Influence of Information Literacy on Farmers’ Green Production Technology Adoption Behavior: The Moderating Role of Risk Attitude

Abstract

Green production technology is a critical component of contemporary agricultural development, playing a pivotal role in the promotion of sustainable agricultural practices. Information literacy is the basic ability for farmers to engage in agricultural production, including information awareness, information knowledge, and information ability. In order to investigate the impact of information literacy on farmers’ green production technology adoption behavior, this paper constructs information literacy indicators using factor analysis, based on survey data from 1316 farming households in Shanxi and Hebei provinces, and empirically analyzed the impact of information literacy on farmers’ green production technology adoption behavior and the moderating effect of risk attitude using the Heckman model and moderating effect model, respectively. The empirical evidence indicates the following: (1) Information literacy can significantly contribute to farmers’ adoption decisions and adoption degree of green production technologies. (2) Information literacy has a significant effect on the adoption decisions of all five green production technologies. (3) The results of the heterogeneity analysis indicate that the coefficient of information literacy on the degree of adoption of green production technologies by farmers with different endowment characteristics varies significantly. The promotion effect is more pronounced for small-scale farmers and farmers with a high share of agricultural income. (4) The interaction term between risk attitude and information literacy has a significant effect on farmers’ green production technology adoption degree. Based on these findings, the paper recommends that relevant institutions focus on enhancing farmers’ information literacy, strengthening the agricultural production guarantee system, and developing differentiated strategies for promoting green production technologies.

1. Introduction

The promotion of green production technologies in agriculture is a vital component in advancing the new green development paradigm and fostering sustainable agricultural practices. However, despite rapid socio-economic development, agricultural production in developing countries continues to encounter significant challenges, including low resource utilization [1,2], the inefficient application of chemical fertilizers [3], and the use of unscientific pest control methods [4,5]. These challenges impede the transformation and advancement of agricultural production towards sustainable green development. The Chinese government has attempted to alleviate these challenges by popularizing green production techniques in agriculture. In December 2024, the Ministry of Agriculture and Rural Affairs released the Guiding Opinions on Accelerating the Comprehensive Green Transformation of Agricultural Development and Promoting Rural Ecological Revitalization, which underscores the necessity of advancing agricultural sustainability within the limits of resource availability and environmental capacity. This policy outlines targeted measures to foster the adoption of green production technologies, such as reducing the use of chemical fertilizers and pesticides, promoting straw incorporation, and implementing water-saving irrigation techniques. However, farmers are limited by resource scarcity, technical barriers, information asymmetry, and insufficient risk-resistant ability, and the enthusiasm for green production technology adoption is low [6], resulting in an actual adoption rate of green production technology that is not high [7]. In this context, it is important to explore how to improve the adoption behavior of green production technology among farmers to promote the green and sustainable development of agriculture.

A considerable amount of research has been conducted by academics on the factors that influence farmers’ adoption of green production technologies. This research can be categorized into two main aspects: internal factors, including individual characteristics such as gender, age, political identity, education, health, and farmers’ perceptions [8,9,10]; household characteristics, including farm size, number of farm workers, land fragmentation, share of agricultural income, and participation in e-commerce [11,12,13,14,15]; and farmers’ cognitive characteristics, such as technology cognition, ecological cognition, and market cognition [16,17,18]. Second, external factors, including social networks, participation in cooperatives, government subsidies, green credit, technical training, and other social characteristics, have been shown to significantly influence farmers’ adoption of green production technologies [19,20,21,22].

Information literacy is the ability of farmers to comprehensively utilize information [23], which is integral to the whole process of green production technology adoption by farmers. Enhancing farmers’ information literacy can enable them to alleviate the technical constraints caused by information asymmetry, thus improving their understanding, adoption, and practical application of green production technologies in agriculture. Numerous studies have explored the relationship between information literacy and farmers’ adoption of green production technologies, with varying research emphases. Some scholars have focused on the influence of a single dimension of information literacy on adoption behavior. For instance, empirical evidence suggests a positive correlation between farmers’ information acquisition skills and their adoption of various green production technologies [24,25]. Moreover, some studies have explored the influence of information literacy on the adoption of specific green production technologies, such as conservation tillage [26], fertilizer application [27], and green pest prevention and control technologies [28,29]. However, limited research has examined the impact of comprehensive information literacy indicators on farmers’ adoption behavior, particularly in relation to different types of green production technologies.

Agricultural production is uncertain and risky, and farmers tend to consider the possible risks and benefits when adopting agricultural green production technologies. The effect of information literacy on farmers’ green technology adoption behavior will be constrained to some extent by risk attitude. However, risk attitude as a non-negligible factor influencing farmers’ behavior has received relatively little attention in the discipline of agricultural economics, and few scholars have comprehensively analyzed the effects of information literacy and risk attitude on farmers’ green production technology adoption behavior. Therefore, it is necessary to integrate all three into the same analytical framework for a more comprehensive assessment.

In order to explore the relationship between information literacy, risk attitude, and farmers’ green production technology adoption behavior, this paper constructs an information literacy indicator system for farmers based on the field survey data of 1316 pear farmers in Shanxi and Hebei Provinces, using factor analysis to construct an indicator system for farmers’ information literacy based on the three dimensions of information awareness, information knowledge, and information ability. On this basis, this paper empirically analyzes the impact of information literacy on farmers’ green production technology adoption behavior, explores its differential impact on different types of green production technology and different endowment groups of farmers, and further verifies the moderating effect of risk attitude between information literacy and green production technology adoption behavior.

2. Theoretical Analysis and Research Hypotheses

Farmers’ information literacy is viewed as a multidimensional competency that encompasses awareness, knowledge, and skills [30]. Specifically, it refers to farmers’ ability to recognize the significance of agricultural information in relation to their own needs, proactively acquire and process relevant information, and effectively apply it in agricultural production and management [31]. Information literacy comprises three key dimensions: information awareness, which reflects farmers’ sensitivity to and recognition of information; information knowledge, which pertains to their understanding of information sources and content; and information ability, which encompasses their capacity to access, analyze, and utilize information to enhance decision-making in agricultural activities.

