Employing Binary Logistic Regression in Modeling the Effectiveness of Agricultural Extension in Clove Farming: Facts and Findings from Sidrap Regency, Indonesia

Abstract

The research objective was to examine the factors influencing the effectiveness of agricultural extension in clove farming in Sidrap Regency. Primary data were collected using structured interviews with 140 selected clove farmers to achieve this objective. The binary logistic regression model was employed to test the influence of fourteen independent variables on the dependent variable. The research results indicated that the variables of family size, mass communication approach, electronic media, and clove cultivation material significantly and positively impacted the effectiveness of agricultural extension in clove farming. Meanwhile, the variables of educational attainment, farming experiences, farmers’ cosmopolitanism, print media, and land area had a negative significant effect, and the other variables had no significant effect. These findings are important indicators and valuable insights in promoting and encouraging the effectiveness of extension in clove farming. The findings emphasize the significant role of family characteristics, agricultural extension approaches, and communication delivery methods in the diffusion of agricultural technology and enhancing the effectiveness of agricultural extension. Based on the research results, policy recommendations are proposed to enhance extension effectiveness in clove farming, specifically, augmented farmer education on clove cultivation and the intensification and extensification of mass extension methods, as well as using electronic media in extension activities.

1. Introduction

Along with Madagascar and Tanzania, Indonesia is listed as one of the countries that produce the most clove (Syzygium aromaticum (L.) Merr. & Perry) in the world [1,2,3]. According to Siringoringo et al. [1], although its production is low, it has the largest harvest area. Kabote [4] and Kusuma et al. [5] said that clove is one of the spices with a high economic value. Additionally, according to Cahyani et al. [6] and Lestari et al. [7], clove is one of Indonesia’s most important non-oil and gas commodities. This crop increases the total amount of money available to the government [8,9]. As Mahulette et al. [10] and Riptanti et al. [11] pointed out, clove is a tremendously important crop in Indonesia. This is primarily because they contribute to the country’s income and foreign exchange through the excise tax levied on cigarettes. Cloves are also very important to Indonesian farmers because most cloves (98%) are cultivated by smallholders [3,12]. The main yield of clove plants is flowers harvested while still in bud [3,13]. In addition, cloves also have considerable health benefits, and they are used by the pharmaceutical, cosmetic, and other industries [1]. In fact, in Indonesia, cloves have been used for generations as traditional medicine [14]. Cloves also contain an important chemical compound called eugenol. The content of eugenol in clove oil is 75–90% [2,15]. In addition to its prospective applications as an analgesic, local anesthetic, and antimicrobial agent, this chemical is exploited as an antibacterial, antifungal, insecticide, and antioxidant [14]. Additionally, it is used to treat pain. Moreover, it is acknowledged as a material that can be used as a starting point for the creation of synthetic vanillin [2].

Clove cultivation is now widespread throughout Indonesia. However, based on the currently available statistical data, it is known that ten provinces in Indonesia are the largest centers of clove production. This information can be seen in Figure 1, based on the average clove production data for 2017–2021. The clove center provinces are Maluku, South Sulawesi, Central Sulawesi, Southeast Sulawesi, East Java, West Java, Central Java, North Sulawesi, Aceh, and North Maluku. The ten provinces contributed a cumulative 83.11% to clove production in Indonesia during this period. The main center of cloves is Maluku Province, which has an average production of 20.73 thousand tons or contributes 15.50% to clove production in Indonesia annually. The second rank is occupied by South Sulawesi, which has an average production of 19.73 thousand tons or contributes 14.76% per year. The average production of cloves in Central Sulawesi, Southeast Sulawesi, and East Java was 15.82 thousand tons, 13.48 thousand tons, and 10.76 thousand tons, respectively. Meanwhile, the next five provinces have an average production below 10 thousand tons. The 10 provinces with the highest clove production centers in Indonesia and their contribution percentages are presented in detail in Figure 1 [16].

From Figure 1, it can be seen that South Sulawesi Province is the second largest producer of cloves after Maluku. Even though it is in second place, this is an important indicator of the fact that the clove commodity in South Sulawesi Province has considerable potential to support the plantation subsector development program. Based on data on the area of clove plantations in 2021, South Sulawesi Province has a land area of 67,337.00 hectares with a production of 21,431.00 tons [17]. Furthermore, Sidrap Regency, located in South Sulawesi Province, is listed as one of the locations that cultivates cloves. Figure 2 illustrates the harvest area, production, and productivity of cloves in Sidrap Regency for 2017–2021. As shown in Figure 2, the harvest area of clove farming continuously increased from 2018 until 2021. This is based on the data that were supplied. After reaching 1545 hectares in 2018, the harvest area in Sidrap Regency was constantly expanded to 1849.61 ha in 2019, 1860.66 ha in 2020, and 2202.00 ha in 2021. Meanwhile, from 2017 to 2021, there were different fluctuation levels in this commodity’s production. These situations, the rise in harvest area, and the fluctuation in productivity of this commodity are interesting facts that were explored further.

Theoretically, the achievement of increased clove production in Sidrap Regency cannot be separated from the role of agricultural extension workers, who are the source of information for farmers in extension activities, which include maintenance, fertilization, and harvesting [18,19,20]. Extension agents are responsible for providing farmers with the information and education services they demand [21,22]. Extension agents are responsible for agricultural extension in a technical and administrative sense [23,24,25,26,27]. The ultimate objective of agricultural extension is to enhance land productivity, input allocation efficiency, effectiveness, and farmers’ income and welfare. Agricultural extension is also carried out as a response to the issues that farmers have in farming that are adapted to the advancement of science and technology [28,29,30,31,32,33,34]. This is done to address the challenges that farmers face. Addressing the problems that farmers face is done to provide solutions. Moreover, agricultural extension workers are required, according to Khairunnisa et al. [35], Darmawan [36], and Suratini et al. [37], to develop their capacity to make use of high-tech agricultural equipment to increase their production yields to assist farmers in their farming. Extension workers need to increase their capabilities and quality of their skills in many areas of education and training to increase the level of competence and motivation of their performance [38,39,40]. Extension workers must also adopt extension strategies adapted to meet farmers’ needs [41,42,43]. Extension workers put in greater effort to develop methods, media, and materials and increase their knowledge and abilities in diverse sectors [27,44,45,46,47,48] in order to demonstrate that the extension program is effective, increase knowledge and skills, and implement a change in attitude in the proper management of their farms [25,44,49,50,51]. Therefore, it is expected that an agricultural extension worker will be able to create a strong work plan and provide counseling to farmers depending on the requirements of their targets [52]. As a result, there must be genuine cooperation and common goals between farmers and the government to overcome all the challenges that farmers face [53]. To accomplish the goals of the extension program, it is necessary to have a high level of extension potential and efficacy [23,54,55].

The conviction and impression were conveyed by the explanation and description that came before it, which was that agricultural extension considerably increases agricultural production and productivity. As a result, this research aimed to examine the factors influencing the effectiveness of agricultural extension in clove farming by using binary logistic regression (BLR). Within the scope of this research, the term “effectiveness” refers to the degree to which the agricultural extension program is successful in educating clove growers. Furthermore, the utilization of BLR was regarded as one of the novelties that were discovered during the course of this research. For the purpose of the efforts that are currently being made to increase the production and productivity of clove cultivation, it had been hoped that the findings would help develop an important policy underpinning such activities.

2. Literature Review

Agricultural extension is a key instrument in the development of agriculture. The Indonesian government has implemented various strategies, including at the individual, group, and mass levels, to improve the effectiveness of agricultural extension. The same is true for the information sources used as extension media. Farmers no longer rely solely on print media for information in the digital age. However, social media and the internet have grown in prominence, and they are now the principal extension media used by farmers. The literature review revealed five key aspects hypothesized to influence the effectiveness of agricultural extension: farmer characteristics, agricultural extension approaches, farmer information sources, farm characteristics, and capital. We elaborate further on each of these aspects below.

