because SME entrepreneurs generally agree that Halal Food Certification provided benefits.¹⁶ However, the procedure for obtaining halal certificates is relatively complex, thereby reducing the intention of SMEs to register halal certification. Referring to the Planned Behavior Theory³⁷, the intention to register for halal certificates is influenced by attitude to produce halal products, supported by halal awareness, and perception of the ease of procedure to obtain halal certificates.

Referring to Giyanti and Indriastiningsih,¹⁶ Waluyo,²⁷ and Ajzen,³⁷ this study identifies that the variables are knowledge of halal (KH), perception of benefits (PB), perception of procedures (PP), halal awareness (HA), attitude to produce halal foods (AHC) and intention to register a halal certificate. Measurement items of each variable are then compiled from the previous studies.^{10,16,27,29,30,33,39}

Conceptual model

Referring to Giyanti and Indriastiningsih,¹⁶ and Waluyo²⁷ this study considers that halal awareness is influenced by knowledge of halal and MSEs’ perception of the benefit of the halal certificate. Referring to the Planned Behavior Theory, halal awareness (as a subjective norm), attitude to produce halal foods (as a positive attitude), and MSEs’ perception of ease of procedures (as a behavioural control) will encourage the intention for registering halal certification. The relationship of these variables represents the conceptual model of halal awareness and intention toward halal certification (see Figure 1). We introduce a halal awareness between knowledge of halal or perception of benefit and intention to register halal certification.

Figure 1.

Conceptual model of awareness and intention to register halal certification.

As can be seen in Figure 1, halal awareness is a dependent variable, while knowledge of halal and perception of benefits are independent variables. Furthermore, halal awareness, perception of procedures, and attitude to produce halal foods are independent variables, while the intention to register halal certification is a dependent variable.

Hypothesis

The conceptual model (Figure 1) shows that the level of halal awareness is influenced by the knowledge of halal which may include MSEs’ understanding of the types of non-halal food as mentioned in Law of Republic Indonesia No. 33/2014.¹⁰ Therefore, we hypothesize that:

Halal certificates and halal logos are perceived to have benefits in increasing consumer confidence and competing with other producers,³⁰ using as a promotional tool.¹⁶ Hence, halal certificates are expected to improve the MSEs’ performance. Based on this point of view, we hypothesize that:

Halal awareness is measured by awareness of the importance of using halal materials in producing halal products,⁹ and perceiving the benefits to be gained despite the process being very strict.³⁰

Consumer awareness is a strong predictor of their intention to consume/purchase foods.^24,26 In this study, halal awareness is expected to influence on intention to register halal certification. Hence, we hypothesize that:

MSE entrepreneurs perceive that the procedure to achieve halal certification is complex due to the lack of information from respondents regarding halal certification procedures.¹⁶ This will negatively influence the intention of producers to register halal certification. In the light of this, we hypothesize that:

A positive relationship between attitude and intention has been shown by Rezai and Teng,²⁶ Jaafar et al.,³⁴ Yang et al.³⁶ These show that attitude influences intention. Hence, we hypothesize that:

Study design and participants

This study adopts a quantitative method to evaluate the hypothesis, i.e., analyze the data using descriptive statistics. This study has followed the STROBE guidelines/checklist for cross-sectional research.

Measurement items

The measurement items for each variable are identified in the following explanation. All variables, indicators, and measurement items are described in Table 1.

According to the previously defined, knowledge of halal in this study only includes knowledge about halal/non-halal (haram) materials according to the Quran and Hadith, as well as knowledge about the separation of equipment used to process halal/non-halal (haram) materials. The measurement items for the knowledge halal variable were adapted from Giyanti and Indriastiningsih¹⁶ and Waluyo.²⁷ The survey section includes four items, such as knowledge of; slaughtering methods (KH1), haram material and products (KH2 and KH3), and processing equipment (KH4).

The measurement items for the perception of benefits variable were adapted from Giyanti and Indriastiningsih,¹⁶ and Abdul.³⁰ These items asked about respondents’ perception of the benefits they would get if they had a halal certificate, such as a promotional tool (PB1), more convincing consumers to buy (PB2), improving business performance (PB3), and more competitive (PB4).