Information awareness constitutes a foundational element of information literacy and can be further subdivided into two key components: information value awareness and information demand awareness. The issue of information awareness is a critical determinant of farmers’ information poverty [32]. First, information value awareness pertains to whether farmers recognize the significance of green agricultural production technology promotion services provided by relevant institutions. Farmers with an enhanced sense of information value are more likely to understand the social function and role of information [33], which increases their possibility of adopting green production technology. Second, farmers with a high level of information demand awareness exhibit greater information sensitivity and initiative. These individuals are not only better able to articulate their information needs but also more proactive in seeking new agricultural information. Such farmers are better equipped to navigate information asymmetry during the technology adoption process [34]. Furthermore, they tend to embrace new ideas and technologies more readily, recognizing the economic and environmental benefits associated with innovative practices. As a result, they are more inclined to make informed decisions that enhance both the adoption and intensity of new technology use.

Information knowledge constitutes the foundation of information literacy, which primarily encompasses theories, knowledge, and methods related to agricultural information. Farmers who possess a certain degree of information knowledge can overcome the barriers to adopting green production technology and enhance the utilization of information resources and the application of new technologies through the mastery of network knowledge and the use of modern communication tools [35].

Information ability is the guarantee of information literacy, which contains information acquisition ability, analysis ability, and application ability. In terms of information acquisition ability, farmers with stronger information acquisition ability have more resources and master information advantage, which is conducive to the rational allocation of agricultural resources, thus enhancing confidence in the adoption of green production technology [36,37]. With regard to the ability to analyze information, farmers, as “rational economic agents,” seek to maximize profits. Independent information analysis ability enables farmers to identify and interpret the information resources they have acquired in order to adopt agricultural green production technologies more effectively [38]. In terms of information application capacity, farmers are limited by their own literacy level and learning ability and usually need the assistance of others to use information effectively. The stronger the information application ability of farmers, the greater the possibility of using green production technology in practice and the stronger the ability to master green production technology [39].

To summarize the above analysis, farmers need to have a certain level of information literacy to understand and absorb green agricultural production technologies. The improvement of farmers’ information literacy can help them to grasp the information of new agricultural green production technologies more sharply, better understand the principles, advantages, and operation methods of these new technologies through effective information screening and learning, and cultivate their innovative consciousness and practical ability. This in turn can help farmers to adopt green production technologies. The following hypotheses are thus proposed in this paper:

Hypothesis H1.

Information literacy and its three dimensions of information awareness, information knowledge, and information competence all positively affect farmers’ green production technology adoption decisions and adoption degree.

Risk attitude has been found to be an important factor influencing farmers’ behavior [40]. Risk attitude has been defined as the tendency that decision makers exhibit when faced with different levels of risk [41]. The effect of information literacy on farmers’ green production technology adoption behavior may vary depending on risk attitude.

Specifically, as information literacy increases, farmers with higher risk attitudes are more likely to take positive steps to change their production behavior [42]. On the one hand, they tend to actively access information resources related to sustainable agricultural development [43] in pursuit of higher returns. On the other hand, their confidence in agricultural investment decisions increases after realizing the specific functions and roles of green production technologies in agriculture, and they may increase their inputs in agricultural production. Therefore, they are more likely to make green production technology adoption decisions and increase the adoption degree of green production technology. In contrast, farmers with lower levels of risk attitudes may stick to their inherent production experiences for fear of technical difficulties and possible economic losses [44,45]. Even if their information literacy is improved, they may not be able to effectively utilize information resources due to concerns about risk, thus hindering the adoption of green production technologies. Accordingly, H2 is proposed:

Hypothesis H2:

Risk attitude plays a moderating role in information literacy affecting farmers’ green production technology adoption behavior.

In summary, this paper integrates information literacy, risk attitude, and farmers’ adoption of green production technology into a unified analytical framework. The theoretical framework is depicted in Figure 1.

3. Materials and Methods

3.1. Data Source

The data presented herein was obtained from a farmer household survey administered by our team in August 2022 in Xi County, Shanxi Province and Wei County, Hebei Province. The pear industry is a typical technology-intensive industry with an obvious need for green production technologies. These two provinces have a long history of large-scale pear cultivation and are important pear production areas in China. Over the years, the governments of the two places have attached great importance to the development of the pear industry and have actively promoted the green development of the pear industry through policy incentives, branding, technical support, and other measures, and farmers have accumulated a wealth of experience in adopting green production technologies. Therefore, it is typical and representative to analyze the green production technology in the sample area.

The research used a combination of multi-stage and random sampling methods, based on the regional economic development level, planting scale, degree of marketization, and other factors to determine the 17 sample townships in two provinces. Each township randomly selected three to five villages, and each village randomly selected 15 to 25 farmers as the subjects of the study. Our team obtained 1338 questionnaires through one-on-one interviews, and after eliminating invalid questionnaires such as missing data and logical contradictions, we got 1316 valid questionnaires, with a validity rate of 98.36%. In order to have an in-depth understanding of the basic situation of the farmers in the study area, this paper analyzes the descriptive statistics of the sample farmers in terms of their individual characteristics, family, and cultivation characteristics, as shown in

Table 1. Results of the descriptive statistical analysis of the sample.

Categorical Variable		Sample Size (Households)	Proportion (%)
Gender	male	1264	96.05%
	female	52	3.95%
Age	(0, 400]	63	4.79%
	(40, 50]	201	15.27%
	(50, 60]	516	39.21%
	(60, 70]	425	32.29%
	>70	111	8.43%
Educational level (years)	0	108	8.21%
	(0, 6]	488	37.08%
	(6, 9]	553	42.02%
	>9	167	12.69%
health status	yes	1201	91.26%
	no	115	8.74%
the proportion of agricultural income	[0, 0.2]	467	35.49%
	(0.2, 0.4]	258	19.60%
	(0.4, 0.6]	222	16.87%
	(0.6, 0.8]	171	12.99%
	(0.8, 1]	198	15.05%
membership in cooperatives	yes	741	56.31%
	no	575	43.69%
cultivation scale (mu)	[0, 10]	708	53.80%
	(10, 20]	308	23.40%
	(20, 30]	158	12.01%
	(30, 40]	47	3.57%
	>40	95	7.22%
soil fertility	yes	1187	90.20%
	no	129	9.80%

.

