Determinants of Farmer Participation and Development of Shallot Farming in Search of Effective Farm Management Practices: Evidence Grounded in Structural Equation Modeling Results

Abstract

The objective of this research was to examine the determinants of farmer participation and shallot-farming development in search of effective farm management practices. The study used structural equation modeling data analysis. The primary data were collected from direct structural interviews with 150 randomly chosen shallot farmers in Bantaeng Regency, Indonesia. It was found that the latent variables of Physical Aspects of Land, System of Economy Peasant Society, and System of Political Peasant Society were fundamental factors that exerted a positive and significant influence on the latent variable of Farmer Participation. Therefore, improvements in the physical aspects of the land, the economic framework, and the political structure of agricultural communities could promote farmer participation. Furthermore, the latent variable of Farmer Participation and System of Political Peasant Society had a positive and significant impact on shallot-farming development. Thus, by increasing the influence of government officials and community leaders, shallot farming can be promoted. The farmers can then enhance their participation in shallot-farming plan formulation and implementation, providing the continued development of shallot farming. The findings of this study contribute significantly to the body of knowledge by validating previous research and proposing different ways to improve effective farm management practices in shallot farming.

1. Introduction

Farmers’ active participation in agricultural development is one indicator of how successful agricultural development initiatives are. It has a significant impact on the success of the development process [1], influences the success of agricultural development [2], and is a critical factor in the sustainability of agricultural development in rural regions [3]. However, farmer participation in agribusiness institutions’ activities has mostly been limited to production activities, and it has not been fully optimized [4]. The agricultural sector is essential in economic development, particularly in developing countries [5], where it is a priority [3] and an essential part of domestic economies [6], as well as a strategic sector at the national and regional levels [7]. Furthermore, the sector is strategically responsible for accomplishing the first Millennium Development Goals (MDGs), which aim to eradicate extreme poverty and hunger [8]. Thus, rural farming communities play a significant role in economic growth and are fundamental parts of a country. A country’s economic development is successful if it improves its citizens’ well-being [9]. As a result, when managing the sustainability of agricultural growth, the development process must be viewed as a negotiation between the community’s expectations and the government’s desires [10]. Efforts in this direction undoubtedly necessitate an integrative strategy considering different dimensions within the agricultural system [11].

Highly effective progress in the economy necessitates the active and timely involvement of all interest groups in the formulation, execution, and assessment of development initiatives that impact their concerns [12]. Development initiatives will likely experience an upsurge when these interest groups perceive their involvement as significant, productive, and streamlined [13]. Community participation, as defined by Adi [14], encompasses the proactive involvement of individuals in the identification of societal challenges and prospects, the formulation of decisions concerning potential resolutions, the execution of approaches to tackle issues, and the assessment of the outcomes. Community participation generally refers to voluntary community involvement in development programs based on the community’s self-awareness and determinants [12]. Farmers will engage if they have the ability, willingness, and opportunity to do so [15]. In order to stimulate active community participation in the development process, revitalizing farmers’ organizations has emerged as an important paradigm to examine [16] and an essential strategy for establishing an agricultural system [17]. In Indonesia, strengthening farmer participation in farmer groups, cooperatives, and unions is essential to revitalizing farmer organizations [16]. Farmers actively take a part in the participation process by identifying problems, defining problems, exploring alternative solutions, creating solution plans, developing processes, and monitoring implementation and active evaluation [18]. To encourage farmers’ active participation in agricultural development activities, agricultural extension workers should be more assertive in educating farmers about new agricultural technologies that can be implemented to improve their welfare.

The Indonesian government has implemented initiatives to enhance farmer participation in agricultural activities since the presidency of Soeharto from 1966 to 1998. These initiatives have involved establishing and growing farmer organizations nationwide specializing in diverse agricultural commodities. A recent policy from the Indonesian Ministry of Agriculture is Regulation 67/Permentan/SM.050/12/2016, which pertains to the institutional development of farmers. The government defines farmer groups as entities that advocate for and fortify the concerns of farmers [19] in this regulation. Furthermore, in implementing regional development, Law No. 16 of 2006 on Extension Systems, Agriculture, Fisheries, and Forestry stresses the significance of local knowledge and community participation [20]. The two previously mentioned government regulations are significant indicators of the Indonesian government’s commitment to fostering farming community participation to accelerate agricultural development. In Indonesia, Bantaeng Regency is one of Indonesia’s regencies in the Province of South Sulawesi. It assertively expands farmer participation in the management of shallot cultivation. Bantaeng Regency’s local administration has developed a shallot farmer organization to assist farmers in increasing their shallot crop production. More specifically, the government forms farmer groups to facilitate learning, serve as a production unit, create cooperation, and conduct business operations [21]. The farmer associations are founded with the motto “from, by, and for farmers” [19]. Furthermore, farmer organizations for shallot farming are developed to maintain family, familiarity, and trust among farmers.

The above explanation demonstrates the essential importance of farmer participation in the advancement of farming. Facilitating farmer participation is an essential aspect of promoting the development of agriculture. Farmer participation and the growth of agricultural commodities are influenced by various factors, including the economic system, political system, communication system, socio-cultural system, level of education, and farming management elements established by the farming community. Hence, the objective of this research was to examine the determinants of farmer participation and shallot-farming development in search of new effective farm management approaches. It is expected that the findings of this research will serve as a guide for the development of shallot farming in the near future. Furthermore, it is hoped that these findings will have a significant influence on the expansion of shallot farming and the promotion of farmer participation. It is our intention that by the end of this research, we will have found a number of important factors that have the potential to contribute to an increase in the number of farmers participating in shallot farming in particular, as well as an innovative management approach that will lead to effective farm management practices (EFMPs) in shallot farming. In this study, effective farm management practices refer to the best management practices (BMPs) in farm management. In the existing literature, BMPs refer to a method of enhancing agricultural sustainability [22]. It is a practical approach to farming that integrates traditional techniques with suitable modern technology, aiming to achieve both profitable crop production and effective environmental management [23]. The fundamental elements of BMPs include crop management, nutrition management, pest control, and financial management [23]. Hence, in this study, we define EFMPs as the ability of farm management to successfully meet production goals within a given timeframe. Finally, in the next section, we will delve into four key themes: (i) in Literature Review, we review all aspects of latent variables examined in this research, including the construction of the conceptual framework; (ii) in Research Methods, we present the research site, structural equation modeling (SEM) development, data collection and sample size, and data analysis; (iii) Results and Discussions; and (iv) Conclusions and Recommendations.

2. Literature Review

A literature review serves the crucial function of gathering, analyzing, and scrutinizing the theories and research findings of past researchers that pertain to the topic under investigation. Armed with this knowledge, a researcher can effectively guide their research plan. Based on our literature review results, we identified many factors that influence farmer participation and the development of shallot farming, including land suitability, production costs, marketing costs, length of education, and so on. Next, we categorized the various factors that theoretically influence farmer participation and shallot-farming development based on their respective characteristics. This category resulted in six exogeneous latent variables in the study area that were hypothesized to influence farmer participation and shallot-farming development, as described below.

2.1. Physical Aspects of Land

The land’s topography is a substantial factor that can influence farmers’ decisions regarding various agricultural management facets. Furthermore, an analysis of the land’s physical attributes is imperative, as it exerts both direct and indirect influences on the level of achievement of the cultivators [24]. The term “land suitability” refers to the capacity of a specific land type to facilitate the growth and development of particular commodities [25]. In the context of sustainable agriculture, it is the utility of land [26].

Land suitability is essential to enhancing land productivity and optimizing land use utilization of [27,28]. Topography refers to the slope of the land contour, which includes elevation changes and significantly impacts biological processes. According to Suparno and Marlina [29], a more extensive land contour indicates a steeper slope. The influence of topographical features on the productive capacity of crops has been chiefly disregarded till now [30]. The larger the land contour, the greater the slope of land [29]. The impact of topographical characteristics on crop production potential has been largely ignored [31]. Hence, it is imperative to thoroughly evaluate the impact of topography to understand the appropriate strategies for shallot cultivation comprehensively.