2.1. Farmer Characteristics

2.1.1. Effect of Farmer Age, Educational Attainment, and Family Size on Agricultural Extension

The success of agricultural extension in improving the productivity and welfare of farmers cannot be separated from the various characteristics they have. Generally, a person is most productive between the ages of 15 and 64. A farmer maximizes his work productivity by carrying out his agricultural tasks at an optimal level in that age range [56]. Sumekar et al. [57] and Ullah et al. [58] added that people have the ability to absorb information only at a productive age. Farmers of a productive age are faster at accepting and understanding training material than farmers of an unproductive age. In addition, a farmer’s formal education also affects learning activities to increase knowledge and skills and change one’s attitude [59]. In addition, elderly farmers may be more open to implementing sustainable agricultural methods since they have amassed more expertise and knowledge over time [60]. Furthermore, education can change a person’s thinking mindset to reason about knowledge, thus influencing decision making, problem solving, the implementation of their actions [51,61], the ability to absorb, interpret, and apply new information, and the adoption of better farming techniques [62]. Next, the findings of the studies by Windani et al. [63] and Mizab [61] demonstrated that individuals who have completed higher levels of education are more open to receiving knowledge, information, and ideas from others. The conclusions of the research led to the discovery of this piece of information. Previous studies have shown that the degree of education farmers have is crucial in deciding the implementation of a new invention. A study by Azizah and Sugiarti [64] discovered that farmers who had completed higher levels of education had a more comprehensive understanding, which made it simpler for them to embrace new technology. Education has a significant impact, both directly and indirectly, on the effectiveness of agricultural extension workers, as stated by Arifianto et al. [64]. Education is another factor that plays a significant role in determining the effectiveness of agricultural extension workers. Furthermore, it not only stimulates the engagement of participants in training activities but also plays a vital role in determining the level of success that a training program achieves in knowledge transfer [65].

Moreover, the number of family dependents in a household is highly tied to the amount of needs a family has, in terms of their bodily and spiritual wants. According to Jamil et al. [59], Aniagyei et al. [51], and Ullah et al. [58], a person who has high family needs may also be motivated to use all of his knowledge and talents to perform work outside of his normal employment with the hopeful expectation of being rewarded with additional financial compensation. Salukh et al. [66] said that one of the primary reasons for household members to assist the head of the household in working for revenue is the number of family members who rely on the head of the household. This is one of the reasons that the head of the household is supported by other family members through their employment. Salukh et al. [66] also noted this. It was found that the bigger the number of family members that are reliant on a home, the greater the incentive to work more, which in turn helps them to work more efficiently [58]. In addition, Atube et al. [67] and Ndamani [68] stated that it is reasonable to anticipate that increasing the number of family members dependent on the family’s income will assist in better appreciating the acceptance of extension innovations. This is a belief that is grounded on possibility. Nabila et al. [69] said that farmers are driven to spend more on their day-to-day expenses, and many family members are reliant on them. This is because farmers are responsible for providing for their families. The family size variable has a significant and unfavorable impact on the degree to which people are willing to adopt innovations that extend their reach, according to Berhanu et al. [70].

2.1.2. Effect of Farming Experience and Farmers’ Cosmopolitanism on Agricultural Extension

According to Oktafiani et al. [71], farmers who have been working in farming for a long time tend to be more skillful at making decisions that follow their specialized knowledge and capabilities. This is a result of the fact that they have more relevant experience. Moreover, Malila et al. [72] observed that the length of time that farmers have spent working in the agricultural industry is directly proportional to the level of competence and skills they possess in terms of using their knowledge and talents in every farming activity. This statement is consistent with the findings of Kotur [73] and Darmawan [36], who explained in their research that the amount of work experience an individual has substantially influenced their performance.

Farmers’ cosmopolitan ability is the ability to orientate outside the area of a farmer to establish broad interpersonal relationships [74]. Farmers’ activities outside the village are a measure of cosmopolitan interactions [75,76]. In addition, cosmopolitanism is also measured by a farmer’s activities outside related institutions, such as the extension center, the agriculture office, the Agricultural Technology Assessment Center (BPTP), and universities, to seek information about supporting facilities for their farms [75]. The results showed that the cosmopolitanism of farmers was in the low category because farmers rarely sought information outside their village [77]. The distance between the village and the information center and the difficulty in accessing public transportation make farmers reluctant to seek information on their own [75]. Farmers prefer to utilize their time for gardening and wait for extension workers who come to visit their villages to receive information related to their farms [75].

2.2. Agricultural Extension Approaches: Effects of Individual, Group, and Mass Communication Approaches on Agricultural Extension

This individual communication approach is intended for farmers who receive special attention from field extension workers. The individual approach extension method is implemented to inform farmers by conducting individual communication [78]. Additionally, individual communication comprehensively addresses an issue with farmers and provides them with answers [79]. This communication is achieved based on the level of trust developed with the farmers so that they can discuss the issue, according to the findings of a study that concluded that these data reveal that individual approaches, such as extension methods with home visits, land visits, informal connections, and inquiry techniques, have various effects on farming operations. However, some research findings suggested that different approaches can affect farming operations. According to Ramadhana [78] and Fangohoi et al. [45], the group approach method is also utilized to target farmer groups in the context of agricultural practices. This is in addition to approaching things individually, which is typically observed. Group counseling is carried out routinely and scheduled following the program by extension workers and the agricultural office [80]. Saputra et al. [81] and Tumurang et al. [79] added that the group method (lecture) is the most economical method to convey information and knowledge verbally about the benefits and importance of innovation in farming. The research results show that the group method has a significant impact because farmers can apply an innovation and development program to support their farming activities [82]. The general populace, often known as the public, is the target audience for the aforementioned transmission of information. According to Ramadhana [78], agricultural extension workers can convey information to farmers frequently situated in rural towns and communities.

Furthermore, according to the findings of Tumurang et al. [79], Imran et al. [83], and Azumah et al. [84], mass communication is one of the tactics that can be utilized to effectively transmit information to a large number of targets in a short time. Extension workers can accomplish this. Furthermore, this is consistent with what was mentioned in the aforementioned statement. Tambo et al. [85] showed that in a preliminary attempt to manage fall armyworm, there was a substantial association between exposure to campaign channels (either independently or in combination) and better knowledge. This was the case regardless of whether the campaign channels were used individually or in combination.

2.3. Farmer Information Sources

2.3.1. Effects of Print, Electronic, and Social Media on Agricultural Extension

Print media is a means of mass media that is printed and published periodically, such as books, brochures, and leaflets, which are used by extension workers to convey information in a clear manner to farmers who need it [86,87]. The results showed that extension workers used print media to support consecutive extension activities, namely books/literature sources, brochures, leaflets, tabloids, and technical instructions [88,89]. In contrast to the findings of Jaya [90] and Munthali et al. [91], the multimedia website portal is one of the effective platforms for delivering information to the public, farmers, and other related parties. The results of their studies showed that the agricultural extension website portal provides significant benefits to the community and farmers, where people can easily access the latest information on agricultural techniques, crop management, fertilization, pest and disease control, and sustainable agricultural practices [92]. In addition, audiovisual media is also a series of electronic images equipped with audio sound elements and images expressed through video tapes. In agricultural extension materials, electronic media, such as the radio, LCD, slides, VCD, and television, is used by farmers to receive information [27,44,48,76]. Also, the material presented in the media contains material that farmers need to support better farming [76,81]. The results of these studies also showed that farmers’ attitudes toward the extension media are in the acceptance category. According to farmers, the extension workers delivered the correct form of media (audiovisual). In line with this opinion, Cummins et al. [93] stated that the videos used were designed to engage agricultural workers and raise farmers’ awareness of extension activities [94,95].