The individual characteristics of farm households, including gender, age, education level, and health status of household heads are the main focus of the analysis. In terms of gender, the heads of households in the sample area are predominantly male, with 96.05% male heads of households and 3.95% female heads of households, an imbalance in the proportion of male and female heads of households, with men playing a major role in household decision-making. In terms of age, 54.48% of household heads in the sample area are between 40 and 60 years old, and 40.72% are over 60 years old, which shows that there are more middle-aged and old groups in the region, and the aging trend is more serious, which is in line with the current stage of China’s population development. In terms of education, 108 of the sample farmers, or 8.21%, had no education and only 167 had more than 9 years of education, resulting in an overall low level of education that may affect the level of adoption of green production technologies. In terms of health status, about 90% of the household heads were healthy and able to work in agriculture. Analysis of the above data shows that current agricultural producers are characterized by male dominance, a clear trend of aging, and a low overall level of education. When providing technical support and services, relevant organizations should take into account the personality characteristics and learning and comprehension skills of the target group.

In terms of household and cultivation characteristics, the main analyses were the proportion of agricultural income, whether they joined cooperatives, the scale of cultivation, and whether the soil was fertile. From the perspective of agricultural income ratio, 725 families’ agricultural income ratio is less than 40%, accounting for 55.09% of the total sample, and 947 families’ agricultural income ratio is less than 60%, accounting for 71.96% of the total sample, which indicates that most of the families’ dependence on agriculture is relatively low. On the one hand, this may be due to the fact that pear fruits are greatly affected by natural disasters, and farmers will choose to go out and work as laborers in order to diversify the risks. On the other hand, pear fruit cultivation is a long cycle, and farmers need part-time jobs to maintain basic living expenses. From the point of view of joining cooperatives, the proportion of households joining cooperatives is 12.62 percentage points higher than that of households not joining cooperatives, indicating that there is still room for further development of new agricultural management subjects in the sample area. From the point of view of planting scale, 708 households planted pear trees on an area of 10 mu or less, accounting for about 53.8%, and 308 households cultivated pears on 10–20 mu, accounting for 23.4%, indicating that farmers mainly operate on a small scale. This is probably because of the low level of application of modern technology in the research area. Pear fruit planting requires the participation of a large number of laborers, and the problems of an aging labor force limit the expansion of pear fruit planting. From the point of view of soil fertility, 90.20% of the farmers believe that their soil is fertile, reflecting that farmers attach importance to the protection and sustainable use of arable land resources, which is conducive to the promotion of green production technology in wide agriculture.

3.2. Variable Settings

3.2.1. Dependent Variable

The present study explores the adoption decision and degree of adoption of green production technology by farmers. This paper builds upon extant studies [46,47] and integrates the production characteristics of pears to select key technologies—soil testing and formulation, chemical fertilizer reduction, scientific medication, bagging, and green food certification—as the variables characterizing green production technology for farmers from the perspectives of resource conservation and environmental friendliness. The adoption decision regarding green production technology is a binary variable. If farmers adopt one of the technologies, the value is 1; otherwise, the value is 0. Due to the wide variety of green production technologies, the adoption decision is more complex. Equal weighting calculation may lead to an error; therefore, the coefficient of variation method was selected to determine the weight coefficients of the five green production sub-technologies. The degree of adoption of green production technology by farmers is calculated with weighted values.

The coefficient of variation (CV) method is a statistical technique that can objectively assign values to the weights of the indicators. The larger the CV, the more valid information the indicator carries and the higher its importance. Following the calculation process, the weights assigned to the five green production sub-technologies are listed in descending order: green food certification technology (0.273), chemical fertilizer reduction technology (0.254), soil testing and formulation technology (0.237), scientific medication technology (0.166), and bagging technology (0.070).

3.2.2. Core Independent Variables

Information literacy measures a person’s overall ability to handle information based on the three dimensions of information awareness, information knowledge, and information ability. This paper constructs the information literacy index system of farmers on the basis of the previous analysis, derives the factor variance contribution rate and factor loading of each question item through factor analysis, and finally calculates the information literacy score, information awareness score, information knowledge score, and information ability score of farmers in the sample area. The Likert five-point scale, comprising 15 questions, was utilized, with all variables assigned a value of “1 = strongly disagree, 2 = comparatively disagree, 3 = fairly agree, 4 = comparatively agree, or 5 = strongly agree”. The KMO test value of the sample was 0.86, and the Bartlett’s spherical test value was 6384.412 (sig = 0.000), indicating that the correlation between the variables was significant and suitable for factor analysis. The Cronbach’s alpha coefficient was 0.844, indicating good internal consistency and high reliability of the question items. The factor loading coefficients were all greater than 0.6, suggesting that the indicator settings demonstrated good convergent validity. The cumulative method contribution rate extracted using the maximum variance method is 72.045%, and the variance contribution rates of the indicators, in descending order, are as follows: information acquisition ability (14.675%), information value awareness (14.603%), information application ability (13.525%), information analysis ability (10.989%), information demand awareness (9.814%), and information knowledge (8.529%). The design of specific indicators is shown in

Table 2. Factor analysis results.

Primary Variable	Secondary Variable	Item	Mean	SD	Factor Loading
Information awareness	Information value awareness	Agricultural policies, government and pear association guidelines are important for the marketing of pear fruits.	3.764	1.102	0.843
		Guidance on the use of relevant information technology is important in your agricultural production and marketing processes.	3.884	1.034	0.817
		A variety of information is important for the sale of your pears or to improve your standard of living.	3.932	0.942	0.750
	Information demand awareness	Other farmers are using the Internet to sell their pears, and you’ve taken the initiative to learn about it.	3.068	1.302	0.724
		You often need to use the Internet to search for information on the sale of pears.	2.600	1.371	0.799
Information knowledge	Information knowledge	knowledge of safe pesticide application.	3.373	1.074	0.880
		Understanding of policies related to soil environmental protection.	2.589	1.148	0.636
Information ability	Information acquisition ability	You have more access to information when you need it.	2.940	1.102	0.861
		When you encounter difficulties in the sale or production of pears, you can find a good source of counseling.	3.003	1.085	0.819
		You can always get more accurate information on pear sales through various channels.	3.108	1.025	0.736
	Information analysis ability	When you hear negative information, you are usually able to analyze it more calmly or communicate it to others.	3.546	1.032	0.786
		When you learn some very valuable pear production and sales information, you can explore it with more knowledgeable people and make in-depth self-analysis.	3.441	1.086	0.676
	Information application ability	You will adjust your production and sales behavior in a timely manner based on the information you obtain.	3.253	1.111	0.663
		You will independently complete the entire process of selling pears based on the information you obtain.	3.288	1.196	0.841
		You can understand well what the experts or technicians explain about pear fruit production and sales.	3.247	1.127	0.744
Information ability	Information acquisition ability	You have more access to information when you need it.	2.940	1.102	0.861
		When you encounter difficulties in the sale or production of pears, you can find a good source of counseling.	3.003	1.085	0.819
		You can always get more accurate information on pear sales through various channels.	3.108	1.025	0.736
	Information analysis ability	When you hear negative information, you are usually able to analyze it more calmly or communicate it to others.	3.546	1.032	0.786
		When you learn some very valuable pear production and sales information, you can explore it with more knowledgeable people and make in-depth self-analysis.	3.441	1.086	0.676
	Information application ability	You will adjust your production and sales behavior in a timely manner based on the information you obtain.	3.253	1.111	0.663
		You will independently complete the entire process of selling pears based on the information you obtain.	3.288	1.196	0.841
		You can understand well what the experts or technicians explain about pear fruit production and sales.	3.247	1.127	0.744

