In production theory, production functions explicitly model the relationship between inputs and output, providing a theoretical framework for analyzing the impact of specific input allocation decisions on agricultural output (Suh and Moss, 2021). The Cobb-Douglas production function, widely used in agricultural economics, provides valuable insights into how changes in factor utilization affect output, allowing the elasticity of inputs and their marginal contributions to be estimated (Gautam, 2024). According to Motbaynor Workneh and Kumar (2023), production functions can be used to analyze the technical efficiency of farms and identify areas for improvement in input allocation. Furthermore, research has shown that optimal input allocation can lead to significant increases in agricultural productivity and efficiency (Wake, 2019). For example, a study by Gaviglio et al. (2021) found that farms that optimized their input allocation were more likely to achieve technical efficiency and increase their output.

Land size is a crucial resource (Cuong et al., 2019) and an absolute requirement for farmers to be able to produce agricultural products (Singirankabo and Ertsen, 2020). However, land ownership is minimal for small communities (Slätmo, 2019). Thus, land use must be adjusted to the suitability of the commodity being grown (Taghizadeh-Mehrjardi et al., 2020). Increased land use and agricultural inputs can directly increase production (Emran et al., 2021). Therefore, to support the increase and sustainability of production, national food security must be supported by the per capita land area owned by farmers (Adhi et al., 2022). Given the impact of land size on domestic potato production, researchers from various government agencies, colleges, non-governmental organizations, and corporations in Indonesia have focused on land. The conduct of many studies in multiple locations in Indonesia demonstrates this concern. Ananda Lubis et al. (2021) have researched the impact of Land area on potato production in Karo Regency, Wahyuni and Saputra (2023) in Gunung Kidul Regency, Amrullah et al. (2024) in Bantaeng Regency, Nugraha and Nugroho Sumarjiyanto Benedictus Maria (2021)) in Grobogan Regency, Kharismawati and Karjati (2021) In 10 regencies in East Java, Rozi et al. (2020) in Kediri. The results produced by these researchers demonstrate that land size has a significant impact on potato production. In addition, similar studies have been conducted in several countries worldwide, such as Lun et al. (2023) in China, Raina et al. (2024) in India, Zarzyńska (2023) in Poland and Andaregie and Astatkie (2020) in Ethiopia, they discovered that land size had a favorable and considerable impact on potato yields.

In addition to land area, seed use was also found to affect potato production. The selection and use of quality seed is essential in potato farming (Wasilewska-Nascimento et al., 2020). Various scientific studies have revealed this. Other researchers and academics at multiple research institutions and universities worldwide have researched the impact of seed characteristics on potato production. Rumallang (2019) in Indonesia, Okello et al. (2017) in Kenya said that certified seed significantly increases potato farm production. These results align with those of Boubaker et al. (2023) in Tunisia and Zhou et al. (2022) in China, which showed that the higher the seed class used, the higher the increase in production. Furthermore, based on their research results, Bist et al. (2023) in Nepal said that shoot and tuber weight really affected the number of potato tubers. Then, Bajgai et al. (2018) in Bhutan, a small South Asian country, and Kumar et al. (2023) in India stated that the adoption of quality seeds substantially impacts potato yield. Likewise, an Investigation by Husen et al. (2021) in Indonesia and Lal et al. (2023) in India found that seed from stem cuttings can increase potato growth and production. Furthermore, the use of tuber sizes grown by farmers was also able to increase potato production (Boubaker et al., 2023).