In this study, food entrepreneurs’ perception of the procedure for obtaining halal certificates in Indonesia involves the availability of information about the certification process and the fact that the certification process is simple, inexpensive, and quick. The measurement items for the perception of the procedure variable were asked about the respondent’s perception of the procedure for obtaining halal certificates in Indonesia, such as the existence of information about the certification process (PP1), the certification process is easy (PP2), cheap (PP3), and fast (PP4). These items were adapted from.^16,29

The Law of Republic Indonesia No. 33/2014 defines the halalness of products, including material supply, processing, storage, packaging, distribution, sales, and product presentation. According to our observations, most MSE food producers in West Java only produce one type of food, either sold directly at their “warung” (a little shop) or entrusted to someone else’s “warung”. Thus, they buy raw materials, process them into finished goods, and then store or ship one kind of product only. Hence, this study limits the scope of halal awareness as the awareness of using halal materials and processing them in a halal manner.

The measurement items for the halal awareness variable were asked about the awareness of respondents about the importance of producing halal food (HA1), using halal raw materials (HA2), the importance of having a halal certificate (HA3), and the rigorous of the halal certification process (HA4). These adapted from Abdul22 and Law of Republic Indonesia No 33/2014.¹⁰

This study conceptualizes attitude as thinking about producing a halal product that includes a focus on halal issues, a guarantee of selling halal products, halal product attention, and checking to use of halal raw materials. Hence, the measurement items for attitude to produce halal foods variable include respondents’ attitudes to always pay attention to halal issues (AHC1), ensure consumers buy halal products (AHC2), pay attention to halal products (AHC3), and use halal materials (AHC4). These items were adapted from Ambali and Bakar21 and Menteri Hukum dan HAM (Minister of Law and Human Rights-English).¹⁰

The intention is conceptualized as an act of registering the product to obtain a halal certificate. The measurement items for the variable of intention to register halal certification were adapted from Law of Republic Indonesia No 33/2014,¹⁰ Abdul et al.,³³ and Ngah et al.³⁹ These items include respondents’ intention to be responsive to the certification process (IHC1), strive to meet halal quality standards (IHC2), immediately implement halal assurance system (IHC3), and immediately register halal certification (IHC4).

Population and sample

The study was carried out in West Java Province, Indonesia. The population is the MSE food producers listed in the Central Bureau of Statistics of West Java, such as producers of cassava chips, shredded catfish, “bagelen” cakes, chocolate “rangginang”, candied vegetables, market snacks, and so on. A copy of the list of West Java MSEs registered in the Office of Cooperatives and Small Businesses in West Java Province obtained from the Central Bureau of Statistics of West Java can be found in Underlying Data.⁴⁰

The Categorization of business scale refers to the “Undang-Undang Republik Indonesia Nomor 20/2008” (Law of the Republic of Indonesia Number 20/2008).⁴¹ According to that law, the criteria for micro-enterprises are as follows:

Whereas in article 6 paragraph 2 it is stated that the criteria for small businesses are as follows:

Table 2 shows the categorization of the business scale.

MSE food producers chosen as respondents include those who meet the legal criteria. Furthermore, the purpose of this study is to determine the awareness level of MSE food producers regarding halal certification. Hence, the determined criteria for respondents were as follows:

The purposive sampling technique was used in this study. The total number of food producers of MSE listed in the Central Bureau of Statistics of West Java was estimated to be around 2300 people. We worked with “Sahabat UMKM Jawa Barat” (West Java MSME association-English), a local association of micro, small, and medium-scale entrepreneurs who are engaged in a variety of fields such as culinary, fashion, crafts, and other businesses or industries. The local association has over 1,000 members, with 68 percent of them being food and beverage entrepreneurs, or approximately 680 entrepreneurs. This study’s sample consists of food producers who meet the determined criteria among the 680 entrepreneurs.

We received written permission from the Chairman of the West Java MSME association to contact his members for data collection. The written consent to participate from the Chairman of the West Java MSME Community was gained according to document number: 015/SKIP/IV/2020. To guarantee that there is no conflict of interest in this study, survey responses are kept anonymous.