:

3.2.3. Moderator Variable

Risk attitude. Drawing on the existing literature [48,49], two indicators were selected, with the question item “You are willing to purchase insurance against personal accidents” assigned the value “1 = Never agree, 2 = Disagree, 3 = Neutral, 4 = Agree, or 5 = Strongly agree” and the question item “You are more worried about suffering losses than earning income” assigned the value “1 = Strongly agree, 2 = Agree, 3 = Neutral, 4 = Disagree, or 5 = Never agree”. The farmers’ risk attitude scores are averaged by summing up the scores of the two items using the arithmetic mean method to obtain the risk attitude scores of the farmers. The higher the value of farmers’ risk attitude, the more risk the farmers prefer.

3.2.4. Control Variables

Drawing on existing studies, the individual characteristics of pear farmers and household and planting characteristics were selected as control variables to be included in the model. Among them, individual characteristics included respondents’ gender, age, education level, and health status, and household and cultivation characteristics included the proportion of agricultural income, membership in cooperatives, cultivation scale, and soil fertility. Specific descriptive statistics are shown in

Table 3. Variable definitions and descriptive statistics.

Variables	Variable Name	Variable Description and Assignment (Unit)	Mean	SD
Dependent variable	Adoption decision	At least one green production technology is used in the planting process: yes = 1; no = 0.	0.850	0.358
	Adoption degree	Adoption degree calculated using the coefficient of variation method.	0.281	0.214
Core independent variables	Information literacy	Information literacy is calculated using factor analysis.	8.563	1.643
	Information awareness	Information awareness is calculated using factor analysis.	3.112	0.697
	Information knowledge	Information knowledge is calculated using factor analysis.	0.706	0.215
	Information ability	Information ability is calculated using factor analysis.	4.738	1.087
Moderator variable	Risk attitude	Risk attitudes are calculated on an arithmetic average basis.	2.598	1.028
Control variables	Gender	Gender: male = 1; female = 0.	0.960	0.195
	age	Age: head of household’s age in years.	58.302	9.470
	Educational level	The number of years of formal education received by the household head.	7.247	3.282
	Health status	Health status of the household head: yes = 1; no = 0.	0.913	0.283
	The proportion of agricultural income	Ratio of income from pear cultivation to total household income.	0.403	0.309
	Membership in cooperatives	Whether the household is a member of a cooperative: yes = 1; no = 0.	0.563	0.496
	Cultivation scale	The land area farmed by the household/mu.	19.139	43.626
	Soil fertility	Whether the soil is fertile: yes = 1; no = 0.	0.902	0.297
Identifying variable	Government publicity	Participated in technology promotion activities organized by the government: yes = 1; no = 0.	0.666	0.297

.

3.3. Model Construction

The present study delineates the adoption behavior of green production technology in two distinct components: the technology adoption decision and the technology adoption degree. The technology adoption decision is concerned with whether farmers decide to adopt green production technologies.Farmers can only determine the adoption degree of green production technologies after adopting them, and the level of adoption reflects the acceptance and implementation effect of green production technologies. However, the sample may suffer from selectivity bias because the adoption behavior of farmers does not occur randomly, but is influenced by a variety of complex factors. To overcome this difficulty, the Heckman two-stage model was introduced in this study. This analytical approach considers both the decision-making process of whether or not to adopt green production technologies and explores the actual adoption degree of farmers after adoption, thus revealing more fully the important role of information literacy in the adoption behavior of green production technologies among farmers. Therefore, this paper adopts the Heckman two-stage model to measure and analyze the green production technology. Referring to existing research [50], government publicity is set as an identifying variable. The specific steps are as follows:

The first stage is the selection equation for the green production technology adoption decision, which is regressed using a binary Probit model.

$${\mathrm{y}}_{\mathrm{i}1}={\mathrm{x}}_{\mathrm{i}1}\mathsf{\alpha }+{\mathsf{\mu }}_{\mathrm{i}1}$$

(1)

$${\mathrm{y}}_{\mathrm{i}1}\left\{\begin{array}{c}1, \mathrm{if} {\mathrm{y}}_{\mathrm{i}1}>0\\ 0, \mathrm{if} {\mathrm{y}}_{\mathrm{i}1}\le 0\end{array}\right.$$

(2)

In Equation (1), ${\mathrm{y}}_{\mathrm{i}1}$ denotes the latent variable of green production technology adoption decision of the ith sample of farmers, ${\mathrm{x}}_{\mathrm{i}1}$ denotes the influencing factors affecting the green production technology adoption decision of farmers, $\mathsf{\alpha }$ denotes the coefficient to be estimated, and ${\mathsf{\mu }}_{\mathrm{i}1}$ denotes the error term. In Equation (2), ${\mathrm{y}}_{\mathrm{i}1}$ indicates whether the ith farmer adopts green production technology or not.When ${\mathrm{y}}_{\mathrm{i}1}>0$, it is observed that ${\mathrm{y}}_{\mathrm{i}1}=1$, which means that the farmer adopts green production technology; when ${\mathrm{y}}_{\mathrm{i}1}\le 0$, it is observed that ${\mathrm{y}}_{\mathrm{i}1}=$0, which means that the farmer does not adopt green production technology.