Furthermore, the use of pesticides in farming significantly increases crop productivity and prevents losses from pests and diseases (Tudi et al., 2021). Prevention of pests and diseases in plants can be controlled by using synthetic pesticides and vegetable pesticides (Ngegba et al., 2022). Vegetable pesticides are a type of pesticide obtained from plants that are used to reduce yield losses due to pest attacks (Kumar et al., 2021). Furthermore, the use of vegetable pesticides has a good ability to control pest attacks on potato plants (Hidayah et al., 2017). Research by Fianda et al. (2016) showed that pesticides have a significant effect on potato production. Furthermore, the efficient use of fungicides can increase crop yields (Lamichhane et al., 2020). Prima et al. (2020) potato tuber rot disease caused by Phytophthora infestans can be controlled by bio-fungicides made from antagonistic fungi, namely Trichoderma sp., because it can inhibit the growth of Phytophthora infestans. This aligns with the findings of Yanti Azzahra et al. (2022) using fungicides can overcome blight and fungal blight caused by Phytophthora infestans. Research by Nurdiana and Fatima (2018) discovered that natural fungicides had a considerable effect in inhibiting the growth of the fungus Fusarium Oxysporum. Likewise, Boydston et al. (2012), in the United States, Abdallah et al. (2021) in Egypt, and Ginter et al. (2022) in Poland, these researchers found that herbicides can increase potato production. Furthermore, Kaleem Ullah et al. (2023) in China, Thornton et al. (2010) in America and Yusdian et al. (2022) in Indonesia, they found that insecticides can overcome pests of leaf caterpillars, aphids, and grasshoppers in potato plants. In contradiction to the studies by Devi et al. (2018) in India and Barbaś et al. (2024) in Poland, they found that pesticide use decreased potato production.

Fertilizers are natural or artificial substances used in plants or soil to provide nutrients needed for plant growth and boost production (Jote, 2023; Akkamis and Caliskan, 2024). Chemical fertilizers are one of the most commonly applied methods in intensive agricultural practices (Gao et al., 2023). The practical use of inorganic fertilizers increases the satisfaction in cultivation because the results can be seen immediately on the plants (Neoriky et al., 2017). In potato crops, inorganic fertilizers such as urea and NPK fertilizers are widely used (Fang et al., 2023). Applying urea fertilizer can significantly increase plant growth and yield (Jiang et al., 2022). Mokrani et al. (2019) said that potato yield is affected by NPK fertilizer, cultivar, and tuber development stage, primarily through its effect on sugar accumulation and distribution associated with photosynthetic efficiency. Other researchers and academics at various research institutions and universities have researched the impact of urea and NPK fertilizers on potato production. Wang et al. (2024) in China, Nagar and Yadav (2019) in Rajasthan, and Oliveira et al. (2021) in Brazil, said that the use of urea and NPK fertilizers affects potato production.

In addition to the effect of inorganic fertilizers, it was also detected that organic fertilizers can increase potato production. Organic fertilizers contain a high percentage of organic matter. They are rich in nutrients, which can improve soil physical properties by increasing aggregate stability and decreasing soil density. In addition, these fertilizers can also enhance the natural and biochemical qualities of the soil and support the improvement of the soil microbial community structure (Ilahi, 2021). The use of organic fertilizers has a significant positive impact on crop production (Lin et al., 2022) because it can increase P nutrients in the soil (Zhou et al., 2022; Bako et al., 2023). Furthermore, combining organic fertilizers with chemical fertilizers also increases nitrogen use efficiency (Lin et al., 2022). Organic fertilizers help improve sustainable agricultural practices by enhancing long-term soil health and decreasing reliance on chemical inputs (Fan et al., 2023). Furthermore, using organic fertilizers improves crop quality by increasing the soil's levels of key nutrients, such as phosphorus and nitrogen. Also, it improves fruit quality by increasing the content of lycopene and vitamin C (Gao et al., 2023). Manure is one of the organic fertilizers that can enhance soil nutrients and increase production. This evidence was presented by Nyiraneza et al. (2021) in Canada, Ahmed et al. (2019) in Bangladesh, and Grandy et al. (2002) in America. Their research found that manure use affects potato production. The same thing was stated by Ierna and Distefano (2024) cow manure application resulted in better growth yields, namely plant height and number of leaves. Furthermore, the significant impact of using cow waste and inorganic fertilizer in combination can increase potato production (Gelaye, 2023).