Furthermore, the utilization of social media in agricultural extension activities demonstrates that the utilization of social media in extension activities has been frequently carried out. It can favorably impact the extension’s precision, efficiency, and effectiveness to increase agricultural productivity [91,96]. This is demonstrated by the utilization of social media in extension activities. According to the views expressed by Safitri et al. [97], it is of the utmost importance for extension operations to utilize social media to disseminate extension materials, training, and various forms of socialization. This is deemed significant, to the extent that it may be used to locate and provide information related to agriculture. According to the studies carried out by Humaidi et al. [98] and Suratini et al. [37], the use of social media platforms like WhatsApp, Facebook, Instagram, and YouTube has an impact on the knowledge and behavior of farmers who are employed in agricultural extension. Following the results of a research project that was carried out by Humaidi et al. [98], it is feasible to utilize social media as a learning medium and a source of information concerning agriculture. This is because agricultural extension uses a wide variety of social media platforms. However, while social media can help agricultural extension, several obstacles prevent agricultural extension workers from using it as a tool. These include the age of farmers, inadequate internet networks, farmers’ lack of familiarity with technology, and the fact that some farmers do not yet own an Android device [99].

2.3.2. Effect of Clove Cultivation Material on Agricultural Extension

Extension materials are a collection of information and knowledge developed to assist farmers in improving their yields and farming efficiency [99,100]. It delivers information, training, and education to farmers on best practices in crop cultivation, pest and disease control, natural resource management, and agricultural technology innovation [22,34,101,102]. Agricultural extension materials aim to educate farmers to adopt modern sustainable farming techniques, increase their productivity and income, and promote economically, socially, and environmentally sustainable agricultural development.

Clove cultivation involves a series of human activities, which include land management, planting, maintenance, and harvesting [34,103,104]. Clove is an important plant that is cultivated in several patterns, including monoculture and polyculture. According to Le et al. [105], monoculture farming is simpler and yields a single plant species that grows as large as possible. This is accomplished without considering the possibility of crop failure due to attacks from pests and disorders. According to Lele et al. [106], polyculture leads to increased profits from a wider variety of crops, and even if one crop commodity is discounted, other commodities are still available. According to Nuthall [107], this is an essential requirement for the membership of the farming community inside the agricultural system. A crucial necessity is the need for high-quality, relevant, and practical agriculture information. Information about agriculture plays a key role in educating farmers, increasing their level of knowledge and ultimately supporting them in decision making that involves activities related to agriculture.

Some farmers stated that extension services helped them overcome frequent challenges that they had in farming, which is relevant to the topic at hand. According to Danjumah et al. [108], Sujianto et al. [34], and Gebremariam et al. [39], this statement indicates that extension services can respond to the practical issues that farmers experience and offer them answers that would help them improve their farming methods. The findings of the studies by Adamu et al. [109] and Baah [110], which suggested positive perceptions related to an increase in information credibility, indicate that extension services efficiently serve the information demands of farmers by ensuring factual accuracy, expert validation, and alignment with the existing knowledge. The studies revealed favorable perceptions exhibited by the participants in response to extension services that promoted information credibility. As stated by Nuthall [107], extension services are an essential resource for farmers since they enable them to contribute to developing their capabilities and assist in applying sustainable agricultural methods. The total beneficial impact of extension services exemplifies their success as a valued resource for farmers.

2.3.3. Farm Characteristics and Capital: Effect of Land Area and Capital on Agricultural Extension

The term “land used for crop cultivation” refers to a location utilized for agricultural crop production. The amount of cultivated land that farmers manage influences the amount of output and income that farmers obtain from the farms that they operate, as stated by Akbar et al. [111], Marding et al. [112], and Sujianto et al. [34]. This is because farmers are directly involved in land management. The findings of several previous studies [35,53,61,113] have proven that the variable of land area has been shown to significantly impact the total amount of agricultural production. Furthermore, this impact has been proven to be favorable. Thus, an increase in land area will be followed by increased clove production [112]. Furthermore, Alberth [49] and Sujianto et al. [34] showed that there was a significant relationship between extension participation among respondents and farm size. In contrast to this, Habun et al. [114] stated that the land area variable did not significantly affect clove production. This finding is due to, among others, the fact that the land used by farmers is intercropped; so, the land area has no significant effect on clove production [115].

Another factor that plays a role in producing more complicated farming methods is the availability of financial financing. The terms “fixed capital” and “current capital” are used collectively to refer to these two categories of capital. Capital that can be replenished in the short term is called current capital [116]. Examples of current capital include seeds, fertilizers, pharmaceuticals, labor, and other commodities. In contrast to current capital, which may be replenished immediately, fixed capital is secured by land, agricultural equipment, buildings, and other assets. Agricultural capital, which includes land area, labor, the application of fertilizers and pesticides, and farming instruments and machinery, is said to affect the effectiveness of agricultural extension, as stated by Jamil et al. [25]. According to Jamil et al. [25] and Sujianto et al. [34], the amount of money a farmer has is directly proportional to the amount of human resources.

3. Materials and Methods

3.1. Construction of Conceptual Framework

Research variables, including independent and dependent variables, are defined in behavioral studies to help conceptualize and organize research methods. These are the goals behind the creation of research variables. In the context of a specific research activity, a “variable” is any event, entity, or characteristic that changes in value and is intended to be researched, measured, reported, and assessed [117]. Kerlinger [118] defined variables as entities that can take on many values. We utilized dependent variables to measure attribute behavior in response to changes to one or more independent factors. To acquire a full grasp of the variables used and discussed in this study, we divided them into two categories: dependent variables and independent variables. The effect of the independent variables on the dependent variable’s behavior is interdependent. We selected to analyze the characteristics that reflect this study’s independent variables to understand better their link to the dependent variable, i.e., extension effectiveness. This approach was taken based on a previous study, which we refer to as the conceptual framework, as shown in Figure 3. We hypothesized that the 14 independent variables shown in Figure 3 are the most important predictors of the effectiveness of agricultural extension in clove farming.

3.2. Research Site, Data Collection Method, and Research Sample

This study was conducted in Pitu Riase District, Sidrap Regency, South Sulawesi Province in Indonesia (Figure 4). The regency was purposefully selected as the research site because it was more accessible than other regencies. In 2020, the Bureau of Statistics (BPS) estimated that 23,350 farmers in Pitu Riase District were cultivating clove crops. So, the minimum sample size of the research was 137 clove farmers (rounded up to 140) based on the Slovin formula. The villages of Compong, Leppangeng, Dengengdengeng, Buntu Buangin, and Belawae were the five villages in Pitu Riase District from which we collected primary data. Because they were the most critical clove-producing communities in the district, we decided to focus on these five villages, with the primary selection criteria comprising farmers who had grown clove crops during the year 2023. One hundred and forty clove farmers were selected randomly and interviewed directly for this study. The number of samples collected during the research was adequate to accurately depict the population of clove growers in Sidrap Regency. A pre-designed questionnaire was used to collect primary data. To obtain primary data, a structured interview was carried out, and a pre-tested questionnaire was conducted with 15 individual clove farmers. The purpose of the pre-tested questionnaires was to assess the questionnaire’s clarity, validity, reliability, and suitability to the respondents, obtain an opportunity to revise the questionnaire, and assess the time needed for the actual interview. Before conducting the interview, each respondent who participated in the research provided was informed about the research objectives, methods, benefits, and publication purposes. This study was conducted according to the guidelines of the Declaration of Helsinki and was approved by the Institutional Review Board (Local Government) of Pitu Riase District of the Sidenreng Rappang Regency Government via Permit Letter No. 800/136/Kec.Pitu Riase.

3.3. Binary Logistic Regression Analysis

3.3.1. General Model of Binary Logistic Regression

A regression model is a statistical method applied to determine independent variables’ influence on what is considered a dependent variable, as stated by Tampil et al. [119] and Getu [120]. This method is utilized to determine the relationship between the two variables. There are several various kinds of regression models. However, the linear regression model, which is represented by Equation (1), is the one that is considered to be the most fundamental of all of them.

Y = β₀ + β₁X₁ + β₂X₂ + … + β_nX_n

(1)

where:

Y = the dependent variable; X_1−n = the independent variables; β₀ = the constant; and β_1−n = the regression coefficient.

According to Tampil et al. [119], the binary logistic regression model analyzes the interaction between a single response variable and several predictor variables and examines possible connections between the variables. The response variable is provided as dichotomous qualitative data, where 1 indicates the presence of a characteristic and 0 indicates the absence of a personality trait. In other words, a value can be either positive or negative. That is to say, the response variable can either be present or absent. Furthermore, Tampil et al. [119] said that the binary logistic regression model is typically utilized when the response variable creates two categories with values of 0 and 1. Equation (2), a representation of the Bernoulli distribution, is adhered to by this model throughout its use.

f(y_i) = π_i^yi(1 − π_i)^1−yi

(2)

where π_i = the probability of the i-th event and y_i = the i-th random variable consisting of 0 and 1.