The second stage is the outcome equation for the degree of adoption of green production technologies, which is estimated using a linear regression model. It was when ${\mathrm{y}}_{\mathrm{i}1}=1$ that ${\mathrm{y}}_{\mathrm{i}2}$ could be observed. In addition, the inverse Mills ratio ($\mathsf{\lambda }$) calculated in the first stage was included in the second stage regression as a correction parameter for the second stage.

$${\mathrm{y}}_{\mathrm{i}2}={\mathrm{x}}_{\mathrm{i}2}\mathsf{\beta }+\mathsf{\lambda }\mathsf{\gamma }+{\mathsf{\mu }}_{\mathrm{i}2}$$

(3)

$${\mathrm{y}}_{\mathrm{i}2}\left\{\begin{array}{c}a, \mathrm{if} {\mathrm{y}}_{\mathrm{i}2}=1\\ 0, \mathrm{if} {\mathrm{y}}_{\mathrm{i}2}=0\end{array}\right.$$

(4)

In Equation (3), ${\mathrm{y}}_{\mathrm{i}2}$ denotes the latent variable of the degree of adoption of green production technology of the ith sample of farmers, ${\mathrm{x}}_{\mathrm{i}2}$ denotes the influential factors affecting the degree of adoption of green production technology of farmers, $\mathsf{\lambda }$ denotes the inverse Mills ratio, $\mathsf{\gamma }$ and $\mathsf{\beta }$ denote the coefficients to be estimated, and ${\mathsf{\mu }}_{\mathrm{i}2}$ denotes the error term. If $\mathsf{\gamma }$ passes the significance test, it indicates the presence of sample selectivity bias. In Equation (4), ${\mathrm{y}}_{\mathrm{i}2}$ denotes the degree of adoption of green production technology by the ith farmer, and when ${\mathrm{y}}_{\mathrm{i}2}=1$, ${\mathrm{y}}_{\mathrm{i}2}=\mathrm{a}$ is observed, and the degree of green production technology adoption is a.

4. Results

4.1. Influence of Information Literacy on Green Production Technology Adoption Behavior

The analysis for this study was conducted using Stata 17.0 software. To address potential issues of multicollinearity among the explanatory variables, all variables were tested, and the variance inflation factor (VIF) values were found to be below 1.22, indicating no significant multicollinearity issues. Furthermore, in the Heckman two-stage model, the inverse Mills ratio for information literacy was significant at the 5% level, while the inverse Mills ratios for information awareness, information knowledge, and information competence were all significant at the 1% level. These results suggest the presence of sample selection bias, validating the use of the Heckman two-stage model for analysis. The regression results for the relationship between information literacy and its three dimensions (information awareness, information knowledge, and information competence) and the adoption of green production technologies are presented in Models 1–4 of

Table 4. Regression results of the effect of information literacy on green production technology adoption behavior.

Variables	Model 1		Model 2		Model 3		Model 4
	Adoption Decision	Adoption Degree	Adoption Decision	Adoption Degree	Adoption Decision	Adoption Degree	Adoption Decision	Adoption Degree
Information literacy	0.112 ***	0.037 ***
	(0.035)	(0.004)
Information awareness			0.155 **	0.080 ***
			(0.076)	(0.010)
Information knowledge					0.915 ***	0.158 ***
					(0.253)	(0.033)
Information ability							0.137 ***	0.040 ***
							(0.052)	(0.007)
Gender	0.251	0.013	0.252	0.014	0.218	0.022	0.257	0.016
	(0.266)	(0.031)	(0.266)	(0.031)	(0.272)	(0.033)	(0.265)	(0.032)
Age	−0.003	−0.004	−0.004	−0.001	−0.005	−0.001 *	−0.004	−0.001
	(0.006)	(0.007)	(0.006)	(0.001)	(0.006)	(0.001)	(0.006)	(0.001)
Educational level	0.027*	0.002	0.028 *	0.002	0.031 **	0.003	0.028 *	0.002
	(0.016)	(0.002)	(0.01)	(0.002)	(0.016)	(0.002)	(0.016)	(0.002)
Health status	0.376 **	−0.020	0.396 **	−0.023	0.414 ***	−0.022	0.393 **	−0.023
	(0.159)	(0.025)	(0.159)	(0.025)	(0.158)	(0.026)	(0.159)	(0.025)
The proportion of agricultural income	1.316 ***	−0.029	1.310 ***	−0.037	1.283 ***	−0.056 **	1.303 ***	−0.040
	(0.193)	(0.025)	(0.192)	(0.026)	(0.192)	(0.027)	(0.192)	(0.027)
membership in cooperatives	−1.459 ***	0.006	−1.435 ***	0.019	−1.442 ***	0.028	−1.446 ***	0.020
	(0.166)	(0.020)	(0.163)	(0.021)	(0.162)	(0.021)	(0.163)	(0.021)
Cultivation scale	0.015 ***	0.001 ***	0.016 ***	0.001 ***	0.017 ***	0.001 ***	0.016 ***	0.001 ***
	(0.005)	(0.000)	(0.005)	(0.000)	(0.005)	(0.000)	(0.005)	(0.000)
Soil fertility	−0.617**	−0.033*	−0.650**	−0.024	−0.759 ***	−0.034	−0.602**	−0.022
	(0.256)	(0.020)	(0.259)	(0.020)	(0.261)	(0.022)	(0.255)	(0.021)
Government publicity	0.392 ***		0.419 ***		0.412 ***		0.395 ***
	(0.105)		(0.104)		(0.105)		(0.105)
Constant	0.357	0.021	0.864	0.109	0.868	0.279 ***	0.649	0.166 **
	(0.643)	(0.077)	(0.613)	(0.073)	(0.583)	(0.071)	(0.622)	(0.076)
Mills ratio	−0.109 **		−0.135 ***		−0.201 ***		−0.167 ***
	(0.047)		(0.047)		(0.047)		(0.048)
prob>chi2	0.000		0.000		0.000		0.000

Note: Values in parentheses are standard errors; *, ** and *** indicate significant at 10%, 5%, and 1% significance levels, respectively. Same as below.

.

From the perspective of the core explanatory variables, Model 1 shows that information literacy significantly and positively affects farmers’ green production technology adoption decision and adoption degree at the 1% statistical level, indicating that the improvement of information literacy not only increases the likelihood of farmers’ green production technology adoption, but also improves the adoption degree of farmers’ green production technology. Hypothesis 1 was tentatively supported. Possible explanations are that improved information literacy increases farmers’ sensitivity to information about green production technologies, helps them to access, filter, and analyze information more effectively, and deepens their use of green production technology skills. Ultimately, improved information literacy increases the likelihood that farmers will adopt green production technologies and the extent to which they do so.