Furthermore, labor is also a concern, as it is an essential factor in agriculture (Mihi-Ramirez et al., 2022). Labor availability is essential to the farm's sustainability. The presence of labor can have a beneficial or bad impact on the advancement of the farm, depending on how much energy the worker contributes (Kurniati and Sisca, 2020). In addition, labor is vital in managing business inputs (Basri et al., 2020). The more labor is required or used in production, the more a product is produced. Conversely, the less labor used or required in production activities, the less a product is produced (Safira and Juliansyah, 2019). Labor has a significant, positive impact on production (Adam and Sadimantara, 2022; Siaw et al., 2021). This discovery is consistent with the outcomes of studies by Budi Setiawan and Inayati (2020) in Central Java, Lidyana and Sulistiyowati (2022) in East Java and Salim et al. (2019). They said that, simultaneously, labor expenditure has a significant effect on the quantity and productivity of granola potato farming. Vu et al. (2020a) in Vietnam, where the study informs that manpower has a significant and positive influence on potato agricultural production.

Age plays a vital role in agriculture, affecting productivity and sustainability. As they age, farmers often face challenges related to physical capacity, which can impact their productivity and work efficiency (Szabo et al., 2021). Experienced and age-mature farmers tend to be more effective in adopting agricultural innovations, thereby increasing their yields and farm efficiency (Emran et al., 2021). In addition, the age of farmers plays a role in land management, including determining the business alternatives to be implemented on each land plot (Rumallang et al., 2023). Research conducted by Key (2022) in the United States, young farmers can achieve higher production levels than older farmers. Meanwhile, Marbun (2024) and Kurniati and Sisca (2020) in Indonesia, farmers' productivity increased with age but declined when they entered middle age. In contrast, Susanti et al. (2016) in Karanganyar Regency and Ramdhan et al. (2020) in Tasikmalaya Regency found that farmers' age did not have a significant effect on production.

In addition to the farmer's age, farming experience is also part of what supports farming activities because it refers to the amount or length of time a farmer has been involved in agricultural activities (Zhou and Li, 2022). This farming experience covers various aspects, such as knowledge of farming practices, land maintenance, pest and disease management, use of agricultural technology, and adoption of innovations. The longer a farmer runs a farm, the more skills and insights he or she generally gains, which can positively impact farm productivity and efficiency (Widi and Isyanto, 2020). Farmers should utilize their experience while incorporating innovations, sustainability principles, and effective land management as core elements in production activities (Gadanakis, 2024). In terms of agricultural innovation adoption, farming experience also has a significant effect, thus increasing production (Munawaroh et al., 2020; Emran et al., 2021). In line with this, Kurniati and Sisca (2020) in Kuantan Singingi Regency, Pambudi and Bendesa (2020) in Buleleng Regency, and Tunas et al. (2023) in the Minahasa Regency, it was discovered that, consequently, farming experience has a supportive and important impact on production, unlike the research of Gustiana and Irwanto (2017) in Aceh Tamiang Regency, and Burano and Siska (2019) in West Sumatra, which informed that the farming experience variable had no impact on production.

Furthermore, education plays a crucial role in improving the capacity of farmers and agricultural workers to adopt new technologies and more efficient and sustainable farming practices. Education in developing countries has been one of the main focuses of agricultural development research, as it can increase productivity (Luh, 2017). Education is recognized as an essential factor in increasing productivity through improving the skills and abilities of the workforce (Rizzo et al., 2024). One of the determinants of success in farming activities is the level of formal education that farmers have (Widi and Isyanto, 2020). Knowledge helps farmers access and apply advanced technologies, such as improved seed varieties and sustainable farming practices, resulting in higher yields and better incomes. In addition, educated farmers are more likely to participate in extension services and farmer groups, which increases their ability to adopt innovations and improve productivity (Rizzo et al., 2024). Research by Gille (2020) in India showed that education level contributes to increased production and productivity of agricultural products. Similarly, Paltasingh and Goyari (2018) in India believe that the knowledge level of farmers impacts the adoption of modern technology that supports increased agricultural productivity. Likewise, Susanti et al. (2016) in Indonesia said that the education level impacts farmers' education level.