Furthermore, accessibility is a catalyst of development [32], which is practically related to the mode of reaching a location. Accessibility is a criterion in site selection [33,34] for agricultural business development because it affects transportation costs [35] for inputs and agricultural production. Therefore, talking about accessibility is not only about the accessibility of the nearest markets in rural areas but also about access to opportunities and any valuable possession [32]. Accessibility can be described through the quality and availability of road and transportation networks. According to Baja [36], roads with higher quality and nearer to residential areas are projected to be more accessible. This accessibility is anticipated to increase agricultural product production and productivity. Shallot plants often develop tubers in regions where the average air temperature reaches 22 °C. Shallots develop larger bulbs when cultivated in regions receiving over 12 h of sunlight. Therefore, shallot plants prefer thriving in low-lying areas with a sunny climate [37]. These findings indicate that the growth and production of shallots are limited by soil fertility and high rainfall features.

2.2. System of Economy Peasant Society

Production costs are the value of all production factors used, in the form of both goods and services, during the production process [38]. Shallot production and productivity are closely related to production factors [39,40] and production costs allocated by farmers to their shallot farming [41]. In the study by Syam’un et al. [42], the primary obstacle to increasing shallot production and productivity was the producers’ high production costs, which ranged between 50 and 75 million IDR per hectare. Consequently, shallot cultivation incurs additional marketing costs for producers, including transportation, storage, and other fees. The marketing expenditures that farmers are responsible for are contingent upon the farm’s proximity to the market and the prevailing market structure [41].

Additionally, marketing expenses are influenced by collection, sorting, packing, and distribution costs [43]. Furthermore, the scale of marketing agency operations and the level of use of marketing facilities also have a role in determining marketing costs [44]. The costs associated with marketing affect the direct advantages farmers gain and influence the strategic development of shallot farming.

The ownership of agricultural capital by farmers is a crucial determinant of the success of shallot farming, as it directly affects the substantial costs involved in the production process [41]. The farming capital is cash saved for investment purposes [45] in farming activities. Capital plays a significant impact in acquiring industrial facilities and determining salaries for labor [46]. Thus, the availability of farming capital for shallot farmers is needed to ensure a smooth production process. Then, the availability of labor in the shallot production process is a production factor that is no less important. A labor shortage will delay planting time, ultimately affecting plant growth, productivity, and farmers’ income [46]. In contrast, a rise in the workforce can effectively enhance the productivity of agricultural enterprises [47].

2.3. System of Political Peasant Society

The political system of the agricultural community encompasses the community’s function, community leaders’ engagement, and government officials’ participation in the administration of shallot farming, together with the prevailing price policies and politics. The empirical data from rural areas illustrate the substantial impact of community leaders on village development politics, namely, in comprehending and meeting the community’s aspirations [48]. Further, the village government utilizes the outcomes of collecting community aspirations as primary data, input, and supplementary information in formulating policies for village development [49]. The participation of the community in an activity will have an impact on the attainment of shared objectives. The research findings indicate that both the community and the government have a significant role in the limited political engagement of the community in development planning [50]. More substantial community participation in development initiatives directly correlates with higher rates of success in development programs [51], which can be felt by the community together.

Extension staff and community leaders continue to play a prominent role in motivating farmers. Therefore, the choice of communication channel can be primarily influenced by interpersonal media, such as demonstration plots, technological degrees, field meetings, or group meetings at the regional level. There is a requirement for active assistance from institutions to support farmers [52]. Community engagement and participation in politics are also shaped by the dynamics of political communication and the circulation of public opinion among them [53]. Finally, an integral aspect of the political system within the farming community is the formulation of a price policy that encompasses factors of production and agricultural output, such as shallots. The price policy pertains to determining prices for the components of production involved in shallot cultivation and the management of production yields and price stability [54]. The objective is to minimize uncertainty for farmers and guarantee the availability of shallot commodities to customers at fair rates. Therefore, it is imperative for the government to coordinate the establishment of production centers, allocate crop outputs among different locations, and oversee and assess rules regarding the pricing of shallots [55].

2.4. Communication System of Farming Society

The assessment of the communication system in agricultural communities is conducted by evaluating many factors, including the frequency of farmer group meetings, the frequency of visits by extension workers, the level of contact between farmers and extension workers, and the presence of communication media. The degree of engagement exhibited by farmers in farmer organizations is positively and significantly correlated with their proficiency in agricultural land management [56]. The study findings demonstrate the successful attainment of the objective to establish farmer groups, primarily aimed at enhancing and advancing the competencies of farmers as critical participants in agricultural progress [57] so that farmers can run their farming business together [58,59]. Then, the visits of agricultural extension workers to farmers and their farmer groups are also important parameters in measuring the effectiveness of the farming community communication system.

The agricultural extension system using the training and visiting approach, which has been implemented since 1996, was very effective in increasing farmers’ knowledge, attitudes, and skills so that the Indonesian government was able to achieve self-sufficiency in rice in 1984 [60]. The training and visiting approach is carried out by using two methods: visiting and working [61]. Agricultural extension workers carry out the visit method by visiting farmers at their homes. In contrast, the visitation method is carried out by visiting farmers and their farmer groups on their farming fields while they are carrying out their activities [61]. The contact between farmers and extension workers is a crucial indicator of the communication system within the farming community. The study findings clarify that the frequency of interactions between farmers and agricultural extension agents substantially impacts farm income [62]. However, there are still many barriers to communication and interaction between farmers and agricultural extension workers. The obstacles farmers mostly perceive include attention and familiarity, prejudice, differences between expectations and needs, farmer experience, and the cosmopolitan character of farmers [63]. Therefore, the effectiveness and frequency of interaction between farmers and extension workers are significant, and they are expected to influence farmers’ behavior and adoption rate toward innovations in the shallot commodity. In this regard, the competency of extension workers in mastering science, technology, and knowledge transfer is the main requirement. Therefore, Bahua [64] explains that the competence of extension workers is closely related to the skills of shallot farmers through the dimensions of personality competence, andragogic competence, professional competence, and social competence. Then, the availability of communication and information media in the form of brochures and leaflets in information dissemination activities has advantages because it can reach more targets and spread farther than face-to-face communication [65]. The study’s findings indicate that extension workers prefer using flipchart media, images, videos, slides, and pamphlets as counseling tools [66].

2.5. Socio-Cultural System of Peasant Society

The government should prioritize the socio-cultural system of the agricultural community while aiming to enhance farmers’ involvement in the management of shallot farming. The farmers in many places in South Sulawesi Province maintain a strong adherence to the traditional practices of tudang sipulung (sitting together for discussion and consensus), gotong royong (cooperation), sistem ijon (pre-harvest purchase and payment of crops), and patron–client relationships. Tudang sipulung, as a form of group communication, is a forum for sharing information with other groups [67]. Etymologically, tudang sipulung means sitting together as part of a group meeting. This activity means gathering to deliberate matters considered important by the local community [68]. Deliberation can be interpreted as negotiating, brainstorming, or saying and proposing something or deliberations, which are known as syuro (a group of community members who are elected to consult to find solutions to problems that arise), “village deliberations”, or deliberations/negotiations [69]. In carrying out these activities, there is a shared value system (for example, deliberation, religious, solidarity, obedience, modesty, and togetherness values), which serves as guidelines for implementation [70]. The tudang sipulung culture effectively intermediates between the local government and the community. The government’s perspective promotes an open mindset throughout the planning and budget reporting processes, fostering efficient management and transparency in sharing information [71]. Furthermore, this will foster a sense of social responsibility and enhance confidence in the policies implemented by the local government [71]. Then, the culture of gotong royong, a community cultural system in various regions throughout the archipelago, including at the research locations, continues to be a way of life for Indonesian people [72]. This culture is practiced as a reflection and implementation of “Pancasila” (Five Principles of Indonesia), “Bhinneka Tunggal Ika” (Unity in Diversity), and the democratic system [72]. The cultural qualities of gotong royong manifest in collective village cleaning activities, exemplifying the village community’s solidarity reinforcement [73]. Another form of gotong royong culture in the countryside is building food barns. The presence of a food storage facility holds significant influence, not only in enhancing food security but also in bolstering the town’s economic vitality [74].