Ethical consideration

The Ethical Licensing Committee of the Islamic University of Bandung approved this study by Protocol number: 495/B.04/Bak-k/XII/2019. We provided all respondents with a consent statement after consultation. In the questionnaire, there is a statement that by filling out the questionnaire the respondents gave their consent to participate. Respondents gave their consent to take part when they filled out the questionnaire. Respondents had given their consent truly and without coercion. Furthermore, to protect respondents’ rights and privacy, all forms of data obtained will be kept confidential.

Data collection

A questionnaire is chosen as the research instrument. The questionnaires were re-translated from English to Indonesian, except for those references that were already in Indonesian. Each variable is made up of measurement items which are scored on a Likert scale of 1 (strongly disagree) to 5 (strongly agree).

The West Java MSME association, which has over 1,000 members with 680 food and beverages entrepreneurs, has set up several WhatsApp groups to help them communicate with one another. Some groups exist due to the limited number of members who can join a single WhatsApp group. We collect data through these WhatsApp groups. A Google Form with a cover letter and a set of questionnaires were sent out electronically to the potential respondents who are members of the West Java MSME association WhatsApp groups. without separating food and non-food entrepreneurs. This method was chosen because of the COVID-19 pandemic outbreak. In addition, the designed questionnaires could be collected without conducting direct visits to the respondents. The respondents could not participate in the survey unless they gave their written consent. Data were collected from March to May 2020. A copy of the distributed questionnaires can be found in Extended Data.⁴²

The questionnaire was pretested with a small sample of members of the West Java MSME association before being distributed to the actual respondents. Based on pretest feedback, the wording of some items was refined and modified to guarantee that the validity and reliability of each variable meet the required standard. The question items were scored on a Likert scale of 1 (strongly disagree) to 5 (strongly agree). The follow-up of this plan is described later in the data preparation section.

The purposive sampling method is used in the following manner.

Structural model analysis

Data are analyzed with descriptive statistics to provide a description of the respondents’ profile, and to describe the results of the assessment of the level of knowledge of halal, perception of benefits, halal awareness, perception of procedure, attitude, and intention to register halal certificates. The research analysis is intended to assess the model and objectively describe the hypotheses.

The adopted theoretical models are evaluated using Partial Least Square – Structural Equation Model (PLS-SEM) with a path model because this study is exploratory research to predict certain constructs by focusing on explaining the variance in the dependent variables when examining the model.⁴³ A Smart-PLS software is chosen for data processing.

There are two elements in the PLS-SEM with path models: the outer model and the inner model. The outer model (also known as the measurement model) describes the relationships between latent variables and their indicators, whereas the inner model (also known as the structural model) depicts the relationships between latent variables.^43,44

The assessment of the PLS-SEM model begins with the measurement models (outer model) by evaluating the quality of the reflective or formative measurement models. The reflective measurement models are estimated by assessing the construct measures’ reliability and validity. Composite reliability (as a means of assessing internal consistency reliability), convergent validity, and discriminant validity are among the specific measures. Formative measures are evaluated for convergent validity, indicator weight significance and relevance, and the presence of collinearity among indicators.^43,44

Following confirmation that the construct measures are reliable and valid, the structural model (inner model) results are evaluated. This entails investigating the model’s predictive abilities for theory testing, as well as the relationships between the constructs. The PLS-SEM model fit is evaluated using standardized root mean square residual (SRMR), root mean square residual covariance (RMS_theta), or the exact fit test to determine how well it predicts endogenous variables/constructs.⁴³

The first step in evaluating the PLS-SEM results for the structural model is to look at the significance and relevance of the coefficients. The bootstrapping routine and examining t values, p values, or bootstrapping confidence intervals are required to test the significance. Despite this, the bootstrapping confidence interval is less common.⁴³ Following that, the relative sizes of path coefficients, total effects, f² effect size, Q² effect size, and q² effect size can be compared. By interpreting these findings, the key constructs with the greatest relevance to explaining the endogenous latent variable(s) in the structural model can be identified.^43–45 Concisely, the systematic evaluation of PLS-SEM output is shown in Table 3.