A research gap refers to an area or issue within a field where existing knowledge is insufficient, hindering the ability to draw definitive conclusions or findings from a particular study (Ajemba and Arene, 2022). It represents aspects of the literature related to a specific topic that have either not been fully explored or are still under investigation (Baako et al., 2022). Finding these holes so that future studies can fill them is the primary goal of a literature review (Müller-Bloch and Kranz, 2015). Although it is acknowledged that finding research gaps in a literature review is essential, there aren't yet well-defined methodological standards. This is a problem because it makes it hard to guarantee accuracy and repeatability (Müller-Bloch and Kranz, 2015). Azeez and Elegunde (2022) found seven different kinds of knowledge gaps in their research: There are several gaps in our current understanding of the following topics: (1) evidence, (2) knowledge, (3) the relationship between theory and practice, (4) methodology, (5) data, (6) theory, and (7) population. On the other hand, Ajemba and Arene (2022) offered an alternative classification, dividing research gaps into four categories: (a) gaps in research techniques and design, (b) gaps in research variables, (c) sampling methods, and (d) gaps in data collection.

As summarized in Table 1, previous studies focused on productivity determinants using various econometric and frontier models. However, none specifically analyzed the probability of achieving high potato yields based on combined input and socioeconomic factors using binary logistic regression. This study fills methodological and empirical gaps by integrating input allocation and spatial factors into a single analytical framework.

Two variables are studied in this study: the variables that affect (independent) and those that are affected (dependent). Dependent variables are those influenced by independent variables. Independent factors produce changes in the dependent variable (Yuniarsih et al., 2024). As the name implies, variables are objects, events, things, and attributes that vary in value and are to be studied, measured, described, and interpreted in research. The dependent (bound) variable in this study is potato production (Figure 2), and the independent variables are described in the previously stated literature review. Based on the previous section, 14 variables are suspected to affect potato production in this research. The conceptual connection between independent and dependent variables is presented in Figure 2.

This study was conducted in the sub-districts of Tinggimoncong and Tombolopao, located in Gowa Regency, South Sulawesi Province, Indonesia (Figure 3). These areas were purposively selected for their role as major centers of potato cultivation and production, supported by highly favorable agroclimatic conditions for potato farming. Both regions are situated at an altitude of 1,500 meters above sea level and experience temperatures ranging from 16 to 20 degrees Celsius (Sastrika Anindita et al., 2024), making them ideal for potato cultivation. The fieldwork was carried out between October and December 2023. The data utilized in this study were primarily obtained from potato farmers. The primary data set is quantitative and was collected through structured interviews with pre-designed questionnaires. A total of 223 potato producers were selected as respondents through random sampling. This figure represents approximately 32% of the active potato farmers in the research area (680 farmers). To determine how many samples to take, we use the Slovin formula (Santoso, 2023).

Where: n = Sample amount; N = No. of population; e = The precision of 5%, with the following calculation:

Binary logistic regression is a data analysis tool that can help you examine the connection between your continuous or categorical predictor variables (x) and your binary (dichotomous) response variables (y) (Getu and Bhat, 2024; Yuniarsih et al., 2024). Equation 2 represents the simplest linear regression model.

Where: Y = Predicted value (response variable); X = predictor variable; β₀ = Constant; β₁ = Coefficient of regression (decrease or increase in value); e = Random error.

Linear and Ordinary Least Squares (OLS) regression are two examples of predictive modeling approaches comparable to logistic regression. The key difference is that researchers use a binary scale to predict dependent variables in logistic regression. According to Getu and Bhat (2024), Yuniarsih et al. (2024), and Bordeaux and Couto (2024) the nominal data scale in question is a dichotomous scale with two distinct categories. When one answer variable is provided as dichotomous qualitative data, where 1 denotes the presence of a characteristic and 0 denotes its absence, the binary logistic regression model is employed to examine the connection between the response variable and multiple predictor variables. Finally, this regression model is based on the Bernoulli distribution (Equation 3).

Where: π= signifies the odds of the i-th occurrence; yi= indicates the i-th binary variable, which can be either 0 or 1.

Typical structure of a single-predictor logistic regression model (Salam et al., 2024a,b), depicted in Equation 4.

In addition, as shown in Equation 5, the logit version of logistic regression is used to simplify the estimation of regression parameters by changing π(x) in the previous equation (Salam et al., 2024a,b).

This binary logistic regression method uses three indicators to evaluate the impact of independent variables on the dependent variable. Table 2 lists these measures, which include the odds ratio, regression coefficient, and significance level (Salam et al., 2024a,b).