Moreover, the farming community still upholds the socio-cultural practice known as the bondage system. Sistem ijon is a culture of farmers who sell crops to intermediaries before their agricultural production is ripe [75]. It is considered an informal credit transaction system [76]. This system is prohibited in Islamic law [77] but still occurs in farming communities. Farmers use this system when they need funds [75]. The last socio-cultural system we use as a parameter in our research is the patron–client cultural system, namely, the vertical relationship between superiors and subordinates [78]. This patron–client relationship already exists in various agrarian community groups, which are symmetrical or asymmetrical [79] and have quite an important role in developing the village economy. This culture, apart from existing in food crop farming communities [80] and perennial crops, is also commonly found in fishing communities [81,82,83].

2.6. Education Level

The educational attainment of the farming population is a significant determinant of their involvement in agricultural management [84]. From the theoretical perspective, significant determinants of educational attainment that currently have considerable influence are educational duration, informal education, social media literacy, and proficiency in utilizing social media. In their study, Ross and Lappin [85] discovered that education is a significant determinant of community involvement in a program. In addition, various empirical findings indicate that non-formal education factors strongly correlate with the level of member participation in the development of farmer groups [86]. Conversely, formal education does not significantly correlate with the level of member participation in the development of farmer groups [87]. Regarding the shallot commodity, research revealed that the duration of farming and the farmer’s age were the key characteristics that substantially impacted farmers’ adoption of environmentally friendly shallot cultivation methods. However, the education degree did not have a noteworthy effect [88]. Further, Suroso et al. [89], in their study, revealed that education substantially impacts community engagement. Conversely, educational media is crucial to cultivating media literacy [90].

3. Research Methods

3.1. Construction of Conceptual Framework

In this subsection, we created a SEM framework with the intention of making the execution process a great deal less difficult. The conceptual framework is a model that acts within the area of research. According to Adom et al. [91], a conceptual framework is defined as a framework that is based on an existing theory in a field of inquiry that is related to and/or reflects the hypothesis of a study. Its purpose is to outline concepts, variables, and the interrelationships that exist between them [92]. Alternatively, a conceptual framework can be a structure that the researcher believes can best explain the natural progression of the phenomenon that is going to be studied [93]. Additionally, it is a metacognitive, reflective, and operational component of the entire research process [94]. It is the logical conceptualization of a research project. Next, Yuniarsih et al. [92] explained that its primary function is to direct the design of research, to make it easier to understand phenomena, to provide assistance in making decisions, and to effectively convey the findings of research to other people within the community. Therefore, referring to the literature in the previous section, in this study, there were eight latent variables (LVs) in total, including of six exogenous LVs, namely, the LV of Physical Aspects of Land (X₁), the LV of System of Economy Peasant Society (X₂), the LV of System of Political Peasant Society (X₃), Communication System of Farming Society (X₄), Socio-Cultural System of Peasant Society (X₅), and the LV of Education Level (X₆), and two dependent LVs (Farmer Participation (Y₁) and LV of Shallot-Farming Development (Y₂)). The structural equation model, as the conceptual framework of the research, examined in this study and the connections between underlying variables are illustrated in Figure 1. The corresponding symbols utilized in the model are outlined in

Table 1. The description of the variables and the measurement units of this research.

Latent Variables	Observed Variables
	Symbols	Indicator Variable Names	Measurement Unit *
Physical Aspects of Land (X1)	X1.1	Land Suitability	5-Point Likert scale
	X1.2	Topography	5-Point Likert scale
	X1.3	Accessibility	5-Point Likert scale
	X1.4	Climate Suitability	5-Point Likert scale
System of Economy Peasant Society (X2)	X2.1	Production Costs	5-Point Likert scale
	X2.2	Marketing Costs	5-Point Likert scale
	X2.3	Availability of Venture Capital	5-Point Likert scale
	X2.4	Labor Availability	5-Point Likert scale
System of Political Peasant Society (X3)	X3.1	The Role of Community Leaders	5-Point Likert scale
	X3.2	Community Engagement	5-Point Likert scale
	X3.3	The Role of Government Officials	5-Point Likert scale
	X3.4	Pricing Policy	5-Point Likert scale
Communication System of Farming Society (X4)	X4.1	Farmer Group Meeting	5-Point Likert scale
	X4.2	Extension Visit	5-Point Likert scale
	X4.3	Farmer and PPL interaction	5-Point Likert scale
	X4.4	Availability of Communication Media	5-Point Likert scale
Socio-Cultural System of Peasant Society (X5)	X5.1	Tudang Sipulung	5-Point Likert scale
	X5.2	Mutual Cooperation	5-Point Likert scale
	X5.3	The Ijon System	5-Point Likert scale
	X5.4	Patron–client	5-Point Likert scale
Education Level (X6)	X6.1	Length of Education	5-Point Likert scale
	X6.2	Non-formal Education	5-Point Likert scale
	X6.3	Literacy Level of Social Media Use	5-Point Likert scale
	X6.4	Literacy on the Use of Agricultural Extension Media	5-Point Likert scale
Farmer Participation (Y1)	Y1.1	Participation in Planning	5-Point Likert scale
	Y1.2	Participation in Execution	5-Point Likert scale
	Y1.3	Participation in Monitoring	5-Point Likert scale
	Y1.4	Participation in Evaluation	5-Point Likert scale
Shallot-Farming Development (Y2)	Y2.1	Shallot Production Quality	5-Point Likert scale
	Y2.2	Increase in Shallot Production	5-Point Likert scale
	Y2.3	Shallot Productivity Increase	5-Point Likert scale
	Y2.4	Shallot Revenue Increase	5-Point Likert scale

* Description: Unit division standard for 5-point Likert scale: 1 = Strongly Disagree, 2 = Disagree, 3 = Neutral, 4 = Agree, 5 = Strongly Agree. For example, if a respondent answered “1” for the indicator variable of Land Suitability, it meant very low land suitability for shallot cultivation.

.

3.2. Research Site

The research site is located inside the administrative boundaries of Uluere District, Bantaeng Regency, South Sulawesi Province, Indonesia (Figure 2). The Uluere District comprises six settlements across 67.29 km², which accounts for 17% of the entire area of Bantaeng Regency. The horticulture cropland area is 4431 hectares, which includes shallots, chilies, potatoes, cabbage, petsai, and tomatoes [95]. This study was conducted according to the guidelines of the Declaration of Helsinki and was approved by the Institutional Review Board (or Ethics Committee) of the South Sulawesi Province Governor’s Research Licensing Division Committee (Dinas Penanaman Modal dan Pelayanan Terpadu Satu Pintu) on 25 November 2022, via Permit Letter No. 12534/S.01/PTSP/2022.

Uluere is a sub-district in Bantaeng Regency, South Sulawesi, Indonesia. The capital of Uluere District is located in Bonto Marannu Village, which Gowa and Jeneponto Regencies border to the north, Jeneponto Regency to the east, Bantaeng Regency to the south, and Sinoa District to the west. The location is approximately 120 km south of Makassar, the capital of South Sulawesi Province. Its precise coordinates are 5°21′13″–5°35′26″ south latitude and 119°51′42″–120°05′27″ east longitude.

3.3. Research Process and Design

The research process and design is a set of procedures and methods used to analyze the variables under study and collect data. We had a minimum of two alternatives at our disposal regarding the software utilized for the analysis of the study data. The available solutions included Smart-PLS Software (https://www.smartpls.com/) and AMOS Software (https://www.amossoftware.com/), with each offering distinct advantages. Academics have consistently preferred Smart-PLS over the years because of its exceptional predictive accuracy, especially when dealing with complex situations [96]. Smart-PLS offers greater flexibility in modeling compared with other methods [97]. Additionally, it uses a variance-based relationship approach instead of covariance [98], and it is reliable in handling sample distribution and small sample sizes [99]. Further, AMOS Software, a powerful structural equation modeling program, enhances regression, factor analysis, correlation, and analysis of variance when utilizing covariance [98]. It also employs ML estimation in SEM analysis, a commonly used method for confirming theories [100]. Due to the study’s objective, which was to validate and verify the relationship among the latent variables as presented in the conceptual framework of the research (Figure 1), we decided to use AMOS Software for the analysis of the primary data collected.

Moreover, from the literature review, six latent variables that affect the success of shallot farming were identified, namely, the latent variables Physical Aspects of Land (X₁), System of Economy Peasant Society (X₂), System of Political Peasant Society (X₃), Communication System of Farming Society (X₄), Socio-Cultural System of Peasant Society (X₅), Education Level (X₆), Farmer Participation (Y₁), and Shallot-Farming Development (Y₂). The research process and design are shown in Figure 3. Figure 3 shows four important steps in the research process. The four steps are (1) SEM development and data collection and verification, (2) data analysis, (3) model fit evaluation, and (4) SEM effect evaluation.