Then, a logistic regression model with one predictor variable is used, as shown in Equation (3) [119,120,121].

$$\pi (x)={\displaystyle \frac{\exp ({\beta }_0+{\beta }_{1}x)}{1+\exp ({\beta }_0+{\beta }_{1}x)}}$$

(3)

To facilitate the interpretation of the regression parameters, π(x) in Equation (3) is transformed, resulting in the logit form of the logistic regression [119], as presented in Equation (4).

$$\mathrm{g}(x)=\ln [{\displaystyle \frac{\pi (x)}{1+\pi (x)}}]={\beta }_0+{\beta }_{1x1}+{\beta }_{2x2}+\dots +{\beta }_{nxn}$$

(4)

3.3.2. Empirical Model

The empirical model was created using Equations (3) and (4) to examine the influence of 14 independent variables and their relationship to the dependent variable, i.e., extension effectiveness (EE), as demonstrated in Equation (5).

$$\begin{gathered} \mathrm{g}(\mathrm{E}\mathrm{E})={\displaystyle \frac{in\left(\mathsf{\pi }\left(\mathrm{E}\mathrm{E}\right)\right)}{1-\mathsf{\pi }\left(\mathrm{E}\mathrm{E}\right)}} \\ =\beta 0+\beta 1FA+\beta 2EA+\beta 3FS+\beta 4FE+\beta 5FC+\beta 6ICA+\beta 7GCA \\ +\beta 8MCA+\beta 9PM+\beta 10EM+\beta 11SM+\beta 12CCM+\beta 13LA \\ +\beta 14CP+{\epsilon }_i \end{gathered}$$

(5)

where:

g(EE) = effectiveness of agricultural extension in clove farming (1 = effective, 0 = otherwise); β₀ = constant; FA = farmer age (year); EA = educational attainment (years); FS = family size (people); FE = farming experience (year); FC = farmers’ cosmopolitanism (5-Point Likert Scale); ICA = individual communication approach (5-Point Likert Scale); GCA = group communication approach (5-Point Likert Scale); MCA = mass communication approach (5-Point Likert Scale); PM = print media (5-Point Likert Scale); EM = electronic media (5-Point Likert Scale); SM = social media (5-Point Likert Scale); CCM = clove cultivation material (5-Point Likert Scale); LA = land area (ha); CP = capital (IDR); and ε_i = error terms.

• Measurement Unit and Data Type

A variable is a property, attribute, or characteristic of a person, item, or scenario that exhibits the ability to vary or change, as stated by Marudhar [122]. Variables can be found in a variety of contexts, such as a person, an object, or a circumstance. A character, trait, or attribute can be defined as a variable depending on the context. The primary purpose of a study conducted in social science is frequently to understand the causal relationships between distinct social situations. Investigating the influence that one or more independent factors have on the variable that is the subject of the investigation is one of the actions that must be completed in this process. When discussing a cause-and-effect relationship, referring to the theorized result as the dependent variable is standard practice.

On the other hand, the independent variable is often referred to as the postulated cause throughout the context of this relationship. It is essential to understand that a variable does not always have the status of either an independent or a dependent variable as it depends on the context. Given this, it is possible for a variable regarded as independent in one investigation to be utilized as the dependent variable in another investigation. Salam et al. [123] considered including variables in the study hypotheses. They stated that three main techniques may be applied to deal with the phenomenon. The first thing that must be done to assess the impact of the independent variables on the dependent variable is to draw comparisons across groups based on the independent variables. In the second step of the process, one or more independent variables are linked to one or more dependent variables to establish a connection. The third step is to describe the responses to the independent variables, the variables that act as mediators, or the dependent variables.

As shown in

Table 1. The description of the variables, measurement units, and research data types.

No.	Variable Names	Symbols	Measurement Units *	Data Types
A	Dependent Variables
	The effectiveness of agricultural extension in clove farming	EE	1 = effective (for a farmer who has a high clove farming income, i.e., equal or higher than the average income of the farmers); 0 = otherwise	Categorical
B	Independent Variables
1.	Farmer Age	FA	year	Continuous
2.	Educational Attainment	EA	year	Continuous
3.	Family Size	FS	people	Continuous
4.	Farming Experience	FE	year	Continuous
5.	Farmers’ Cosmopolitanism	FC	5PLS	Categorical
6.	Individual Communication Approach	ICA	5PLS	Categorical
7.	Group Communication Approach	GCA	5PLS	Categorical
8.	Mass Communication Approach	MCA	5PLS	Categorical
9.	Print Media	PM	5PLS	Categorical
10.	Electronic Media	EM	5PLS	Categorical
11.	Social Media	SM	5PLS	Categorical
12.	Clove Cultivation Material	CCM	5PLS	Categorical
13.	Land Area	LA	ha	Continuous
14.	Capital	CP	IDR	Continuous

* = 5PLS (Five-Point Likert Scale: 1 = not effective, IDR = Indonesian Rupiah (Indonesian Currency).

, this study investigated the relationship between 14 independent variables and a dependent variable. A measurement unit is provided for each variable. We also effectively identified and classified the variables into two unique groups: continuous and categorical data types.

• Hypothesis Development

A temporary solution to a problem is one of the definitions of a hypothesis in the field of research [124]. The formulations used to express hypotheses vary in complexity. Most researchers engage in quantitative research activities to confirm their hypothesis rather than attempt to find a solution to the problem. It is, therefore, vital for a researcher to have a firm knowledge of the significance and nature of the hypothesis that was established at the beginning of a research activity. One of the goals of establishing a research hypothesis is to either draw and analyze the logical inferences of causal linkages or to foresee causal correlations between variables that have been observed [125]. Both of these objectives are important in the research process.

Using the abovementioned description of hypothesis development, we conducted a literature review to generate predicted hypotheses, hypothesis statements, and significant outcomes for each independent variable in this study. This study’s findings are presented in Table 2. The statistics shown in Table 2 make it abundantly clear that the results related to the independent variables were significant, as expected. The number of family members, the mass approach, electronic media, clove cultivation material, and capital are some aspects that might be considered. Education, time spent working in agriculture, farmers’ cosmopolitanism, print media, and the size of the land are some of the factors that negatively influence the outcome. As a result, it is possible to conclude that the elements of farmer age, individual approach, group approach, and social media do not significantly impact the effectiveness of agricultural extension concerning clove cultivation.

3.3.3. Parameter Estimations

The Maximum Likelihood Estimation (MLE) approach is a potential solution that could be applied when estimating unknown parameters. The most fundamental form of this strategy provides the anticipated value of β in order to maximize the likelihood function. Equation (6) represents the likelihood function for the binary logistic regression model [119,120,121,123,126]. For more information, see the references listed below. An organized approach can be used to demonstrate this.

$$l\left(\beta \right)=\prod _{i=1}^n\pi (x_i{)}^{yi}(1\pi (x_i){)}^{1-yi}$$

(6)

where y_i = the observation of the i-th variable; $\pi \left(x_i\right)$ = the odds ratio for the i-th predictor variable.

To facilitate the calculation, a log-likelihood approach is taken, as shown in Equation (7) [120,121,122,124,127].

$$l\left(\beta \right)=\sum _{i=1}^n\left\{y_1\ln (\pi \right(x_i\left)\right)+(1-y_i)\ln \left(1\pi \right(x_i\left)\right)\}$$

(7)

The value of the logistic regression coefficient (β^) is obtained by making the first derivative of l(β) against β and equating it to 0.

Table 2. Hypothesis development of each independent variable.