The purpose of evaluating model fit is to determine if the model works with the data, which is when the values that are seen are near or even match the values that the model predicted (Yuniarsih et al., 2024). To be considered Goodness of Fit (GoF), the model has to follow specific rules. There must be congruence between the model's input data and the observed data for the model to be deemed Goodly Fitted (Bordeaux and Couto, 2024). In binary logistic regression, the Chi-square value from the Hosmer-Lemeshow Test can be used to evaluate the Goodness of Fit. Table 4 displays the outcomes of the examiner's assessment of the model fit.

The calculation of the Chi-square statistic, with a significance level of 0.05 and degrees of freedom (df) equal to 8, yields a Chi-square value of 15.507. The data in Table 4 indicate that the calculated Chi-square value is 15.167, with a significance level of 0.056. The results indicate that the null hypothesis (H₀) is not rejected, suggesting no statistically significant difference between the observed and expected values. This shows that the model is adequate and reliable. The conclusion is substantiated by the computed Chi-square value of 15.167, less than the Chi-square table value of 15.507. The significance level of 0.056 also surpasses the established alpha threshold of 0.05.

The Cox and Snell R-Square and Nagelkerke R-Square are metrics for evaluating the extent to which the independent variables account for the variation in the dependent variable. Table 7 presents the results of these tests. Table 5 shows a Nagelkerke R-square of 0.922, indicating that the independent variables in this study account for 92.2% of the variation in potato production. The remaining 7.8% of the variation can be attributed to other independent variables not included in the model examined. In other words, the simultaneous influence of the fourteen independent variables analyzed in this study explains 92.2% of the variation in potato production. This remarkably high explanatory value is attributed to the similarity of the potato farming systems practiced by the farmers in the study area and to the strong dependence of potato yield on farmers' input allocation decisions. In addition, the fourteen independent variables examined in this study could account for the unusually high Nagelkerke R-square value of 0.922.

To evaluate the combined impact of independent variables on the dependent variable, the simultaneous test (G-test) is used (Yuniarsih et al., 2024). To determine how well the model fits the data, this test compares the Log-likelihood value with all independent variables to the Log-likelihood value without them, using the likelihood ratio test (LRT) (Bordeaux and Couto, 2024). After that, we compare the computed Chi-square value to the values in the Chi-square table and check its significance at the 5% level. Table 6 displays the study's simultaneous test findings in the omnibus test table.

The simultaneous test results in Table 6 indicate a calculated Chi-square value of 259.92, significantly higher than the Chi-square table value of 23.684. This suggests that the computed Chi-square exceeds the value found in the table. Additionally, the significance level indicated in the table is 0.000, which falls below the threshold of 0.05. According to the decision rule, H₀ is rejected when G exceeds χ²(α, v), and if the significance level of the test statistic falls below α, then H₁ is also rejected. The findings from the G-test indicate that at least one of the independent variables in the model significantly influences potato production (Salam et al., 2024a,b).

Using a partial or Wald test, we examined the impact of each independent variable on the dependent variable. Area of land, potato seed, manure, NPK ponska fertilizer, pesticide, herbicide, fungicide, labor, education, farming experience, age of farmer, watering, and distance from the house were among the fourteen independent variables that were supposedly tested for their effects on potato output. To see how each independent variable affected potato output, see Table 7.

The test results indicate that the Chi-square table value is 3.841, corresponding to a significance level of 0.05. Consequently, the null hypothesis (H₀) gets dismissed, while the alternative hypothesis (H₁) is accepted. The regression results in Table 6 show that there are six independent variables that significantly and positively affect potato production (PK), namely the variables of seed potato (PS), manure (OF), NPK ponska fertilizer (NPK), urea fertilizer (Fu), insecticides (In), and watering (Wa). The analysis shows that the herbicide (He) and the distance of the house from the farm (DH) variables exert a significant negative influence on potato production. In the analysis, the variables of land area (LA), fungicide (UF), labor (FL), education (Fe), experience (EF), and farmer age (FA) were found to have no significant impact on potato production. The regression equation derived from the partial test results, specifically the Wald test, is outlined in Equation 13.