3.3.1. The First Step: SEM Development and Data Collection

This step began with identifying problems and formulating research problems. From the problem formulation, the research variables consisting of six exogenous latent variables (Physical Aspects of Land (X₁), System of Economy Peasant Society (X₂), System of Political Peasant Society (X₃), Communication System of Farming Society (X₄), Socio-Cultural System of Peasant Society (X₅), and Education Level (X₆)) and two dependent latent variables (Farmer Participation (Y₁) and Shallot-Farming Development (Y₂)) were identified. Then, we developed a theoretical model that reflects the relationships among these latent variables, whose direction of influence is a representation of the theory or literature in the previous literature review chapter (Figure 1). This study prioritized parsimony by utilizing concise theoretical models to facilitate easier interpretation of the results. We conducted primary data collection after completing the development of the structural equation model.

• Structural Equation Modeling (SEM) Development

The approach for analyzing the data was structural equation modeling (SEM). SEM is a valuable statistical modeling technique for examining data comprising latent and indicator variables [101]. Equation (1) represents the mathematical relationship among latent variables in SEM.

η = βη + Γξ + ς

(1)

In Equation (1), η, β, ξ, ς ∈ R, Γ ∈ R^n×n, and the relationship is represented by partial least squares as shown in Equation (2).

η_i = Σ_iβ_jiη_i + Σ_iδ_jlξ_l + ς_j

(2)

where β_JI dan δ_jl are coefficients relating the predicted endogenous variable to the exogenous variable, while ς_j is the residual of the endogenous variable.

$$\left(\begin{array}{c}n1\\ n2\end{array}\right)=\left(\begin{array}{cc}0& 0\\ \beta 21& 0\end{array}\right)\left(\begin{array}{c}n1\\ n2\end{array}\right)+\left(\begin{array}{c}\delta 11\\ \delta 21\end{array} \begin{array}{c}\delta 12\\ \delta 22\end{array} \begin{array}{c}\delta 13\\ \delta 23\end{array} \right)\left[\begin{array}{c}\xi 1\\ \xi 2\\ \xi 3\end{array}\right]+\left(\begin{array}{c}\varsigma 1\\ \varsigma 2\end{array}\right)$$

(3)

In this research, a structural equation model incorporating 32 indicator variables, 2 endogenous latent variables, and 6 exogenous latent variables was developed. Path diagrams facilitated the implementation of the measurement and structural models and the relationships among variables. Confirmatory Factor Analysis (CFA) was subsequently implemented as the measurement model. CFA represents a latent variable measurement paradigm in which one or more observable variables are utilized. In Figure 1, it can be seen that the exogenous latent variables of Physical Aspects of Land, System of Economy Peasant Society, System of Political Peasant Society, Communication System of Farming Society, Socio-Cultural System of Peasant Society, and Level of Education each formed an informative model with four indicators. Thus, the total indicator variables in the exogenous latent variables were 24 indicators. At the same time, the endogenous latent variables Farmer Participation and Shallot-Farming Development had four indicators each.

• Data Collection

The data gathering approach utilized in this study was a questionnaire. This research employed a closed questionnaire method, employing a Likert scale to assess the responses provided by participants. Furthermore, the population for this study comprised 1500 shallot farmers who were residents of five villages located within the Uluere District. These communities were selected as research locations based on their significant shallot production within Bantaeng Regency. Following that, the researchers determined the sample size for the study by using the Slovin formula [102], which led to the selection of 150 respondents who were engaged in shallot farming. The proportionate random sampling approach determined the sample size of 150 respondents for this research, which fell into the medium category for SEM analysis [103]. Verifying the data and variables was the next step after completing the primary data collection. This verification aimed to confirm the validity of all collected data and variables for the research model. Once we declared the data and variables valid, we could initiate the data analysis process.

3.3.2. The Second Step: Data Analysis

In this step, the data were prepared first. Once this was achieved, we moved on to determining the reliability and validity of the test. Cronbach’s alpha was the method utilized in the reliability test. All variables were proven to be legitimate. The SEM path model could be estimated, allowing valid and dependable data to continue. The structural equation modeling (SEM) technique was utilized to analyze the collected primary data. Below are some details on our tasks before testing our hypothesis by using the structural equation modeling approach.

• Validity Testing of Questionnaire

Validity refers to the extent to which the data collected during a research study accurately reflect the information researchers can provide [104]. The validity test employed in this work utilized the Pearson product–moment correlation coefficient analysis technique, employing the calculation specified in Equation (4).

$$r_{\mathit{xy}}={\displaystyle \frac{ \sum nXY-\left(\sum X\right)\left(\sum Y\right)}{\sqrt{\left(n\sum X^2-{\left(\sum X\right)}^2\right)\left(n\sum Y^2-{\left(\sum Y\right)}^2\right)}}}$$

(4)

where r_xy = the Pearson correlation coefficient between variable and instrument item; X = the score of instrument items to be used; Y = the score of all instrument items in the variable; n = the number of respondents; ∑ X² = the sum of squares of X values; and ∑ Y² = the sum of squares of Y values. According to the decision-making process in Equation (4), if the calculation result of r_xy or r count is greater than t_0.05, then the instrument is reliable, while if r-cout is smaller than t_0.05, the instrument is not reliable.

• Reliability Testing

Reliability testing is carried out to prove that the instrument used can measure something consistently over time. The instrument tested using the reliability test was a questionnaire. The research instrument utilized in this study comprised a questionnaire and a multilevel scale. Consequently, Cronbach’s alpha formula, which is represented in Equation (5), was employed to assess the instrument’s reliability. The value of the reliability coefficient serves as an empirical indicator of high and low reliability. An r_xx value denotes elevated reliability in proximity to 1. The satisfactory reliability level is generally acknowledged as an alpha value of 0.700 or greater. Then, if the alpha value is greater than 0.80, all items are deemed reliable, and the reliability of the entire test is consistently high.

$$r_{\mathit{xx}}=\left({\displaystyle \frac{n}{n-1}}\right) \left(1-{\displaystyle \frac{\sum \delta t^2}{\delta t^2}}\right)$$

(5)

where r_xx = the reliability sought; n = the number of question items tested; ∑ σ_t² = the total variance; and σt² = the total variance of each item’s score.

3.3.3. The Third Step: Measurement Model Evaluation

This step involved analyzing model fit indices and R-Square values. The model was initially evaluated by using a fit test that comprised Chi-square, Probability Level, RMSEA, GFI, AGFI, CMIN/DF, TLI, and CFI. The criteria for these tests are outlined in

Table 2. Goodness-of-fit index and model fit.

Goodness-of-Fit Index		Cut-Off Value
Chi-square	X2	The smaller, the better (p-value ≥ 0.05)
Probability Level	PL	≥0.05
Root Mean Square Error of Approximation	RMSEA	RMSEA ≤ 0.08 means a good fit
Goodness-of-Fit index	GFI	Good fit if GFI ≥ 0.9 and marginal fit if 0.8 ≤ IFI ≤ 0.9
Adjusted Goodness-of-Fit index	AGFI	The model is said to be a good fit if AGFI ≥ 0.9 and is said to be a marginal fit if 0.8 ≤ AGFI ≤ 0.9
CMIN/DF	CMIN/DF	≤2.02
Tucker Lewis Index	TLI	The model is said to be a good fit if it has a TLI value ≥ 0.9 and is said to be a marginal fit if 0.8 ≤ TLI ≤ 0.9
Comparative Fit Index	CFI	The model is said to be a good fit if it has a CFI value ≥ 0.9 and is said to be a marginal fit if 0.8 ≤ CFI ≤ 0.9
Model fit: R-Square	R2	R2 = 0.040 to 0.19 (weak), R2 = 0.24–0.33 (moderate), R2 = 0.34–0.67 (strong)

Source: [105,106].