In Models 2–4, information knowledge and information ability are positively significant at the 1% level in relation to adoption decisions, while information awareness is positively significant at the 5% level. Regarding the degree of adoption, all three dimensions of information literacy exhibit positive significance at the 1% level. Hypothesis 1 was tested. Possible explanations are that, first, farmers with higher levels of information awareness are more sensitive to available information and are better able to recognize the economic and ecological benefits of green production technologies, which in turn promotes greater demand for information and leads to the adoption of a wider range of green technologies. For example, farmers with a high level of information awareness who are aware of the importance of fertilizer reduction technologies in reducing agricultural input costs and maintaining soil fertility will continue to pay attention to changes in information, keep abreast of the latest agricultural advice and technological advances, and thus adjust their green production technology adoption behavior. Second, farmers with strong information knowledge usually have an in-depth understanding of theories, knowledge, and methods related to agricultural information, which provides a certain knowledge reserve for them to participate in agricultural training, understand national environmental protection policies, and learn about green production technologies, thus lowering the threshold for adopting technologies. In addition, they are more adept at utilizing modern information technology, such as mobile payment systems, social media platforms, and e-commerce skills. Third, taking the scientific pesticide technology as an example, the stronger the information ability of farmers, the more skillfully they can use the Internet, specialized agricultural books, government announcements, agricultural cooperatives, and other channels to obtain more comprehensive pesticide use guidelines, and retrieve different types of pesticides in the use of precautions in the process, the applicable crops, the application method, and other detailed information. After obtaining a large amount of information, they were able to use their own professional knowledge and experience to recognize the authenticity of the information, filter out the most suitable pesticide use information according to their own agricultural production needs, and make the decision to adopt the scientific use of pesticide technology. After experiencing the benefits of scientific pesticide use, farmers will actively adopt other green production technologies.

Regarding control variables, individual farmer characteristics reveal that gender does not significantly affect the decision to adopt green production technologies or the extent of adoption. The negative coefficient for age suggests that older farmers are less likely to adopt green production technologies; however, this result is not statistically significant, likely due to the influence of other factors. The level of education at 10% has a positive and significant effect on the adoption decision because, on the one hand, farmers with a higher level of education master the basic textual knowledge and break through the threshold of technology use; on the other hand, farmers with a higher level of education have a stronger ability to learn new knowledge and analyze new things, and are more likely to practice the concept of environmental protection and make the decision to adopt green production technology. Health status positively influences farmers’ decisions to adopt green production technologies at the 5% statistical level, indicating that the healthier farmers are, the more energy and capacity they have to adopt green production technologies.

In terms of household and crop characteristics, the proportion of agricultural income positively influences farmers’ adoption decision of green production technologies at the 1% statistical level, indicating that the proportion of agricultural income is an important basis for farmers to adopt green production technologies, and that farmers with a high proportion of agricultural income are highly dependent on agricultural activities and are more willing to try new technologies to obtain more profit margins. Joining cooperatives negatively and significantly affects the adoption decision of farmers at the 1% statistical level, probably because the imperfect internal management mechanism of cooperatives and the lack of effective operation strategies do not play a significant role and inhibit the enthusiasm of farmers to adopt green production technologies. Soil fertility negatively affects the adoption decision at the 5% statistical level. This may be because the higher the soil fertility, the lower the input cost of agricultural production, and the lower the demand for green production technology, which inhibits the decision to adopt green production technology. The 1% significant effect of farm size on the adoption decision and the degree of adoption of green production technology may be explained by the fact that, on the one hand, the larger the farm size, the lower the unit cost of adopting green production technology, and the higher the economies of scale of adopting new technology; on the other hand, the larger the farm size, the higher the concern among farmers for the sustainable development of agriculture, and the higher the likelihood that they will try different types of green production technology to improve the efficiency of production.

4.2. The Effect of Information Literacy on Adoption Decisions of Different Types of Green Production Technologies

Given the distinct technical characteristics of various green production technologies, this study further examines the influence of information literacy on the adoption decisions for different types of green production technologies using a binary logit model (

Table 5. Regression results of the effect of information literacy on the adoption decision of different types of green production technology.

Variables	Soil Testing and Formulation Technology	Chemical Fertilizer Reduction Technology	Bagging Technology	Green Food Certification Technology	Scientific Medication Technology
Information literacy	0.329 ***	0.305 ***	0.184 ***	0.576 ***	0.083 *
	(0.052)	(0.055)	(0.048)	(0.064)	(0.043)
Control variables	controlled	controlled	controlled	controlled	controlled
Constant	−4.965 ***	−4.342 ***	−0.760	−6.970 ***	−0.488
	(0.953)	(0.970)	(0.899)	(1.094)	(0.774)
Pseudo R2	0.087	0.084	0.214	0.141	0.101

). The findings indicate that information literacy significantly influences the adoption decisions for soil testing and formulation technology, chemical fertilizer reduction technology, bagging technology, and green food certification technology at the 1% statistical level. Among these, green food certification technology demonstrates the most substantial effect, which may be attributed to its relatively low entry barriers and ease of implementation. Encouraged by local cooperatives and other organizations, farmers are more likely to adopt green food certification technology after recognizing its advantages, including low cost, high economic returns, ease of adoption, and a short return-on-investment cycle. Soil testing and formulation technologies are predominantly offered by institutions such as agricultural suppliers, and farmers with higher information literacy are more proactive in seeking out these services. They are more likely to initiate soil tests and purchase the necessary inputs for agricultural production in a scientifically informed manner, leading to improved land utilization and enhanced ecological protection. Additionally, an increase in information literacy enables farmers to better understand fertilizer usage, facilitating adjustments in their fertilization practices that optimize crop production while minimizing environmental impact. Farmers with higher levels of information literacy are also more likely to adopt bagging technology, recognizing its benefits in improving fruit appearance and quality, and reducing the contamination from airborne pollutants. Furthermore, information literacy significantly impacts the adoption of scientific pesticide use practices, at a 10% statistical level. This may be due to the diverse range of pesticides and the high transmission rates of agricultural pests and diseases, which can be exacerbated by environmental factors. With improved information literacy, farmers are better equipped to analyze and apply pesticide information, promoting more informed and rational decision-making regarding pesticide use.