Independent Variables	Expected Signs *	Hypothesis Development	References
Farmer Age (FG)	+SIG	H0: Farmer age does not influence the effectiveness of agricultural extension in clove farming. H1: Farmer age influences the effectiveness of agricultural extension in clove farming.	[56,57]
Educational Attainment (EA)	+SIG	H0: Educational attainment does not influence the effectiveness of agricultural extension in clove farming. H1: Educational attainment influences the effectiveness of agricultural extension in clove farming.	[59,64,65,75,127]
Family Size (FS)	+/SIG	H0: Family size does not influence the effectiveness of agricultural extension in clove farming. H1: Family size influences the effectiveness of agricultural extension in clove farming.	[59,66,67,68,69,70]
Farming Experience (FE)	+/SIG	H0: Farming experience does not influence the effectiveness of agricultural extension in clove farming. H1: Farming experience influences the effectiveness of agricultural extension in clove farming.	[36,71,73]
Farmers’ Cosmopolitanism (FC)	+/SIG	H0: Farmers’ cosmopolitanism does not influence the effectiveness of agricultural extension in clove farming. H1: Farmers’ cosmopolitanism influences the agricultural extension effectiveness in clove farming.	[75,76,77]
Individual Communication Approach (ICA)	+/SIG	H0: The individual communication approach does not influence the effectiveness of agricultural extension in clove farming. H1: The individual communication approach influences the effectiveness of agricultural extension in clove farming.	[78,79]
Group Communication Approach (GCA)	+/SIG	H0: The group communication approach does not influence the effectiveness of agricultural extension in clove farming. H1: The group communication approach influences the effectiveness of agricultural extension in clove farming.	[45,78,79,80,81]
Mass Communication Approach (MCA)	+/SIG	H0: The mass communication approach does not influence the effectiveness of agricultural extension in clove farming. H1: The mass communication approach influences the effectiveness of agricultural extension in clove farming.	[78,79,83,85]
Print Media (PM)	+/SIG	H0: Print media does not influence the effectiveness of agricultural extension in clove farming. H1: Print media influences the effectiveness of agricultural extension in clove farming.	[86,87,89,90,92,128]
Electronic Media (EM)	+/SIG	H0: Electronic media does not influence the effectiveness of agricultural extension in clove farming. H1: Electronic media influences the effectiveness of agricultural extension in clove farming.	[27,44,48,76,81,93,94]
Social Media (SM)	+/SIG	H0: Social media does not influence the effectiveness of agricultural extension in clove farming. H1: Social media influences the effectiveness of agricultural extension in clove farming.	[37,91,96,97,98,129]
Clove Cultivation Material (CCM)	+/SIG	H0: Clove cultivation material does not influence the effectiveness of agricultural extension in clove farming. H1: Clove cultivation material influences the agricultural extension effectiveness in clove farming.	[22,100,103,107,108,110]
Land Area (LA)	+/SIG	H0: Land area does not influence the effectiveness of agricultural extension in clove farming. H1: Land area influences the effectiveness of agricultural extension in clove farming.	[35,111,112,114,115]
Capital (CP)	+SIG	H0: Capital does not influence the effectiveness of agricultural extension in clove farming. H1: Capital influences the effectiveness of agricultural extension in clove farming.	[25,116]

* SIG = Significant.

Table 2. Hypothesis development of each independent variable.

Independent Variables	Expected Signs *	Hypothesis Development	References
Farmer Age (FG)	+SIG	H0: Farmer age does not influence the effectiveness of agricultural extension in clove farming. H1: Farmer age influences the effectiveness of agricultural extension in clove farming.	[56,57]
Educational Attainment (EA)	+SIG	H0: Educational attainment does not influence the effectiveness of agricultural extension in clove farming. H1: Educational attainment influences the effectiveness of agricultural extension in clove farming.	[59,64,65,75,127]
Family Size (FS)	+/SIG	H0: Family size does not influence the effectiveness of agricultural extension in clove farming. H1: Family size influences the effectiveness of agricultural extension in clove farming.	[59,66,67,68,69,70]
Farming Experience (FE)	+/SIG	H0: Farming experience does not influence the effectiveness of agricultural extension in clove farming. H1: Farming experience influences the effectiveness of agricultural extension in clove farming.	[36,71,73]
Farmers’ Cosmopolitanism (FC)	+/SIG	H0: Farmers’ cosmopolitanism does not influence the effectiveness of agricultural extension in clove farming. H1: Farmers’ cosmopolitanism influences the agricultural extension effectiveness in clove farming.	[75,76,77]
Individual Communication Approach (ICA)	+/SIG	H0: The individual communication approach does not influence the effectiveness of agricultural extension in clove farming. H1: The individual communication approach influences the effectiveness of agricultural extension in clove farming.	[78,79]
Group Communication Approach (GCA)	+/SIG	H0: The group communication approach does not influence the effectiveness of agricultural extension in clove farming. H1: The group communication approach influences the effectiveness of agricultural extension in clove farming.	[45,78,79,80,81]
Mass Communication Approach (MCA)	+/SIG	H0: The mass communication approach does not influence the effectiveness of agricultural extension in clove farming. H1: The mass communication approach influences the effectiveness of agricultural extension in clove farming.	[78,79,83,85]
Print Media (PM)	+/SIG	H0: Print media does not influence the effectiveness of agricultural extension in clove farming. H1: Print media influences the effectiveness of agricultural extension in clove farming.	[86,87,89,90,92,128]
Electronic Media (EM)	+/SIG	H0: Electronic media does not influence the effectiveness of agricultural extension in clove farming. H1: Electronic media influences the effectiveness of agricultural extension in clove farming.	[27,44,48,76,81,93,94]
Social Media (SM)	+/SIG	H0: Social media does not influence the effectiveness of agricultural extension in clove farming. H1: Social media influences the effectiveness of agricultural extension in clove farming.	[37,91,96,97,98,129]
Clove Cultivation Material (CCM)	+/SIG	H0: Clove cultivation material does not influence the effectiveness of agricultural extension in clove farming. H1: Clove cultivation material influences the agricultural extension effectiveness in clove farming.	[22,100,103,107,108,110]
Land Area (LA)	+/SIG	H0: Land area does not influence the effectiveness of agricultural extension in clove farming. H1: Land area influences the effectiveness of agricultural extension in clove farming.	[35,111,112,114,115]
Capital (CP)	+SIG	H0: Capital does not influence the effectiveness of agricultural extension in clove farming. H1: Capital influences the effectiveness of agricultural extension in clove farming.	[25,116]

* SIG = Significant.

3.3.4. Nagelkerke R-Square Test

The Nagelkerke R-Square is a modified version of the R-Square developed by Cox and Snell. The values between 0 and 1 make up the range of possible values for Nagelkerke. This value represents the percentage of the total variation of the dependent variable that can be accounted for by the independent variables in the currently used model. In this study, the Nagelkerke R-Square was employed to test the contribution of the fourteen independent variables on the total variation of the dependent variable, i.e., the effectiveness of agricultural extension in clove farming.

3.3.5. Likelihood Ratio Test

The Likelihood Ratio Test is useful for determining whether each predictor influences a particular model. This test can be used to satisfy the requirements associated with the analysis in this study. The Likelihood Ratio Test is a statistical method that analyzes the ratio between the probability of observing data, assuming a particular parameter is zero (L₀), and the probability of obtaining data when the parameter is evaluated at its MLE (L₁). This ratio is referred to as the likelihood ratio. The research conducted by Tampil et al. [119] and Park [121] led to the following conclusion. To ascertain this ratio, it is necessary to initially assess the probability of detecting data while operating under the assumption that the parameter is negative. Equation (8) provides a mathematical explanation for the G Likelihood Ratio Test, as shown by the research conducted by Salam et al. [123], Yuniarsih et al. [126], Rusliyadi et al. [130], and Getu [120].

$$\mathrm{G}=-2\ln \left[{\displaystyle \frac{{\left(\frac{n_1}{n}\right)}^{n_1}{\left(\frac{n_0}{n}\right)}^{n_0}}{\prod _{i=1}^n{\widehat{\pi }}_i^{yi}{\left(1-{\widehat{\pi }}_i\right)}^{1-yi}}}\right]$$

(8)

where:

• n₁ = the number of observations in category 1;
• n₀ = the number of observations with category 0.