. If a model is deemed unfit, it must be modified. Adjusting the indices is one method to enhance a poorly fitting model. The modification index quantifies the reduction in the Chi-square value while estimating coefficients in a structural equation model (SEM). Model modifications should be based on theoretical justifications and implemented on the initial structure of the structural equation model. Any modifications made to the model must be evaluated by using distinct data before their acceptance. Model measurement can be performed by modifying the index. Models deemed suitable can proceed to R-Square analysis. The R-Square or coefficient of determination quantifies the degree of fit between dependent and independent data. The R-Square value ranges from 0 to 1. A number closer to 1 indicates an improving link between endogenous and exogenous latent variables (Table 2).

The evaluation of the measurement model was the subsequent phase of this research design. During this phase, experiments were conducted to assess convergent validity (CV), composite reliability (CR), Cronbach’s alpha (CA), and average variance extraction (AVE). In the concurrent validity test, the following requirement must be met: the loading factor must be significant (loading factor > 0.50). When the indicators did not match these standards, they were omitted from the analysis. After deleting erroneous data, all analytical indicators were valid. Thus, CR and CA testing were possible. The required dependability was achieved, and CA was >0.60. Next, we ran the AVE test with a minimum cut-off of 0.50. The four tests in the measurement model evaluation had to be passed to enter the structural equation model evaluation step.

3.3.4. The Fourth Step: SEM Evaluation

In this step, verifying the requirements attached to the assumptions is necessary. We proceeded to the SEM effect evaluation after that was completed. Nevertheless, we supposed that not all of the conditions were satisfied. In that case, the model had to be modified and returned to the initial step, the data collection and verification step. First, the structural equation modeling (SEM) analysis, which included direct and indirect effect tests, was performed in this step. In direct effect testing, we first identified the hypotheses about the direct relationships among latent variables. This study formulated the hypotheses at a later stage. Following the formulation of the hypothesis, Figure 1 illustrates the relationships through a path diagram. AMOS Software used this path diagram to visualize the model for testing. It then calculated the path coefficient and provided an estimate and the significance level for the direct path coefficient. This coefficient indicated the magnitude of the direct effect of one variable on another. We will further interpret the value of this coefficient in the Discussion chapter of this research paper. After that, the test’s outcomes were assessed to determine an efficient management strategy for shallot-farming procedures. The following step was to arrive at conclusions and make recommendations for policy, which will be briefly discussed in the concluding section of this research paper.

Furthermore, structural equation model evaluation was carried out. This phase began with testing the Goodness-of-Fit (GoF) of the model. In testing the GoF of the model, the cut-off value requirements had to be met: CMIN/DF value < 2.00, RMSEA ≤ 0.08, TLI, and CFI ≥ 0.90. If a model is declared unfit, the covariance index must first be modified. This modification requires the modification index (M.I) value to indicate a misfit model. Index modification is performed by covariance of the M.I value generated. In addition, it should be noted that the covaried M.I value must be based on supporting theory. Then, if the model has been declared fit, one may proceed with the R-Square analysis. This analysis is performed by examining the squares’ multiple correlation values. R-Square ranges from 0.040 to 0.19 (weak), 0.24–0.33 (moderate), and 0.34–0.67 (strong). An R-Square value approaching 1 is indicative of a stronger relationship between endogenous and exogenous latent variables.

3.4. Hypothesis Model Test

The final step in this research design was model hypothesis testing, interpretation, and drawing conclusions and recommendations. Model hypothesis testing can be performed after all SEM requirements are met. However, if all conditions are not met, then the model that has been formed needs to be improved and return to the model development step. In this study, a direct effect test was conducted. The direct effect is known through the C.R number generated by AMOS Software > 1.96 or p-value < 0.05. Then, based on the theoretical framework model presented in Figure 1, 13 hypotheses (H) were proposed in this study:

• H1 = The LV of Physical Aspects of Land (X₁) influences the LV of Farmer Participation (Y₁).
• H2 = The LV of System of Economy Peasant Society (X₂) influences the LV of Farmer Participation (Y₁).
• H3 = The LV of System of Political Peasant Society (X₃) influences the LV of Farmer Participation (Y₁).
• H4 = The LV of Communication System of Farming Society (X₄) influences the LV of Farmer Participation (Y₁).
• H5 = The LV of Socio-Cultural System of Peasant Society (X₅) influences the LV of Farmer Participation (Y₁).
• H6 = The LV of Education Level (X₆) influences the LV of Farmer Participation (Y₁).
• H7 = The LV of Physical Aspects of Land (X₁) influences the LV of Shallot-Farming Development (Y₂).
• H8 = The LV of System of Economy Peasant Society (X₂) influences the LV of Shallot-Farming Development (Y₂).
• H9 = The LV of System of Political Peasant Society (X₃) influences the LV of Shallot-Farming Development (Y₂).
• H10 = The LV of Communication System of Farming Society (X₄) influences the LV of Shallot-Farming Development (Y₂).
• H11 = The LV of Socio-Cultural System of Peasant Society (X₅) influences the LV of Shallot-Farming Development (Y₂).
• H12 = The LV of Education Level (X₆) influences the LV of Shallot-Farming Development (Y₂).
• H13 = The LV of Farmer Participation (Y₁) influences the LV of Shallot-Farming Development (Y₂).

4. Results and Discussion

4.1. Measurement Model Evaluation Results

This research model had six latent variables (LVs), including Physical Aspects of Land (X₁), System of Economy Peasant Society (X₂), System of Political Peasant Society (X₃), Communication System of Farming Society (X₄), Socio-Cultural System of Peasant Society (X₅), Education Level (X₆), Farmer Participation (Y₁), and Shallot-Farming Development (Y₂). Then, the evaluation of the measurement model in this study tested the validity and reliability of the latent variables.

4.1.1. Validity Test Results

The LV validity test evaluated the degree of alignment between the indicator variables and the LV theory. Indicator variables were considered valid if their loading factor exceeds 0.5.

Table 3. The reliability test results.

Variables	Code	Cronbach’s Alpha	Description
Physical Aspects of Land	X1	0.790	Reliable
System of Economy Peasant Society	X2	0.747	Reliable
System of Political Peasant Society	X3	0.825	Reliable
Communication System of Farming Society	X4	0.793	Reliable
Socio-Cultural System of Peasant Society	X5	0.782	Reliable
Level of Education	X6	0.787	Reliable
Farmer participation	Y1	0.798	Reliable
Shallot-Farming Development	Y2	0.821	Reliable

displays the outcomes of the validity test. Table 3 demonstrates that all indicators for each latent variable, namely, Physical Aspects of Land (X₁), System of Economy Peasant Society (X₂), System of Political Peasant Society (X₃), Communication System of Farming Society (X₄), Socio-Cultural System of Peasant Society (X₅), Education Level (X₆), Farmer Participation (Y₁), and Shallot-Farming Development (Y₂), had loading factor values greater than 0.5 and AVE values greater than 0.5. These results indicate that the items measuring the research variables met convergent validity, so all valid indicators were used for further analysis.

4.1.2. Reliability Test Results

The reliability of data criteria can be established when Cronbach’s alpha (α) value exceeds 0.6, as indicated by Amanda et al. [107]. The reliability test results are shown in Table 3. The reliability test findings indicate that all variables were deemed reliable since Cronbach’s alpha value was above 0.6, thus rendering them suitable for use in research.

4.1.3. Confirmatory Factor Analysis Test: Initial and Fit Models

Confirmatory Factor Analysis (CFA) is a research instrument quality test that validates latent construct indicator statement parts. Testing is incremental until it approaches the threshold-based model. Chi-square, significance (P), RMSEA, CFI, GFI, AGFI, TLI, and CMIN/DF are used for Goodness-of-Fit. A suitable model must be adjusted to meet model feasibility requirements if it is not found. CFA testing uses theoretically based structural equation models. Figure 4 shows CFA output based on research model Goodness-of-Fit indices.

Table 4. The results of validity and AVE tests.