One way to compare the likelihood of an event occurring in two groups is to use the probability ratio, the odds ratio divided by the expected value of the outcome, to quantify the association between the two groups. The regression coefficient in logistic regression analysis indicates how the outcome variable's likelihood is expected to change for a one-unit increase in the predictor variable. According to (Chen et al. 2010), the odds ratio can be anything from zero to infinity.

In Figure 4, we can see the odds ratios calculated for this study and the Exp(B) coefficients for all the independent factors of potato production. Figure 4, which summarizes the Wald test results from Table 3, shows that all independent variables significantly affect potato production (PK) in this study. In Figure 4, three critical indicators are presented, viz: odds ratio (OR), regression coefficient (B), and significance value (Sig.). The following subsections explain the effects of each independent variable on potato production (PK).

Potato farmers in the research location choose and plant granola varieties because they produce more than other potato varieties. The research results show that the seed potato variable (PS) significantly affects potato production (PK). The significant effect of this variable is indicated by its significance value of 0.013. This figure is smaller than the alpha value (α = 0.05). The seed variable (PS) has an odds ratio of 1.014 and an estimated value (B) of 0.014, as shown in Figure 4. We may infer that if potato farmers increase the volume of seed they use, their potato farms will produce more potatoes, based on this variable's positive estimated value and odds ratio. In other words, potato farmers who can increase their seed potato have a chance or possibility of increasing their potato farm production by 1.014 times compared to potato farmers who cannot increase their seed use, assuming other variables are fixed (ceteris paribus). These findings support previous research conducted by Fianda et al. (2016) in Aceh Province, Kasutjianingati et al. (2018) in East Java Province and Gultom (2018) in North Sumatra Province, Indonesia, Vu et al. (2020b) in Vietnam, Boubaker et al. (2023) in Tunisia, Zhou et al. (2022) in China, Lal et al. (2023) in India, and Okello et al. (2017) in Kenya. The findings of studies conducted by these researchers across different provinces in Indonesia and several other countries show that seed potatoes have a significant, positive effect on production. Then, theoretically, seed potatoes significantly affect production because good seed quality, the right size, varieties suitable for agro-climatic conditions, seed health free from disease, optimal physiological age, healthy sprouts, and proper management will support optimal plant growth and increase yields. Meanwhile, low-quality or infected seeds can significantly reduce production and productivity Zhang et al., (2023).

Furthermore, the results also show that manure (OF) positively and significantly impacts potato production (PK). The significant influence of this variable is indicated by its significance value of 0.022, which is smaller than the alpha value (α = 0.05). The estimated value (B) for the manure variable (OF) is 0.013, and the odds ratio (OR) is 1.013, as shown in Figure 4. An increase in the volume of manure used by potato farmers is likely to boost potato farm productivity, as indicated by the odds ratio and positive estimate. Assuming all other factors remain constant, another way of looking at it is that potato farmers who can increase their use of manure have a 1.013 times greater probability of increasing their potato farm production than farmers who can't. This is consistent with the research of (Zhou et al. 2022), (Lin et al. 2022), (Lin et al. 2024) and (Yan and Gong 2010) in China, El-Ghamry et al. (2024) in Italy, Hammed et al. (2019) in Nigeria and Wurr and Morris (1979) in Italy, all of which explained that manure application significantly improved soil fertility and positively affected production. Theoretically, the use of manure contributes to production because it fertilizes the soil, increases the availability of nutrients in the soil, improves soil biology, and supports environmentally friendly agriculture Taye and Husvéth, (2020). Empirically, the research location uses chicken manure mixed with husks. Then, this fertilizer is applied at the beginning of planting as the main nutrient for potato plants at the beginning of their growth. However, the findings of (Kumar et al. 2016) in India, reminded that the use of manure does not always affect production. They found that manure did not have a significant effect on production.