A Chi-square distribution was utilized to determine the test statistic G. The value of the χ² table was compared to the degree of freedom (df) = k − 1, where k is the number of predictor variables. This comparison was carried out to determine the degree of freedom in order to make the necessary decisions. The null hypothesis (H₀) was rejected when the value of G was more than χ² (db, α) or the p-value was lower than α.

3.3.6. Partial Hypothesis Test

Based on the model that was obtained, partial testing was utilized to investigate the impact of each β_i individually. There was a possibility that a predictor variable could be added to the model, and the results of the partial test showed whether or not this was doable. Concerning each variable, the following is the hypothesis that was utilized:

• H₀: β_i = 0
• H₁: β_i ≠ 0

The Wald test statistic (WST) can be seen in Equation (9). Equation (10) is a way to obtain the standard error estimator value of β_i [119,121,123,126,130].

$$\mathrm{W}\mathrm{S}\mathrm{T}={\displaystyle \frac{\widehat{\mathsf{\beta }}\mathrm{i}}{\mathrm{S}\mathrm{E}\left({\widehat{\mathsf{\beta }}}_{\mathrm{i}}\right)}}$$

(9)

and

$$\mathrm{SE}\left({\widehat{\mathsf{\beta }}}_{\mathrm{i}}\right)=\sqrt{\left({\mathsf{\sigma }}^2\left({\widehat{\mathsf{\beta }}}_{\mathrm{i}}\right)\right)}$$

(10)

where:

• SE (${\widehat{\mathsf{\beta }}}_{\mathrm{i}}$) = the estimated standard error for the coefficient of β_i;
• ${\widehat{\mathsf{\beta }}}_{\mathrm{i}}$ = the expected value for the parameter β_i.

Because the ratio generated by the Wald statistic agrees with the standard normal distribution, comparing it to the standard normal distribution (Z) is essential to arrive at a specific recommendation. If the W value exceeds Z α/2 or the p-value is lower than α, the null hypothesis (H₀) is rejected.

3.3.7. Interpretation of Binary Variable Coefficients

An odds ratio is a set of odds divided by other odds. The odds ratio coefficient shows a person’s tendency to do or not do an activity. The odds ratio equation (ψ) is presented in Equation (11) [30,120,123,126,130,131].

$$\mathsf{\psi }={\displaystyle \frac{\raisebox{1ex}{$\mathsf{\pi }\left(1\right)$}\!\left/ \!\raisebox{-1ex}{$\left[1-\mathsf{\pi }\left(1\right)\right]$}\right.}{\raisebox{1ex}{$\mathsf{\pi }\left(0\right)$}\!\left/ \!\raisebox{-1ex}{$\left[1-\mathsf{\pi }\left(0\right)\right]$}\right.}}={\displaystyle \frac{{\mathrm{e}}^{{\mathsf{\beta }}_{0+{\mathsf{\beta }}_1}}}{{\mathrm{e}}^{{\mathsf{\beta }}_0}}}={\mathrm{e}}^{{\mathsf{\beta }}_1}$$

(11)

Under the assumption that the value of ψ is equal to 1, it is possible to conclude that there is no relationship between the two variables. If the value of ψ is less than 1, it signifies a negative link between the two variables and the category change of the value of x. In contrast, if the value of ψ is greater than 1, there is no association of this kind.

3.3.8. Hosmer and Lemeshow Test

Testing the goodness of fit is an essential step in evaluating a statistical model. The Hosmer–Lemeshow goodness-of-fit test was used to assess the binary logistic regression models constructed in this research. The test aims to determine whether or not the observed event rates correspond to the expected event rates in subgroups of the population described by the model.

4. Results and Discussion

4.1. Results of Binary Logistic Regression Analysis

4.1.1. Result of Nagelkerke R-Square Test

The Nagelkerke R-Square test yielded a value of 0.760. The results indicated that the independent variables accounted for 76.0 percent of the dependent variable, specifically, the effectiveness of extension in clove farming, whereas the other independent variables, which were excluded from the analyzed model, accounted for the remaining 24.0 percent of the explanation.

4.1.2. Likelihood Ratio Test (LRT)

The LRT test is utilized to determine whether or not the model is appropriate for a particular circumstance. The omnibus test table is used to compare the estimated Chi-square value to the Chi-square table value and the significance value with a critical threshold of 5%. The estimated Chi-square value of 83.376, as presented in

Table 3. Simultaneous test results (G-test).

Omnibus Tests of Model Coefficients
	Chi-Square	Df	Sig.
Step	83.376	14	0.000
Block	83.376	14	0.000
Model	83.376	14	0.000

, was more than the Chi-square table value of 23.685. Since the significance value in the table is 0.000, which is less than the threshold of 0.05, the null hypothesis (H₀) was rejected. This rejection was carried out according to the rule that says that H₀ is rejected at a significance level (α) if G is more than the Chi-square (χ² (α, v)) value and the significance value in the test statistic is less than α. This finding establishes a causal relationship between the dependent variable and one or more independent factors.

4.1.3. Partial Test (Wald Test)

The Wald test is just one example of the many different forms of significance tests available. One way to obtain an estimation of the parameter β is to square the quotient of the parameter estimate with the standard error by using this method. In light of this, the conclusion that can be drawn from the estimation is as follows. Whenever a variable is incorporated into a model, this test is carried out before running the model. The decision-making criteria are applied in this test, and a significant threshold of α = 0.05 is utilized. If the significance value is found to be less than α or if the value of W is found to be greater than χ² (α, df), then, the null hypothesis (H₀) is rejected at a level of significance (α), as seen in

Table 4. Wald test results.

Independent Variables	B	S.E.	Wald	df	Sig.	Exp(B)
Farmer Age (FA)	0.188	0.117	2.590	1	0.108	1.207
Educational Attainment (EA)	−0.365	0.134	7.435	1	0.006 *	0.694
Family Size (FS)	2.929	0.876	11.186	1	0.001 *	18.703
Farming Experience (FE)	−0.335	0.138	5.864	1	0.015 **	0.715
Farmers’ Cosmopolitanism (FC)	−3.645	1.238	8.672	1	0.003 *	0.026
Individual Communication Approach (ICA)	0.713	0.580	1.512	1	0.219	2.041
Group Communication Approach (GCA)	1.276	0.843	2.295	1	0.130	3.584
Mass Communication Approach (MCA)	1.790	0.845	4.483	1	0.034 **	5.989
Print Media (PM)	−2.415	0.897	7.246	1	0.007 *	0.089
Electronic Media (EM)	1.928	0.680	8.039	1	0.005 *	6.874
Social Media (SM)	1.067	0.599	3.167	1	0.075	2.906
Clove Cultivation Material (CCM)	2.257	0.902	6.257	1	0.012 **	9.554
Land Area (LA)	−1.447	0.762	3.608	1	0.058 ***	0.235
Capital (CP)	0.000	0.000	4.175	1	0.041	1.000
Constant	−15.113	6.681	5.117	1	0.024	0.000
Dependent Variable: The Effectiveness of Agricultural Extension in Clove Farming (EE)

Note: * significant at 99% confidence level; ** significant at 95% confidence level; and *** significant at 90% confidence level.

.

The significance level was set to 0.05, with a degree of freedom of 1; the Chi-square table indicated a value of 3.481. These values provided proof that the results were statistically significant. Furthermore, based on the Wald test results presented in Table 4, it was known that the variables of educational attainment (EA), family size (FS), farming experience (FE), farmers’ cosmopolitanism (FC), mass communication approach (MCA), print media (PM), electronic media (EM), clove cultivation material (CCM), and land area (LA) had a statistically significant effect on the effectiveness of extension in clove farming (EE). Based on these findings, it was discovered that the projected test values for these variables were more than the critical value of 3.481 that was generated from the Chi-square table. The findings of the Wald test support the conclusion that the variables are statistically significant at a level lower than 0.05. The conclusion reached was that H₀ was rejected, with H₁ being accepted. The other independent variables, such as farmer age (FA), the individual communication approach (ICA), the group communication approach (GCA), social media (SM), and capital (CP), did not have a significant impact on the EE.