Variables	Indicators	Code	Loading Factors	Criteria	AVE	Description
Physical Aspects of Land (X1)	Land Suitability	X1.1	0.712	0.5	0.634	Valid
	Topography	X1.2	0.816	0.5
	Accessibility	X1.3	0.915	0.5
	Climate Suitability	X1.4	0.725	0.5
System of Economy Peasant Society (X2)	Production Costs	X2.1	0.672	0.5	0.585	Valid
	Marketing Costs	X2.2	0.899	0.5
	Availability of Venture Capital	X2.3	0.890	0.5
	Labor Availability	X2.4	0.537	0.5
System of Political Peasant Society (X3)	The Role of Community Leaders	X3.1	0.861	0.5	0.657	Valid
	Community Engagement	X3.2	0.810	0.5
	The Role of Government Officials	X3.3	0.845	0.5
	Pricing Policy	X3.4	0.717	0.5
Communication System of Farming Society (X4)	Farmer Group Meeting	X4.1	0.917	0.5	0.647	Valid
	Extension Visit	X4.2	0.890	0.5
	Farmer and Extension Workers Interaction	X4.3	0.871	0.5
	Availability of Communication Media	X4.4	0.446	0.5
Socio-Cultural System of Peasant Society (X5)	Tudang Sipulung	X5.1	0.889	0.5	0.609	Valid
	Mutual Cooperation	X5.2	0.872	0.5
	The Sistem Ijon	X5.3	0.707	0.5
	Patron–client	X5.4	0.622	0.5
Education Level (X6)	Length of Education	X6.1	0.900	0.5	0.616	Valid
	Non-formal education	X6.2	0.757	0.5
	Literacy Level of Social Media Use	X6.3	0.790	0.5
	Literacy on the Use of Agricultural Extension Media	X6.4	0.674	0.5
Farmer participation (Y1)	Participation in Planning	Y1.1	0.866	0.5	0.629	Valid
	Participation in Execution	Y1.2	0.834	0.5
	Participation in Monitoring	Y1.3	0.823	0.5
	Participation in Evaluation	Y1.4	0.628	0.5
Shallot-Farming Development (Y2)	Shallot Production Quality	Y2.1	0.861	0.5	0.659	Valid
	Increase in Shallot Production	Y2.2	0.639	0.5
	Shallot Productivity Increase	Y2.3	0.856	0.5
	Shallot Revenue Increase	Y2.4	0.869	0.5

shows the initial Goodness-of-Fit index results. Table 4 shows that this study’s eight Goodness-of-Fit parameters were not fitted. Despite the GFI, AGFI, TLI, and CFI being reasonably good, the initial structural equation model did not match the data. So, the model had to be adjusted and modified.

As explained above, the Goodness-of-Fit test results did not fulfill the criteria in the initial step. Consequently, adjustments were made to the covariance index. After implementing modifications to the model, it was determined that the structural equation model (Figure 5) was considered appropriate. The findings of the Goodness-of-Fit criterion for the model created in this study are also presented in

Table 5. The results of initial Goodness-of-Fit (GoF) indices and the GoF model fit test after modification.

Goodness-of-Fit Index	Cut-Off Value	Initial Goodness-of-Fit Index		GoF Model Fit Test Results after Modification
		Results	Descriptions	Results	Descriptions
X2	The smaller, the better (p-value ≥ 0.05)	973.959	Not yet fit	658.876	Expectedly small
Probability Level	≥0.05	0.000	Not yet fit	0.000	Fairly good
RMSEA	RMSEA ≤ 0.08 means a good fit	0.091	Not yet fit	0.061	Good fit
GFI	A good fit if GFI ≥ 0.9 and a marginal fit if 0.8 ≤ IFI ≤ 0.9	0.728	Marginal fit	0.803	Marginal fit
AGFI	The model is said to be a good fit if AGFI ≥ 0.9 and is said to be a marginal fit if 0.8 ≤ AGFI ≤ 0.9	0.670	Marginal fit	0.755	Marginal fit
CMIN/DF	≤2.02	2.234	Not yet fit	1.554	Good fit
TLI	The model is said to be a good fit if it has a TLI value ≥ 0.9 and is said to be a marginal fit if 0.8 ≤ TLI ≤ 0.9	0.782	Marginal fit	0.902	Good fit
CFI	The model is said to be a good fit if it has a CFI value ≥ 0.9 and is said to be a marginal fit if 0.8 ≤ CFI ≤ 0.9	0.809	Marginal fit	0.916	Good fit

. Overall, the developed model was well calibrated based on the data employed in this study. The Chi-square value obtained was 658,876, whereas the associated probability value was 0.000. The obtained RMSEA index demonstrated a satisfactory fit, as indicated by a value of ≤0.08. The GFI and AGFI index analyses yielded results of 0.803 and 0.755, respectively. These numbers fall below the threshold of 0.90, classifying them into the marginal fit group. The TLI and CFI analysis results yielded values of 0.902 and 0.916, respectively. These values fall under the good fit group since they are above the threshold of 0.90. Hence, the SEM path diagram generated after the index alteration procedure was deemed appropriate and viable for subsequent investigation.

4.1.4. The Results of R-Square (R²) Analysis

R-Square analysis necessitates the utilization of the squared multiple correlations value. The R-Square value of the endogenous latent variable was derived by using AMOS Software. The predefined measurement parameters were 0.34–0.67 for strong, 0.20–0.33 for moderate, and 0–0.19 for weak. The R-Square analysis findings are displayed in

Table 6. The results of the R-Square (R²) analysis.

Latent Variable	Notation	R-Square Value	Description
Farmer Participation	Y1	0.675	Strong
Shallot-Farming Development	Y2	0.706	Strong

. The R-Square value of the LV of Farmer Participation (Y₁) was 0.675, as indicated in Table 6. This diagram illustrates that the exogenous latent variables of X₁, X₂, X₃, X₄, X₅, and X₆ significantly impacted the endogenous latent variable of Farmer Participation (Y₁), meeting the strong requirement. The R-Square value of 0.675 signifies that the six exogenous latent variables affecting the endogenous latent variable contributed to 67.5% of its effect. In contrast, the remaining 32.5% was attributed to other variables not considered in the study model. The R-Square value of the LV of Shallot-Farming Development (Y₂) was 0.706. This diagram illustrates a strong causal connection between the exogenous latent variables of X₁, X₂, X₃, X₄, X₅, and X₆ and the endogenous latent variable of Y₂. Moreover, these data also demonstrate that the exogenous latent variables of X₁, X₂, X₃, X₄, X₅, and X₆ can account for 70.6% of the impact on the endogenous latent variable of Shallot-Farming Development (Y₂₎. The remaining 29.4% was attributed to additional variables not considered in the research model. Thus, it can be concluded that the LVs of System of Economy Peasant Society, System of Political Peasant Society, Communication System of Farming Society, Socio-Cultural System of Peasant Society, and Education Level strongly influenced the LVs of Farmer Participation and Shallot-Farming Development.

4.1.5. Hypothesis Testing

Hypothesis testing is a statistical technique used to make judgments by analyzing data from controlled experiments and uncontrolled observations. This study conducted hypothesis testing to examine the impact of the independent variable on the dependent variable. The outcomes of hypothesis testing are shown in Figure 6. The output of the hypothesis testing results is presented in Figure 6 and

Table 7. The results of the hypothesis testing.

			Estimate	S.E.	C.R	p	Hypothesis
Y1	<---	X1	0.323	0.088	3.677	***	Accepted
Y1	<---	X2	0.771	0.196	3.933	***	Accepted
Y1	<---	X3	0.413	0.144	2.866	0.004	Accepted
Y1	<---	X4	0.455	0.293	1.552	0.121	Rejected
Y1	<---	X5	−0.509	0.271	−1.880	0.060	Rejected
Y1	<---	X6	0.151	0.119	1.270	0.204	Rejected
Y2	<---	X1	−0.032	0.045	−0.716	0.474	Rejected
Y2	<---	X2	0.010	0.073	0.131	0.896	Rejected
Y2	<---	X3	0.200	0.077	2.617	0.009	Accepted
Y2	<---	X4	0.120	0.133	0.903	0.366	Rejected
Y2	<---	X5	0.066	0.128	0.514	0.607	Rejected
Y2	<---	X6	−0.180	0.066	−2.742	0.006	Accepted
Y2	<---	Y1	0.466	0.078	5.941	***	Accepted

Description: *** = 0.000 or more decimal.

.

Based on Table 7, there are 13 important points that can be seen based on the hypothesis test results:

(1)

There was a significant influence of the LV of Physical Aspects of Land (X₁) on the LV of Farmer Participation (Y₁) with a C.R value of 3.677 and a probability of 0.000.