The application of NPK Ponska fertilizer is carried out by farmers at the beginning of planting and at the age of potatoes, 20–25 days after planting. Farmers do this as the potato plant progresses through its growth phase. The results of this study show that the variable NPK ponska fertilizer (NPK) significantly affects potato production (PK). The significant impact of this variable is also evident in Figure 4, where the p-value is 0.020. This figure is smaller than the alpha value (α = 0.05). In the exact figure, this variable has an odds ratio of 1.018 and an approximate value (B) of 0.018. Based on the odds ratio and positive estimate, the increase in the volume of NPK Ponska fertilizer used by potato farmers is likely to increase their potato production. From the results of this study, it can also be concluded that potato farmers who can increase the volume of NPK ponska fertilizer use can increase potato production by 1.018 times compared to farmers who cannot increase NPK ponska fertilizer use, assuming other variables remain constant. This is consistent with research by Mokrani et al. (2019) in Tunisia, Adhikari (2014) in Nepal, Wang et al. (2024); (Sun 2019) and Bai et al. (2023) in China, Khalofah et al. (2022) in Saudi Arabia, Nagar and Yadav (2019) in Rajasthan, and Oliveira et al. (2021) in Brazil. These studies from different countries all found that using NPK fertilizers affected potato growth and production. Theoretically, it can be understood that the use of NPK Ponska fertilizer can have a significant effect on potato production because its macronutrient content, namely nitrogen (N), phosphorus (P), and potassium (K), is crucial for plant growth and development (Zahoor, 2016). Nitrogen helps promote vegetative growth, such as leaves and stems, which are essential for photosynthesis. Phosphorus plays a role in root formation and accelerates tuber maturation, while potassium enhances plant resistance to disease and improves tuber quality (Bošković-Rakočević et al., 2018).

Furthermore, potato farmers apply urea fertilizer during the early growth of potato plants because potato plants need the nitrogen it contains during the vegetative phase. Statistically, the results show that the urea fertilizer variable (Fu) is significant to potato yield (PK). The significant effect of urea fertilizer is indicated by its significance value of 0.030. This figure is smaller than alpha (α = 0.05). Furthermore, in Figure 4, the urea fertilizer variable (Fu) has an odds ratio value of 1.032 with an estimated value (B) of 0.032. Based on the odds ratio for this variable and its positive value, it can be inferred that an increase in the volume of urea fertilizer used by potato farmers increases the probability of increased potato production. In other words, potato farmers who can add urea fertilizer have a chance or possibility of increasing potato production by 1.032 times compared to farmers who cannot increase the use of urea fertilizer, assuming other variables are fixed (ceteris paribus). This empirical fact aligns with the research by Maulidiyah et al. (2024) in Bantaeng, Indonesia, and Oliveira et al. (2021) in Brazil, Taysom et al. (2023) in the United States, Shunka et al. (2021) and Addis et al. (2024) in Ethiopia, all found that the use of urea fertilizer significantly affected potato production. Slightly different from the study of Clément et al. (2021) in Canada, urea fertilizer did not significantly affect potato production, but it did affect vegetative growth. This study suggests that the correct application rate, timing, and target are the best ways to use urea fertilizer to increase potato yield.

Interviews with potato farmers in the research location indicate that potato plants planted in the dry season are very vulnerable to pest attacks, so farmers must use insecticides much more than in the rainy season to maintain and increase production. The plant's health must be maintained as early as possible to prevent pests and diseases that could affect potato production. The test results show that the insecticide variable (In) has a real and significant effect on potato production (PK). The considerable impact of this variable is indicated by its significance value of 0.002. This figure is smaller than alpha (α = 0.01). Furthermore, in the exact figure, the insecticide variable (In) has an odds ratio of 1.900 with an estimated value (B) of 0.642. Based on the odds ratio value of this variable and its positive estimate, it can be interpreted that an increase in the volume of insecticides used in potato farming increases the probability of increased potato production. Another explanation is that potato farmers who can add insecticides have the opportunity to increase potato production by 1.900 times compared to farmers who do not increase insecticide use on their farms, assuming other variables are fixed (ceteris paribus). These findings reinforce previous research conducted by Wahyuni and Saputra (2023) in Kerinci Regency, Yusdian et al. (2022) in Bandung Regency, Indonesia, Bhatnagar and Subhash (2022) in India, Tudi et al. (2021) and Qiu et al. (2024) in China, all of which explained that insecticides affect potato production. Theoretically, controlling pests as early as possible makes potato plants healthier and more resistant to pests, leading to higher yields (Araújo et al., 2023). However, Kaleem Ullah et al. (2023) in China, it is cautioned that although insecticides increase production, their use should be judicious, as some aphid pests have become resistant to insecticides when overused.