Equation (12) is the binary logistic regression model equation formed in this research. The equation derived from Table 4 contains the results of the Wald test, which were significant variables in the model tested.

$$\begin{gathered} \mathrm{g}\left(\mathrm{E}\mathrm{E}\right)={\displaystyle \frac{in\left(\mathsf{\pi }\left(\mathrm{E}\mathrm{E}\right)\right)}{1-\mathsf{\pi }\left(\mathrm{E}\mathrm{E}\right)}} \\ =0.188\mathrm{F}\mathrm{A}+\left(-0.365\right)\mathrm{E}\mathrm{A}+2.929\mathrm{F}\mathrm{S}+\left(-0.335\right)\mathrm{F}\mathrm{E} \\ +\left(-3.645\right)\mathrm{F}\mathrm{C}+0.713\mathrm{I}\mathrm{C}\mathrm{A}+1.276\mathrm{G}\mathrm{C}\mathrm{A}+1.790\mathrm{M}\mathrm{C}\mathrm{A} \\ +\left(-2.415\right)\mathrm{P}\mathrm{M}+1.928\mathrm{E}\mathrm{M}+1.067\mathrm{S}\mathrm{M}+2.257\mathrm{C}\mathrm{C}\mathrm{M} \\ +\left(-1.447\right)\mathrm{L}\mathrm{A}+0.000\mathrm{C}\mathrm{P}+{\mathsf{\epsilon }}_{\mathrm{i}} \end{gathered}$$

(12)

4.1.4. Model Fit Test

The Hosmer and Lemeshow Test is a method that can be applied within the framework of binary logistic regression to evaluate the degree to which the data correspond to the model. The Chi-square value also can be utilized in this test to evaluate the degree to which the data are consistent with the model. This makes it possible to determine the degree to which the model and the findings are compatible with one another. A value of 1.944 was obtained for the Chi-square statistic, and 0.983 was obtained for the significance evaluation. We performed a computation using the Chi-square table, yielding a result of 15.507. For the sake of this computation, the significance level was established at 0.05, and the degree of freedom was established at 8. Even though the significance value of 0.983 was greater than the alpha level of 0.05, which was computed earlier, the estimated Chi-square value of 1.944 was lower than the crucial Chi-square value of 15.507. This is because the value of the crucial Chi-square was much higher than the value of the significance threshold. So, it was concluded that the observed and predicted values were not statistically different after these modifications.

4.2. Interpretation of the Odds Ratio

A quantitative depiction of the connection between a causative component and the associated effect can be accomplished by the computation of the odds ratio, denoted by the expression OR/Exp(B). Chen et al. [132] and Ospina et al. [131] concluded that it allows researchers to explore whether the likelihood of an occurrence is the same or different between two distinct groups of people from various backgrounds. According to Chen et al. [132] and Ospina et al. [131], the value of the ratio may be anywhere from zero to infinity. A graphical summary of variables that significantly affected the effectiveness of agricultural extension in clove farming is presented in Figure 5. The regression results of this research were interpreted using Figure 5 and are shown in the following section.

4.2.1. Effect of Farmer Characteristics

• Effect of Educational Attainment (EA) and Family Size (FS) on the Effectiveness of Agricultural Extension in Clove Farming

In this study, the independent variable EA was examined for its effect on the variable EE. As presented in Figure 5, this variable was significant in the model tested. Its significance value of 0.006 was less than the alpha value of 0.05. Meanwhile, the odds ratio value for the EA variable was 0.694, with an estimated value (B) of −0.365. This value indicates a negative influence of the EA variable on the effectiveness of agricultural extension in clove farming. Based on this value, it was reasonable to believe that the time a farmer has spent obtaining an education could diminish the effectiveness of agricultural extension in clove farming. We argue that the farmers who devote more time to formal schooling have lower levels of motivation and interaction in the agricultural sector. As a result, they experience less success in managing their farms. This finding is in line with those of Tham-Agyekum et al. [133], who also found a decrease in this perception. On the other hand, the findings of Azizah [127] show that the level of education that farmers possess would impact the decisions they make regarding implementing a new invention they are considering. Higher education provides a more comprehensive understanding, which makes it simpler for farmers to accept innovations. Moreover, Arifianto et al. [64] indicated that the level of education possessed by extension workers substantially impacts their overall performance capabilities. There is still a lack of clarity regarding the reason for this inverse link, which may call for additional research.

Moreover, this study looked at the family size (FS) variable to see how much it affects the effectiveness of agricultural extension in clove farming (EE). This variable significantly affected the EE variable, whose significance value was 0.001 (Figure 5). This value was less than the 99% significance level (α = 0.01). The estimated value (B) for the family size (FS) variable was 2.929, and the odds ratio (OR) value for the variable was 18.703. These specific values indicate that the FS variable positively impacted the EE variable. These values also indicate that there is a possibility that an increase in family size (FS) could have increased the effectiveness of agricultural extension in clove farming. In other words, the likelihood of agricultural extension being successful is increased when there is a greater number of family members who are dependent on the farm. This finding is in line with the research of Suvedi et al. [134] and Aniagyei et al. [51], which showed that one of the factors affecting farmers’ participation in agricultural extension is family resources, which relate to the division of tasks in the household. Therefore, farmers who have many family resources will help implement extension programs. The findings of Khalid [135] asserted that extension activities aim to improve the efficiency of farming families, raise productivity, and generally improve the living conditions of farming families. This suggests that increasing the dependent family’s size will help better understand the acceptance of extension innovations [67,68]. Moreover, Nabila et al. [69] argued that farmers are compelled to spend more on their day-to-day expenses if more family members depend on them. This is because farmers are responsible for providing for their families. On the other hand, compared to the findings of Berhanu et al. [70], the family size variable significantly and negatively influences the acceptability of extension innovations. The findings of this study indicate that the opposite is true.

• Effect of Farming Experience (FE) and Farmers’ Cosmopolitanism (FC) on the Effectiveness of Agricultural Extension in Clove Farming (EE)

This study also investigated the effect of the farming experience (FE) variable on the effectiveness of agricultural extension in clove farming (EE). As shown in Figure 5, the significance value of this variable was 0.015. This variable significantly affected the EE variable because the value was less than the alpha value of the significance level of 95% (α = 0.05). Meanwhile, the odds ratio of the FE variable was 0.715, with an estimated value (B) of −0.335. This value indicated a negative effect of the FE variable and the agricultural extension’s effectiveness in clove farming. Based on this value, we can conclude that the farming experience variable had a negative impact on the effectiveness of agricultural extension in clove farming. There is a possibility that the effectiveness of agricultural extension could decline as farming time grows. It is also likely that experienced farmers find agricultural extension less effective because a farmer with more experience managing his clove farm would no longer be interested in agricultural extension. Consequently, he would experience a significant reduction in the influence of agricultural extension activities. The findings obtained in this study are consistent with the conclusions obtained in previous studies. Hassan et al. [136] stated that the amount of time spent on farming was a significant factor that became a barrier in the agricultural extension process. The evidence for this is that some farmers have been in the farming business for a considerable time, and it may be difficult to convince them to abandon their traditional farming practices in favor of more contemporary farming methods. This is one of the basic explanations for the phenomenon. On the other hand, contrary to the findings of [35,71,73], it has been found that farmers who have a long history of farming experience have a considerable beneficial impact on the effectiveness of extension.

Moreover, the independent variable of farmers’ cosmopolitanism (FC) was examined to see how much it affected the effectiveness of agricultural extension in clove farming (EE). This variable had a significant influence on the EE variable. Its significance value (0.003) was less than the alpha value of 0.01 or the 99% significance level. At the same time, the FC variable had an odds ratio (OR) of 0.026, with an estimated value (B) of −3.645. This value indicates that the farmers’ cosmopolitanism variable had a negative effect on the EE variable. We can conclude that as the number of cosmopolitan farmers increases, the effectiveness of agricultural extension in clove farming may diminish. This result is in line with those of Yusliana et al. [77] and Utami [76], who explained that cosmopolitan farmers were included in the low category because farmers rarely seek information outside their village. Furthermore, Setiyowati et al. [75] and Utami [76] stated that the distance between the village and the information center and the difficulty in accessing public transportation cause farmers to be reluctant in seeking information on their own, leading them to spend more time gardening and waiting for extension workers or guests who come to visit their village to receive information related to their farm.