(2)

The LV of System of Economy Peasant Society (X₂) had a significant influence on the LV of Farmer Participation (Y₁) with a C.R value of 3.933 and a probability of 0.000.

(3)

The LV of System of Political Peasant Society (X₃) significantly influenced the LV of Farmer Participation (Y₁) with a C.R value of 2.866 and a probability of 0.004.

(4)

The LV of Communication System of Farming Society (X₄) had an insignificant effect on the LV of Farmer Participation (Y₁) because the C.R value was 1.552 and the probability was 0.121.

(5)

The LV of Socio-Cultural System of Peasant Society (X₅) did not significantly influence the LV of Farmer Participation (Y₁) because the C.R value was −1.880 and the probability was 0.060.

(6)

There was an insignificant effect of the LV of Education Level (X₆) on the LV of Farmer Participation (Y₁) because the C.R value was 1.270 and the probability was 0.204.

(7)

The LV of Physical Aspects of Land (X₁) had an insignificant effect on the LV of Shallot-Farming Development (Y₂) because the C.R value was −0.716 and the probability was 0.474.

(8)

The LV of System of Economy Peasant Society (X₂) did not significantly influence the LV of Shallot-Farming Development (Y₂) because the C.R value was 0.131 and the probability was 0.896.

(9)

The LV of System of Political Peasant Society (X₃) significantly influenced the LV of Shallot-Farming Development (Y₂) with a C.R value of 2.617 and a probability of 0.009.

(10)

The LV of Communication System of Farming Society (X₄) had an insignificant influence on the LV of Shallot-Farming Development (Y₂) because the C.R value was 0.903 and the probability was 0.366.

(11)

There was no significant influence of the LV of Socio-Cultural System of Peasant Society (X₅) on the LV of Shallot-Farming Development (Y₂) with a C.R value of 0.514 and a probability of 0.607.

(12)

There was a significant influence of the LV of Education Level (X₆) on the LV of Shallot-Farming Development (Y₂) with a C.R value of −2.742 and a probability of 0.006.

(13)

There was a significant influence of the LV of Farmer Participation (Y₁) on the LV of Shallot-Farming Development (Y₂) with a C.R value of 5.941 and a probability of 0.000.

4.2. Discussion

In this section, we are going to discuss the results of the structural equation modeling test presented in Table 7, which aimed to determine if the hypotheses in this study could be accepted or rejected and to demonstrate the magnitude of the variables’ influence.

4.2.1. The Effect of the LV of Physical Aspects of Land (X₁) and the LV of System of Economy Peasant Society (X₂) on the LV of Farmer Participation (Y₁)

The test results indicate a strong influence of the LV of Physical Aspects of Land (X₁) on the LV of Farmer Participation. The test of the influence between the LV (latent variable) of Physical Aspects of Land (X₁) on the LV of Farmer Participation (Y₁) resulted in a statistical value of C.R of 3.677 and a probability of 0.000. Then, the estimated coefficient was positive, at 0.323. This figure indicates that an improvement or increase in the LV of Physical Aspects of Land (X₁) by 1 standard deviation (SD) can increase the LV of Farmer Participation (Y₁) by 0.323 SD. The research result is in line with the research results by Krakauer and Temimi [108], who showed a significant relationship among climate, river flow, land, and water absorption. Further, land use change can increase surface water runoff and deplete groundwater [109]. This affects the physical aspects of land, so shallot farmers at the research location actively participate in farmer groups. They plan for shallot varieties that will be cultivated based on the physical condition of the land.

Furthermore, the testing of the second hypothesis revealed that the LV of System of Economy Peasant Society (X₂) significantly impacted the LV of Farmer Participation (Y₁), as indicated by a C.R value of 3.933 and a probability of 0.000. This value proves a significant influence of the economic system of farming communities on farmer participation. Further, the estimated coefficient was 0.771. This figure shows that an increase in X₂ by 1 standard deviation (SD) can increase the LV of Farmer Participation (Y₁) by 0.771 SD. Farmer community participation plays an essential role in agricultural economic activities. The findings of this study are further supported by the findings of Bagheri [110], who elucidates that the assessment of the significance of sustainable agriculture varies across farmers, primarily according to the socio-economic features of farmers and their economic conduct. Another research result by Sari et al. [111] indicates that analyzing socioeconomic factors affecting shallot growth has a probability value of less than 0.05.

4.2.2. The Influence of the LV of System of Political Peasant Society (X₃) and the LV of Communication System of Farming Society (X₄) on the LV of Farmer Participation (Y₁)

The third hypothesis test revealed a substantial impact of the LV (latent variable) of System of Political Peasant Society (X₃) on the LV of Farmer Participation (Y₁). This result is based on the obtained C.R value of 2.866 and probability of 0.004. In addition, the estimated parameter was a positive 0.413. This result shows that an increase in X₃ by 1 standard deviation (SD) can increase the LV of Farmer Participation (Y₁) by 0.413 SD. Activities in the political system at the research location include participating in village head election activities and cooperating with the local government. This result is in line with the statement by Mas’oed [112], who revealed that activities in the political system include taking part in campaigns, participating voluntarily in campaign activities, participating in political party campaigns or political meetings, calling for support, and voting for political parties or major candidates, voting in elections, monitoring the casting and counting of votes, and evaluating candidates.

Moreover, the testing of the fourth hypothesis yielded a C.R value of 1.552 and a probability of 0.121. This figure explains that the LV of Communication System of Farming Society (X₄) had an insignificant influence on the LV of Farmer Participation (Y₁). The agricultural community communication system with farmer participation had no real influence on implementing participatory communication. This result is influenced by the beneficiaries’ lack of experience and their failure to participate in viewing the communication. Moreover, the distance between participants is far, so communication does not improve. The findings of this study suggest the necessity to enhance the amalgamation of governmental political interests with the aspirations and requirements of society. Optimally, collaborative planning can be conducted between the government and the community [113].

4.2.3. The Influence of the LV of Socio-Cultural System of Peasant Society (X₅) and the LV of Education Level (X₆) on the LV of Farmer Participation (Y₁)

The testing of the fifth hypothesis yielded a C.R value of −1.880 and a probability of 0.060. The results suggest no significant influence of the LV (latent variable) of Socio-Cultural System of Peasant Society (X₅) on the LV of Farmer Participation (Y₁). The impact of the socio-cultural framework of agricultural communities on farmer participation is seen in the farmers’ capabilities following the establishment of farmer collectives. The results show that farmers’ abilities are still weak because they admit they do not feel the benefits of participating in farmer groups. This research also confirms that people’s social skills in relation to new cultures still support existing cultural values. In addition, along with the development of new cultures in society, society maintains existing cultural values passed down from generation to generation [114]. In addition, the tudang sipulung culture is seen as an effective intermediary between the local administration and the community [71].

The results of testing the sixth hypothesis, the influence of the LV of Education Level (X₆) on the LV of Farmer Participation (Y₁) show that the C.R value obtained is 1.270, and the probability is 0.204. This value proves that the LV of Education Level (X₆) had an insignificant influence on the LV of Farmer Participation (Y₁). Competency refers to achieving target educational levels. Farmers who join farmer groups are predominantly elementary school graduates, with 80 percent participating in behavioral participation. Furthermore, the farmers’ limited educational attainment adversely affects their ability to enhance their income through their labor. This finding is consistent with the research conducted by Sahara et al. [55], which demonstrates that the involvement of farmer group members in the advancement of horticultural farming is influenced by their perceptions of the farmer group’s role, access to information, and level of formal education. Furthermore, their participation in farmer groups is strongly and negatively impacted.

4.2.4. The Influence of the LV of Physical Aspects of Land (X₁) and the LV of System of Economy Peasant Society (X₂) on the LV of Shallot-Farming Development (Y₂)

The results of testing the seventh hypothesis show no significant influence between the LV (latent variable) of Physical Aspects of Land (X₁) and the LV of Shallot-Farming Development (Y₂). Testing the influence of X₁ on Y₂ produced a C.R value of −0.716 and a probability of 0.474. These values show that the physical aspects of land, including land suitability, topography, accessibility, and climate suitability, are related to soil fertility, which can influence the growth and production of shallot farming. The findings of this study are consistent with the research conducted by Habibi [115] and Shura et al. [116], who demonstrated that farmers’ practices in growing shallots (Allium ascalonicum L.) are influenced by geographical factors, including land conditions, topography, rainfall, and soil conditions. Additionally, non-physical or social factors such as land ownership, capital, labor, knowledge, skills, transportation, and sales also significantly shape these cultivation activities.