4.3. Endogenous Analysis

In the context of analyzing the influence of information literacy on farmers’ adoption of green production technologies, potential endogeneity arising from reverse causality must be considered. Specifically, farmers who have already adopted green production technologies may be more inclined to enhance their information literacy. This study addresses this issue by using “the average information literacy of other farmers within the same village” as an instrumental variable for information literacy. The IV-Probit model is employed to examine adoption decisions, while the 2SLS model is utilized to assess the degree of adoption. The regression results are presented in

Table 6. IV-Probit and 2SLS model estimation results.

Variables	IV-Probit	IV-Probit	2SLS	2SLS
	Phase I	Phase II	Phase I	Phase II
	Information Literacy	Adoption Decision	Information	Adoption
			Literacy	Degree
Information literacy		1.321 ***		0.090 ***
		(0.256)		(0.016)
Instrumental variable	0.559 ***		0.557 ***
	(0.063)		(0.076)
Control variables	controlled	controlled	controlled	controlled
Constant	4.318	−10.376 ***	4.581 ***	−0.507 ***
	(0.686)	(2.400)	(0.848)	(0.170)
F-value	40.46 ***		29.11 ***
sample size	1316	1316	1118	1118

.

The first-stage F-values for both models exceed the critical value of 10, and the instrumental variables pass the 1% significance level test. Additionally, the second-stage Wlad’s test and the Durbin-Wu-Hausman (DWH) test both yield significant results at the 1% level, indicating that the instrumental variables are not weak, and that the selection of the instrument is valid. After addressing the potential endogeneity issue, the absolute values of the regression coefficients for information literacy are found to be larger than those obtained in the benchmark regression. This suggests that ignoring the bidirectional causality between information literacy and green production technology adoption would underestimate the true impact of information literacy on adoption behavior. However, this adjustment did not substantially alter the findings of the benchmark regression, thereby reinforcing the reliability of the original results.

4.4. Robustness Tests

4.4.1. Substitution of Variables

Common approaches to measuring the degree of adoption of green production technologies in agriculture include methods such as the coefficient of variation (CV), the simple sum method, and the equal weight method. To mitigate potential errors associated with the CV method in calculating the weights for different green production technologies, this study adopts an alternative measure, quantifying adoption by the number of green production technologies adopted. This adjustment allows for a further examination of the robustness of the relationship between information literacy and the adoption behavior of green production technologies among farm households. Specifically, the number of adopted technologies is determined by summing the five green production sub-technologies utilized by farmers. As shown in Model 5 of

Table 7. Robustness test.

Variables	Model 5		Model 6	Model 7
	Adoption Decision	Adoption Degree	Logit	Tobit
Information literacy	0.112 ***	0.138 ***	0.165 ***	0.042 ***
	(0.035)	(0.020)	(0.062)	(0.004)
Control variables	controlled	controlled	controlled	controlled
constant	0.357	1.055 ***	0.786	−0.107 *
	(0.643)	(0.357)	(1.214)	(0.065)
Mills ratio	−0.813 ***
	(0.211)
prob > chi2	0.000		0.000	0.000

, the results reveal a consistently significant positive effect of information literacy on the adoption of green production technologies, even after modifying the measurement approach. This provides additional evidence supporting the robustness of the original regression findings.

4.4.2. Replacement Regression Model

To further validate the robustness of the Heckman regression results, this study employs the Logit model to reassess the impact of information literacy on the decision to adopt green production technologies and the Tobit model to re-examine the effect of information literacy on the degree of adoption. The results of these alternative analyses are presented in Table 6. A comparison of the estimation outcomes from Model 6 and Model 7 with the Heckman two-stage regression results from Model 1 indicates a high degree of consistency in the direction and significance of the coefficients. This further reinforces the reliability and robustness of the baseline regression model.

4.5. Heterogeneity Analysis

Considering that the green technology adoption behavior of farmers may be differentiated by planting scale and the share of agricultural income in total income, the planting scale is defined using the median, with less than 10 acres as the small-scale group and greater than or equal to 10 acres as the large-scale group. Farmers with agricultural incomes less than 50% of their total income were categorized as the low percentage group, and those with agricultural incomes greater than or equal to 50% were categorized as the high percentage group. The variables were regressed in groups and the likelihood of no correlation test was used to estimate the difference in coefficients between groups. The details are shown in

Table 8. Heterogeneity analysis.

Variables	Adoption Decision		Adoption Degree		Adoption Decision		Adoption Degree
	Limited	Large	Limited Scale	Large	Low Percentage	High Percentage	Low Percentage	High Percentage
	Scale	Scale		Scale
Information literacy	0.079 **	0.071	0.024 ***	0.050 ***	0.043	0.186 ***	0.043 ***	0.042 ***
	(0.038)	(0.086)	(0.006)	(0.008)	(0.041)	(0.068)	(0.006)	(0.009)
Control variables	controlled	controlled	controlled	controlled	controlled	controlled	controlled	controlled
Discrepancy	0.008		−0.026 **		−0.144 **		0.001
Sample size	639	677	639	677	833	483	833	483

.

The between-group coefficient difference test for adoption decision-making did not yield significant results for the small-scale and large-scale groups. However, significant between-group differences were observed in the degree of adoption, with both differences being statistically significant at the 1% level, suggesting that larger-scale farmers experienced a more pronounced facilitation effect. The estimated coefficients for both adoption decision and adoption level were significant for the small-scale group, indicating that small-scale farmers exhibit a greater need and motivation to adopt green agricultural production technologies.

When examining the differences in information literacy’s impact on adoption behavior across various subgroups based on agricultural income share, the test for variance in adoption decision-making passed, while the test for the degree of adoption did not show significant differences across groups. The estimated coefficients for information literacy’s effect on both adoption decision and adoption level were 0.186 and 0.042, respectively, in the high agricultural income share group, both significant at the 1% level. In contrast, in the low agricultural income share group, only the coefficient for the degree of adoption was statistically significant. This suggests that information literacy has a more substantial effect on the green production technology adoption behavior of farmers with a higher share of agricultural income.

4.6. Analysis of the Moderating Effect of Risk Attitude in Information Literacy Influencing the Adoption of Green Production Technologies

In this study, both the core explanatory variables and the moderating variables are treated as continuous variables. An interaction term, representing the product of “information literacy × risk attitude,” is constructed to test the moderating effect of risk attitude using hierarchical regression (

Table 9. Regression results for the moderating effect of risk attitudes.