Sample size

Barclay, Higgins, and Thompson (1995) in Hair et al.⁴³ explain the Ten Times Rule in determining the number of PLS-SEM samples, which states that the sample size must be greater than (1) ten times the greatest number of formative indicators used to measure a single construct, or (2) ten times the greatest number of structural paths directed at a specific construct in the structural model. In other words, the minimum sample size is equal to 10 times the maximum number of arrows in the PLS path model pointing to the latent variable.⁴³

In this study, the IHC variable is the latent variable with the maximum number of arrows, i.e., 3 (see Figure 1). As a result of the Ten Times Rule, 3.10 = 30 represents the bare minimum of observations required to estimate the PLS path model depicted in Figure 1. In terms of Cohen’s (1992) recommendation for multiple OLS regression analysis, or running a power analysis using the G*Power program, as cited in Hair, et al.,⁴³ 33 observations are required to detect an R² value of about 0.25, assuming a statistical power of 80% and significance level of 5%.

Participants

The questionnaires were distributed to the members of the West Java MSME association’s WhatsApp groups. The questionnaires were returned by 376 respondents, including food and non-food entrepreneurs with micro, small, and medium-scale businesses. All respondents who returned the questionnaires were then selected based on the determined criteria with the following stages:

We examined the responses of the selected and non-selected respondents to see if there was any possibility of non-response bias. Lindner et al.,⁴⁶ proposed three methods for investigating non-response bias: (1) comparing early to late respondents, (2) using “days to respond” as a regression variable, and (3) comparing respondents to non-respondents. This study adopted the third proposed method of Lindner et al.,⁴⁶ by comparing selected and non-selected respondents. We examine 20% of both selected and non-selected respondents. That is, 27 respondents (20% of 137) were taken from the selected group, while 47 respondents (20% of 239) were taken from the non-selected group.

The independent samples t-test was used to compare the responses of the two groups due to the difference in the number of respondents examined.

The findings show that there is no bias when none of the respondents in the two groups tested registered their products to obtain a halal certificate. When the non-selected respondents were a mix of those who had not yet registered and those who had already obtained halal certificates, two indicators were discovered to be biassed in the perception of procedure variable (PP), namely PP2 and PP4 indicators, or in the perception of benefit variable (PB), namely PB4 indicator. This indicates that respondents who have obtained halal certificates have different perceptions than those who have not, particularly regarding perceptions of simplicity (PP2) and the length of time to obtain a halal certificate (PB4), as well as benefits such as increased competitiveness from having a halal certificate.

Table 4 displays the percentage of respondents for each indicator where 98.5% of respondents have Islam as their religion (Muslim), 67.9% are female, and 38% are 26 years old or older. In terms of business size, 89% of respondents are micro-scale entrepreneurs. Non-Muslim respondents are involved because producing halal food regardless of the religion of the food producers.

Descriptive statistics of the research variables

Table 5 shows the descriptive statistics for each measurement indicator. The knowledge of halal (KH) has a high average perception value.

Data preparation

A copy of the dataset of the questionnaire result can be found in Underlying Data.⁴⁷ There are no missing values, invalid observations, or outliers in this data set, which has a sample size of 137.

Initially, the questionnaires were first to be pretested in a small sample of MSE food producers in West Java, Indonesia, to assess the research instrument’s validity and reliability. The questionnaires are planned to be distributed in person starting in March 2020. However, we were unable to meet with the food producer due to the social distancing caused by the COVID-19 outbreak. Finally, we decided to distribute the electronic questionnaire Google Forms in May 2020 to the West Java MSME association’s WhatsApp groups. In just a few days, the data collected reached 137 respondents who met the predetermined criteria, so all collected data were subjected to validity and reliability tests.

Measurement model (Outer model) analysis

Based on the conceptual model shown in Figure 1 and the measurement items in Table 1, the structural model involves the following 2 models:

The constructs or latent variables in this structural model are KH, PB, HA, PP, AHC, and IHC. Each latent variable is made up of some indicators that are highly correlated and interchangeable, allowing them to be reflective. Hence, the causality flow is going from the construct to the indicators. It means that any changes in the construct are expected to be reflected in all of its indicators (see Figure 2). Reflective measurement models should be evaluated for their reliability and validity. The path analysis for the proposed model is obtained by running Smart PLS-SEM software, as shown in Figure 2.