Moreover, testing the eighth hypothesis yielded an estimated parameter value of 0.131 and a probability of 0.896. The data indicate that the LV of System of Economy Peasant Society (X₂) did not have a noteworthy impact on the LV of Shallot-Farming Development (Y₂). In terms of the economic system, the economic condition of the farming community at the research location is still considered weak in its contribution to shallot farming. This result is demonstrated by farmers who borrow money at interest or through a debt bond system when there is an urgent need. Hence, it is important to take into account the advice of key stakeholders in the development process and actively involve them as participants in the effort to enhance the economic viability of farmers within the framework of agricultural sustainability [117].

4.2.5. The Influence of the LV of System of Political Peasant Society (X₃) and the LV of Communication System of Farming Society (X₄) on the LV Shallot-Farming Development (Y₂)

The results of testing the ninth hypothesis show that the LV (latent variable) of System of Political Peasant Society (X₃) significantly affected the LV of Shallot-Farming Development (Y₂). This influence can be seen in the C.R value obtained, 2.617, and the probability, 0.009. Then, the estimated coefficient obtained was positive, 0.200. This figure shows that improving/increasing the system of political peasant society by 1 SD can increase the development of shallot farming by 0.200 SD. In the political system, farming communities in the development of shallot farming view political parties as being in line with the community’s wishes because they have distinctive tribes and political parties. The feeling expressed by the farming community was sympathy because they have a positive attitude towards politics, intermediaries to support aid, and vice versa. This result aligns with the statement by Bagayoko et al. [118], who revealed that a resource management system based on values, norms, and community relationships reflects the power structure. Social, cultural, and economic structures have developed this political system.

Moreover, the results of testing the tenth hypothesis regarding the effect of the LV of Communication System of Farming Society (X₄) on the LV of Shallot-Farming Development (Y₂) resulted in a C.R value of 0.903 and a probability of 0.366. This value shows that the LV of Communication System of Farming Society did not significantly influence the LV of Shallot-Farming Development (Y₂). The communication system in the shallot-farming community is still low. The intensity of meetings/communication between extension workers and farmers via telephone, internet, and telecommunication networks is relatively difficult because not all people in the village fully understand the use of smartphones. In order to seize these opportunities as well as challenges, vegetable farming managers must be able to adjust and capture various opportunities through information systems about shallot farming from technological sources. This result is consistent with the results of research by Sirajuddin and Liskawati [119], who showed that smartphones have insufficient potential to be used in agricultural extension, so strategies are needed to increase the ease of use of smartphones for farmers with low education levels.

4.2.6. The Influence of the LV of Socio-Cultural System of Peasant Society (X₅), the LV of Education Level (X₆), and the LV of Farmer Participation (Y₁) on the LV of Shallot-Farming Development (Y₂)

The testing of the eleventh hypothesis yielded an estimated C.R parameter value of 0.514, with a corresponding probability of 0.607. These data demonstrate that the LV (latent variable) of Socio-Cultural System of Peasant Society (X₅) did not have a noteworthy impact on the LV of Shallot-Farming Development (Y₂). Within the socio-cultural framework of farming communities, the focus is on the management and structure of agribusiness enterprises in the major areas of shallot-farming development. The existing cooperative relations in this area are still considered weak. A study conducted by Elizabeth [120] found that the use of the modernization paradigm in agricultural growth led to alterations in the social structure of rural agricultural communities. The changes that occur include the structure of agricultural land ownership, patterns of employment relations, and the structure of employment opportunities, as well as the income structure of farmers in rural areas. An adverse consequence of the wage system is the erosion of long-standing notions of unity and communal practices, leading farmers to forsake their traditions of mutual collaboration [121]. Additionally, the twelfth hypothesis testing yielded a C.R estimated parameter value of −2.742 with a probability of 0.006. The depicted data demonstrate that the LV of Education Level had a notable impact on the LV of Development of Shallot Farming. The obtained estimated coefficient was −0.180, indicating a negative value. Nevertheless, the findings of this investigation are consistent with the research outcomes obtained by Saputra et al. [122] and Putri et al. [123]. On the contrary, Anwarudin [124], Liani et al. [125], and Anwarudin and Haryanto [126] revealed that the majority of farmers had higher education above primary school level. Moreover, the success of shallot farming can be facilitated by working-age farmers who possess a sufficiently elevated degree of education, such as having completed high school [127,128].

Finally, the testing of the thirteenth hypothesis revealed that the LV of Farmer Participation (Y₁) substantially impacted the LV of Shallot-Farming Development (Y₂). Testing the influence of the LV of Farmer Participation (Y₁) on the LV of Shallot-Farming Development (Y₂) produced a C.R value of 5.941 and a probability of 0.000. Further, the estimated coefficient was positive, 0.466. This figure shows that increasing farmer participation by 1 standard deviation can increase the development of shallot farming by 0.466 SD. Farmers’ participation increases when they can build on existing strengths within the group to mobilize and motivate its members to achieve group goals so that farmer groups develop more dynamically. The role of farmers in developing shallot farming was very participatory at the research location. This finding is consistent with the findings of Berun et al. [129], who revealed that farmer groups played quite a role in Sumlili Village, West Kupang District, and Kupang Regency. Furthermore, Gao et al. [130] showed that farmers’ participation in contract farming can reduce transaction costs and increase agricultural productivity.

5. Conclusions and Recommendations

5.1. Research Conclusions

The objective of this research was to examine the determinants of farmer participation and shallot-farming development in search of effective farm management practices. This research was conducted in Bantaeng Regency, South Sulawesi Province, Indonesia, by utilizing the quantitative approach of structural equation modeling (SEM). A group of 150 farmers was randomly chosen to participate in this research via direct structural interviews. The respondents in the survey were farmers who were associated with shallot farmer organizations. The SEM results suggest that the selected indicators effectively measured all latent variables, and the conclusions of this research are as follows:

(1)

It was found that the physical aspects of the land, the system of economy peasant society, and the system of political peasant society were fundamental elements that exerted a positive and significant influence on farmer participation.

(2)

The factors of communication system, socio-cultural system, and education level within the agricultural community did not significantly influence farmer participation.

(3)

The system of political peasant society exerted a beneficial and noteworthy impact on shallot-farming development. The education level also had a significant role, albeit with a detrimental impact on shallot-farming development.

(4)

The physical aspects of the land, the economic system, the communication system, and the socio-cultural system of peasant society did not play a significant role in the development of shallot farming.

5.2. Relevant Recommendations

The results suggest that improvements in the physical aspects of the land, the economic framework, and the political structure of agricultural communities could promote farmer participation. Furthermore, enhancing the political structure of agricultural communities could enhance the success of shallot-farming development. Hence, community leaders and government officials might enhance their contributions to promote the development of shallot farming. Moreover, based on the research results, it is suggested that increasing farmer participation can help develop shallot farming. Participation in the planning phase and participation in the execution phase are the two metrics with the strongest association. These findings highlight the critical importance of thorough planning and good execution in farmer agricultural operations to ensure the sustained success of shallot farming. As a result, farmers and local governments should make this issue a top priority. Furthermore, farmer participation and the system of political peasant society significantly and positively impact the growth and development of shallot farming. Thus, by increasing the influence of government officials and community leaders, shallot farming can be promoted. Farmers can then enhance their participation in the plan’s formulation and implementation, assuring the continuing growth of shallot farming. The findings of this study contribute significantly to the body of knowledge by validating previous research and proposing different ways for the advancement of shallot farming.

Limitation of the Study

It was discovered that this study had limitations. The study’s sample size of 150 respondents is a medium category in structural equation modeling (SEM) analysis. If the study had a larger sample size, the conclusions may have differed slightly. The limited scope and sample size were due to the research study’s self-funded nature and restrictive budget, which prevented the inclusion of a large sample size. Another limitation of the study was that it missed a very important variable, “respondent experiences”, which can influence farmer participation and shallot-farming development. Therefore, we recommend conducting further research in the future, focusing on obtaining a large sample size that fully reflects the entire population and incorporating “respondent experiences” as a significant variable.