4.2.2. Agricultural Extension Approaches: Effects of Mass Communication Approaches (MCA) on the Effectiveness of Agricultural Extension in Clove Farming (EE)

Regarding the effectiveness of agricultural extension in clove farming (EE), we also explored how the independent variable of the mass communication approach (MCA) impacts clove farming’s effectiveness. A significant value of 0.034, less than the alpha value of 0.05, was reported for the MCA variable concerning its effect on the EE variable based on the data shown in Figure 5. At the same time, it was discovered that the odds ratio (OR) for the MCA variable was 5.989, while the estimated value (B) was determined to be 1.790. This depiction of the data demonstrates that the MCA variable had a large and positive influence on the EE variable. Considering these observations, it is feasible to predict that any new mass communication approach (MCA) might make agricultural extension work more effectively for clove growing. It has been determined from this study’s findings that increasing the adoption of mass communication methods has the potential to improve the effectiveness of agricultural extension. Mass extension is one of the techniques that is used to transmit information extensively from extension workers to a large number of targets in a short length of time, as stated by Tumurang et al. [79], Imran et al. [83], and Azumah et al. [84]. These research groups also announced their findings to the public. A considerable degree of congruence was discovered between the outcomes of this study and the conclusions of the research that Ramadhana [78] carried out. The results of the study conducted by Tambo et al. [85] indicate that exposure to campaign and mass channels, either on their own or in combination, has a large and beneficial influence on the effectiveness of agricultural extension.

4.2.3. Farmer Information Sources

• Effects of Electronic Media (EM) and Print Media (PM) on the Effectiveness of Agricultural Extension in Clove Farming (EE)

The EM variable was also investigated in this study to determine its influence on the effectiveness of agricultural extension in clove farming. This variable was one of the independent variables that substantially affected the dependent variable (EE), as seen in Figure 5. The magnitude of the significance value, which was 0.005, was lower than the alpha value of 0.05. The electronic media variable’s odds ratio (OR) value was 6.874, while the estimated value (B) was 1.928. This value indicates a positive effect of the electronic media variable on the variable of EE. This value showed that any increase in the use of electronic media as a source of information could increase the effectiveness of agricultural extension in clove farming, leading to a considerable and beneficial influence. When this statistic is considered, it is reasonable to assert that the growing exploitation of electronic media could increase the effectiveness of agricultural extension to a greater extent. The findings of this study are consistent with those of other studies, such as [27,44,48,76,81], which demonstrated that the content presented in the media contained material that farmers specifically required to improve farming. In addition, as a result of the findings, which demonstrated a beneficial influence of media factors (audiovisual) on extension, farmers believed that disseminating information through electronic media was suitable [93,94,95].

Furthermore, the PM variable was examined to see how much it affected the effectiveness of agricultural extension in clove farming. In Figure 5, the PM variable significantly affected the variable EE. Its significance value of 0.007 was less than the alpha value of 0.05 for the significance level of 95%. Meanwhile, the PM variable had an odds ratio (OR) of 0.089 and an estimated value (B) of −2.415. This value shows a negative effect of the PM variable on the effectiveness of agricultural extension in clove farming. This means that any increase in the use of print media as a source of information can hamper the effectiveness of agricultural extension in clove farming. Since some farmer respondents were literate while others were illiterate, printed media such as books, brochures, and leaflets that extension workers utilize as information media are less successful. The studies by Al-Zahrani et al. [50] and Anyanwu [88] also confirm this finding, with evidence suggesting that farmers in the region under investigation do not have access to formal education, which ultimately makes it more difficult to employ more efficient agricultural practices.

• Effects of Clove Cultivation Material (CCM) on the Effectiveness of Agricultural Extension in Clove Farming

This research also examined the effect of the CCM variable on the effectiveness of agricultural extension in clove farming. The CCM variable was significant in the model tested, as presented in Figure 5. The significance value was 0.012, less than the alpha value of 0.05, while the CCM variable had an odds ratio (OR) of 9.554 and an estimated value (B) of 2.257. This value demonstrates a positive and significant influence of the CCM variable on the effectiveness of agricultural extension in clove farming, indicating that increasing clove farming material could potentially improve the effectiveness of agricultural extension. The basis for this conclusion was that farmers who learned more about clove production practices would be more successful in managing their clove farming. This finding was corroborated by a study conducted by Nuthall et al. [107], which demonstrated that agricultural information needs have a significant part in illuminating individuals, raising their level of knowledge, and eventually assisting them in decision-making processes about agricultural operations. This implies that extension services address the practical difficulties that farmers experience and provide them with options to improve their farming methods [39,108]. Extension services can effectively satisfy the information needs of farmers, as stated by Baah [110], Sujianto et al. [34], and Mariel et al. [104]. This is accomplished by ensuring that the material is factually accurate, vetted by experts, and linked to the already available knowledge. The research findings indicate that when the credibility of the content is improved, favorable attitudes are displayed. According to Nuthall [107], extension services are an essential resource for farmers since they enable them to contribute to developing their capabilities and assist in implementing sustainable agricultural methods. Extension services are extremely valuable as a resource for farmers, and their total influence highlights their importance.

4.2.4. Farm Characteristics: Effect of Land Area (LA) on the Effectiveness of Agricultural Extension in Clove Farming

The LA variable was one of the significant variables in the model tested in this research. The significance value of this variable was 0.058, which was less than the significance level of 90% (α = 0.10). Meanwhile, the LA variable’s odds ratio (OR) was 0.235, with an estimated value (B) of −1.447. Since this value was negative, it is clear that the land area variable had a negative impact on the effectiveness of agricultural extension in clove farming. Given these data, it is reasonable to assume that the effectiveness of agricultural extension will decrease in proportion to the increase in land area, consistent with the findings of prior research. According to Tham-Agyekum et al. [133], those with smaller land sizes tended to favor extension services more favorably. There could be various reasons for this finding, including that farmers with smaller or bigger land holdings have different requirements or may see the services as more customized to their particular circumstances [137]. However, in contrast to this study, other researchers found that land size has a positive and significant correlation with extension effectiveness, as seen from the increase in knowledge, skills, and changes in farmers’ attitudes to an increase in farm production [35,53,111,112,113]. Furthermore, Alberth [49] argued that land area characteristics are closely related to agricultural extension participation.

5. Conclusions and Recommendations

This research objective was to examine the factors influencing the effectiveness of agricultural extension in clove farming in Sidrap Regency, South Sulawesi Province, Indonesia. A structured interview was employed to collect primary data from 140 clove farmers, who were randomly selected. The data collected were analyzed using binary logistic regression to test the influence of fourteen independent variables on the effectiveness of extension in clove farming, which was the dependent variable. The main finding is that all fourteen independent variables significantly affected the effectiveness of agricultural extension in clove farming. Based on the partial test results, it was found that the variables of family size, mass communication approach, electronic media, and clove cultivation material had a positive significant effect on the effectiveness of agricultural extension in clove farming. The results of this study are important for spurring and encouraging the effectiveness of agricultural extension in clove farming. The significant variables that were identified in this research can be used to determine the policy recommendations for boosting clove production and improving clove farming management in the short and long term.

Based on the findings presented above, the research concludes that family characteristics, approaches to agricultural extension services, and communication methods play an important role in the dissemination of technology and improving the effectiveness of agricultural extension when it comes to delivering extension material and programs to farmers. Based on what we have discovered thus far, policymakers should prioritize teaching more about how to cultivate cloves, employing mass extension methods more frequently, and including more electronic media into extension operations in research sites. Because of these improvements, extension services will become more useful. Furthermore, the national government’s availability of resources has a considerable impact on the implementation of extension services. It is critical to offer farmers and other stakeholders in rural areas the necessary knowledge and services to improve their lives and encourage sustainable farming. Agricultural extension and consultation services are critical in this regard. These strategies are believed to improve the effectiveness of agricultural extension in clove farming.