Figure 2.

PLS-SEM result of the first path analysis.

Outer model loading

The correlation between the latent variable and the indicators in its outer model is evaluated by an outer loading. The outer loading of the first path model is shown in Table 6 where the outer loading of all indicators is greater than 0.7, indicating that it meets the convergent validity criteria. However, when the discriminant validity was examined, it was discovered that the correlation between AHC-HA is higher than AHC-AHC, while the correlation between IHC-HA is higher than HA-HA (see Table 7), indicating that the path model does not meet the Fornell-Larcker criterion.

Hence, the model’s feasibility must be reconsidered⁴⁵ by analyzing the multicollinearity to determine whether any indicators should be merged into one or eliminated.^43,44 Collinearity is assessed by calculating a variance inflation factor (VIF) for each indicator and comparing these VIFs to a threshold. The VIF threshold values of 10, 5, and 3.3 are commonly recommended for collinearity, which means that a VIF equal to or greater than the threshold value indicates a potential collinearity problem.⁴⁸ As a result, the corresponding indicators should be considered to be removed.⁴³

Table 8 displays the VIF value of each indicator of the first path model. As can be seen, it was discovered that the HA2 indicator had a VIF value greater than 10, so the HA2 indicator was considered to be discarded. After the HA2 indicator is removed, the processing is performed on the second path model, which does not include the HA2 indicator. Based on the discriminant validity analysis, it was discovered that the correlation between IHC-HA is higher than HA-HA. Hence, the VIF value should be checked again. The calculation result shows that there are no VIF values greater than 10, but the IHC2 indicator has close to 10 values of 9.02. To meet the Fornell-Larcker criterion, the IHC2 indicator is being considered for removal. The third path model is then constructed without the use of the HA2 and IHC2 indicators. When the discriminant validity of the third path model was examined, it was discovered that all latent variables met the Fornell-Larcker criteria. Consequently, the final path model for the problem discussed in this study is the third path model.

The assessment of PLS-SEM output

In this study, both validity and reliability tests are carried out to measure the goodness of the shared questionnaires. Validity is a test of how well the developed instrument measures the particular construct being measured, while reliability is a test of how the developed instrument consistently measures the construct being measured.⁴⁹ The assessment’s steps are as follows:

According to the Systematic evaluation of PLS-SEM output described in Table 3, the findings of the reflective measurement model evaluation are summarized in Table 9. As can be seen, all items’ loadings exceed 0,7, the AVEs for the indicators are within the range of 0.760 and 0.893, all CR values are higher than 0.6, and the Cronbach Alpha’s are within the range of 0.892 and 0.960. Therefore, all model evaluation criteria have been met, indicating that the instruments are reliable and valid.

The square root of the AVE of each construct should be greater than its highest correlation with any other construct, according to the Fornell-Larcker criterion.⁴³ Table 10 shows that the average variance extracted by the indicators measuring that construct is less than the squared correlations for that construct. To put it another way, the measurement model represents adequate discriminant validity.

Another method to determine discriminant validity is assessed by looking at loading and cross-loading to identify problem items if there are any. The validity test using cross-loading is patterned in that the main loading factor originating from its construct is greater than the correlation value built from these variables on other constructs.⁴³ Table 11 presents an evaluation of validity based on the value of the main loading factor to the value of cross-loading factors with other constructs. As shown in Table 11, the value of the main loading factor of each construct is higher than the value of the loading factor outside of the main loading factor, so it can be concluded that all constructs are declared valid. For example, the loading factor of indicator AHC1 in the AHC construct is highest than its loading factor in HA, IHC, KH, PB, and PP.

This study uses the standardized root means square (SMSR) to assess model fit. The SMSR is 0.067, less than 0.08, indicating that the model meets the model fit criteria. Furthermore, the data can be used to estimate the model.