Apart from insecticides, potato farmers use herbicides, which are inorganic materials, to control weeds. Using herbicides reduces crop failure caused by competition between potato plants and weeds for nutrients, water, and growing space. The study shows that the herbicide variable (He) has a negative effect on potato production (PK), with a significance value of 0.048, which is smaller than alpha (α = 0.05). In addition, Figure 4 shows that herbicide (He) has an odds ratio value of 0.333 with an estimate (B) of −1.100. Based on the negative odds ratio and estimate, it can be concluded that an increase in the volume of herbicides used by potato farmers can potentially reduce potato production. In other words, farmers who increase their herbicide use are 1.100 times more likely to reduce potato production than farmers who do not increase their herbicide use, assuming other factors remain constant (ceteris paribus). The results of this study are consistent with those of Brochado et al. (2023) in Brazil and Thornton et al. (2010) in the United States, who both found that overuse of herbicides had a negative impact on potato production. However, these results differ from those of Ginter et al. (2022) in Poland and Abdallah et al. (2021) in Cairo, where they found that herbicide use had a positive effect on potato production. In theory, excessive herbicide use can deplete soil nutrients, resulting in suboptimal potato plant growth and adversely affecting production (Brochado et al., 2023). In this study, herbicides were applied three times: at land clearing, 14 days after planting, and two weeks before harvest.

Watering is a technical step to ensure water availability in potato plants, either done manually or using sprinklers. The results show that the watering variable (Wa) has a real and significant effect on potato production (PK). The significant effect of this variable is indicated by its significance value of 0.019. This figure is less than alpha (α = 0.05). In Figure 4, this variable watering (Wa) has an odds ratio of 1.214 and an estimated value (B) of 0.194. Based on the odds ratio for this variable and the positive estimate, it can be concluded that increasing the frequency of watering by potato farmers can potentially increase potato production. In other words, potato farmers who can increase the watering frequency can increase potato production by 1.214 times compared to farmers who cannot increase the watering frequency in potato crops planted in the dry season, assuming other variables remain constant (ceteris paribus). These findings are in line with previous studies by Jia et al. (2018), (Soothar et al. 2021), (Guo et al. 2023) and Wu et al. (2022) in China, Medina et al. (2024) in Colombia and Khalifa et al. (2020) in Saudi Arabia, all explained that watering affects production. Based on theory, innovation and technology have driven progress in developing water systems that support sustainability and water use efficiency, especially in arid regions (Velasco-Muñoz et al., 2019). In practice, potato farmers in the research location perform watering varies greatly, at intervals 55–60 as many as 112 people, intervals 61–65 as many as 56 people, and intervals 66–70 as many as 55 people, depending on the availability of water sources owned, the ability of farmers and the availability of farmers' time in conducting watering.

In addition, the distance between the house and the garden is critical to ensure easy accessibility so that the garden owner can take care of the plants regularly without spending too much time traveling. The results show that the variable house distance from the farm (DH) negatively affects potato production (PK). The significant effect of this variable is indicated by its significance value of 0.088. This figure is smaller than alpha (α = 0.05). Still, in the exact figure, the distance of the house from the farm (DH) has an odds ratio value of 0.263 with an estimated value (B) of −1.336. Based on the odds ratio value of this variable and its negative estimate, it can be interpreted that an increase in the distance from the farmer's house to the potato farm increases the probability of reducing potato production. Another explanation is that the farther the distance of the house from the garden, the greater the opportunity or possibility of reducing potato production by −1.336 times compared to farmers who have a closer distance from the garden, with the assumption that other variables are fixed (ceteris paribus). These findings are consistent with the research by Maulidiyah et al. (2024) in Bantaeng Regency, Rosalia and Karyani (2020) in Bandung Regency, and Febriansyah et al. (2021) in West Tanjung Regency, Indonesia, and Kassem et al. (2021) in Saudi Arabia, all of whom found that the distance between the house and the farm will reduce production and technical inefficiency. In addition, the findings of Saragih and Harmain (2021) in Simalungun Regency, Indonesia, indicate that accessibility to farmland also affects farmer performance, which is directly related to yields obtained.