Significance testing of both the inner and outer model in SmartPLS uses bootstrapping procedure to give approximate t-values. In this study, the hypotheses are represented with a positive path coefficient, so that the t-value is determined by the one-tailed t-test. By using a significance level of 5%, the path coefficient will be significant if the t-value is larger than 1.65.⁵⁰ Figure 3 shows the path analysis result of halal awareness and the intention to register halal certification using Smart PLS software, while Table 12 represents the hypothesis testing results of each path.

Figure 3.

Result of the final path analysis.

The path analysis is addressed to ascertain the hypotheses put forward. As can be seen in Figure 3, the R² value of 0.825 for IHC indicates that 82.5 % of the variance in IHC can be explained by HA, PP, and AHC. In addition, Figure 3 also shows that HA, PP, and AHC are positively related to IHC among MSE entrepreneurs with β = 0.696, 0.080, and 0.202, respectively. According to the t-value of the path coefficients, HA has a significant impact on IHC, while AHC and PP do not have a significant impact on IHC. Besides, the R² value of 0.841 for HA means that 84.1% of the variance in HA is influenced by KH and PB, with values β = 0.169 and 0.791 respectively. Hence, KH and PB have a strong and significant impact on the halal awareness of MSE food producers.

According to Hair, et al.,⁴³ R² values of 0.75, 0.50, or 0.25 for the endogenous construct, respectively, can be described as substantial, moderate, or weak. The R² value of an endogenous latent variable (i.e., halal awareness) described by the two predictive constructs in this study is 84.1% (see Figure 3), which is substantial. Hence, KH and PB are genuine predictors of halal awareness of food producers of MSE. The same goes for the R² value of the endogenous latent variable (i.e., Intention to register halal certification) is 82.5% which is substantial. However, the genuine predictor of MSEs’ Intention to register halal certification is only HA.

In addition to evaluating the R² values of all endogenous constructs, the contribution of an exogenous latent variable to the R² value of an endogenous latent variable, known as the f² effect size, must be evaluated. In simpler words, the f² effect size measures the strength of the relationship between the latent variables.⁴³ Table 13 displays the f² values for all endogenous construct combinations (represented by columns) and corresponding exogenous (i.e., predictor) construct combinations (represented by the rows). As can be seen, HA has a large effect of 0.462 on IHC, while PB has a large effect of 2,035 on HA. Meanwhile, AHC (0.037) and PP (0.033)) have a medium effect on IHC, and KH (0.093) has a medium effect on HA.

Predictive relevance is another aspect that can be investigated for the inner model. The value of cross-validated redundancy (Q²) can be used to calculate the predictive value of relevance. If the Q² value is greater than zero it indicates that the model has predictive relevance accurate to certain constructs, otherwise the model lacks predictive relevance.⁴³

The blindfolding procedure in SmartPLS can be used to calculate the Q² value. The output of the blindfolding procedure of the model discussed is shown in Table 14. In the table, SSO represents the sum of squared observations, SSE represents the sum of squared prediction errors, and the final column (i.e., 1 - SSE/SSO) represents the final value Q², which is interpreted to assess the model’s predictive relevance for each endogenous construct.

As can be seen, the Q² values of two endogenous constructs (HA and IHC) are significantly higher than zero where the Q² value of IHC is greater than HA. These findings provide strong support for the model’s predictive relevance for endogenous latent variables. Hence, predictions for HA and IHC are accurate.

Underlying data

Figshare: Dataset of Questionnaire Results from the respondents of Awareness and Intention to Register Halal Certification; https://doi.org/10.6084/m9.figshare.20488317.⁴⁷

This project includes the following underlying data:

Data are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).

Figshare: Dataset of West Java MSE food and culinary in 2021; https://doi.org/10.6084/m9.figshare.21618096.⁴⁰

This project includes the following underlying data:

Data are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).

Extended data

Figshare: List of questions of descriptions of the questionnaire of Awareness and Intention to Register Halal Certification; https://doi.org/10.6084/m9.figshare.20488590.⁴²

This project includes the following extended data:

Data are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).

Figshare: The Profile of Respondent of Awareness and Intention to Register Halal Certification; https://doi.org/10.6084/m9.figshare.20488650.⁵⁷

This project includes the following extended data:

Data are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).