Variables	Model 8		Model 9
	Adoption Decision	Adoption Degree	Adoption Decision	Adoption Degree
Information literacy	0.100 ***	0.036 ***	0.092 **	0.034 ***
	(0.036)	(0.004)	(0.036)	(0.004)
Risk attitude	0.177 ***	0.017 ***	0.163 ***	0.013 **
	(0.056)	(0.006)	(0.058)	(0.007)
Information literacy × risk attitude			−0.032 (0.033)	0.009** (0.004)
Control variables	controlled	controlled	controlled	controlled
Constant	1.347 **	0.324 ***	1.308 **	0.321 ***
	(0.568)	(0.062)	(0.569)	(0.062)
Mills ratio	−0.086 *		−0.109 **
	(0.047)		(0.048)

). To reduce the potential influence of covariance, the variables are mean-centered before creating the interaction term. The inverse Mills ratios in Models 8 and 9 are found to be statistically significant, indicating the presence of sample selection bias.

A comparison of Model 8 with Model 1 reveals that information literacy continues to have a positive and statistically significant impact on both farmers’ decisions to adopt green production technologies and the extent of their adoption, even after accounting for the moderating effect of risk attitude. The results of Model 9 concerning adoption decisions indicate that the interaction between information literacy and risk attitude does not significantly influence the adoption decision, suggesting that information literacy remains a crucial driver in promoting farmers’ decisions to adopt green production technologies, irrespective of their risk attitude. The regression results for adoption decisions in Model 9 do not support Hypothesis 2. This finding may be attributed to farmers’ high sensitivity to potential losses and their limited risk tolerance. On the one hand, the current agricultural loss protection mechanisms, including agricultural insurance, are insufficient, with compensation often falling short of normal agricultural income levels. On the other hand, the cost of premiums may be prohibitive for some farmers, especially when compounded by limitations related to farm size and other factors.

The results of Model 9 on the level of adoption show that the interaction term between information literacy and risk attitudes is statistically significant at the 5% level with a positive regression coefficient. This suggests that the moderating effect of risk attitude is observed only in the degree of adoption stage, and there exists a complementary relationship between the two variables. The regression results for adoption degree in Model 9 partially support Hypothesis 2. Specifically, the stronger the risk preference of farmers, the more pronounced the effect of information literacy on the degree of adoption of green production technologies. This may be because, as farmers accumulate information about green production technologies and apply it to their agricultural practices, their risk attitudes serve to amplify the positive impact of information literacy on adoption levels. Consequently, promoting the adoption of green production technologies requires not only enhancing farmers’ information literacy but also improving their risk attitudes through policy interventions such as subsidies and agricultural insurance, thereby encouraging a deeper commitment to adopting these technologies.

5. Discussion

In the current global context, where sustainability and environmental protection are of paramount importance, many scholars are focusing on sustainability strategies in agricultural production and natural resource utilization. Brazilian scholars Krein et al. found that the scientific use of fertilizers and pesticides is the key to sustainable agricultural development [51]. By precisely controlling the use of fertilizers and pesticides, not only can crop yields be effectively increased, but also the long-term, stable development of agriculture can be achieved without harming the environment. In addition, Italian scholars Pino et al. put forward an important point of view on water resource management [52]. They pointed out that the unplanned use of groundwater will have serious adverse effects on the natural ecology. In order to deal with this problem, the government should actively encourage farmers to adopt water-saving irrigation technology to reduce the over-exploitation of groundwater. It can be seen that promoting the adoption of green production techniques in agriculture by farmers has become a common concern in many countries in the pursuit of sustainability and environmental protection.

Improving farmers’ information literacy is very important in agricultural production [53], and narrowing the gap of information literacy among farmers can greatly increase their utilization of new technologies [54]. It has been well established that information competence promotes the adoption of green production technologies by farmers [24,25]. Unlike existing studies, our study comprehensively considers the information-related qualities of farmers. We not only constructed an information literacy index system containing information awareness, information knowledge, and information capability through factor analysis, but also systematically assessed the impact of information literacy and its three dimensions on farmers’ green production technology adoption behavior. The results of the study showed that information literacy, information awareness, information knowledge, and information competence can all significantly promote farmers’ green production technology adoption behavior. This result expands the research in this area to a certain degree and makes up for the inadequacy of the previous literature in analyzing farmers’ green production technology adoption behavior from a single dimension of information literacy.

In terms of cooperative participation, our findings differ significantly from other studies. Some studies have found that joining cooperatives does not significantly affect farmers’ behavior [55,56]. Kalogiannidis et al. concluded that farmers’ membership in cooperatives can promote resource sharing and technology integration [57]. However, in our study, joining cooperatives actually reduces farmers’ green production technology adoption decisions, which may be due to the slow development of cooperatives in our survey area, the lack of effective operation strategies, and unregulated management systems that inhibit farmers’ motivation to adopt green production technologies.

In addition, we expanded our study based on the heterogeneity of technology types and farmers’ characteristics. Most of the existing research results regard farmers’ green production technology adoption behaviors as a unified whole [58,59,60], and few studies have combined the differences in technology attributes to conduct in-depth analyses. Our study empirically analyzed the effect of information literacy on different attributes of green production technologies and found that information literacy significantly affected farmers’ adoption of soil testing and formulation technology, weight loss technology, bagging technology, and green food certification technology at the 1% statistical level, and the adoption of five technologies of scientific application technology at the 10% statistical level. In addition, information literacy has a greater impact on the green production technology adoption behavior of small-scale farmers and farmers with a higher percentage of farm income. This implies that information literacy affects farmers’ green production technology adoption decisions differently when the types of green production technologies are different and when farmers’ resource endowments are different.

Some scholars incorporate the influence of cognition [61], willingness [62], and other factors when studying farmers’ behavior, but few scholars have paid attention to the relationship between information literacy, risk attitudes, and farmers’ green production technology adoption behaviors. This paper innovatively incorporates the three into the research framework and finds that risk attitudes can further promote the influence of information literacy on farmers’ green production technology adoption degree.

6. Conclusions and Recommendations

Based on the above analysis, this paper proposes the following policy recommendations:

First, the government should focus on enhancing the information literacy of farmers. To achieve this goal, the government can collaborate with agricultural technology extension departments, village collectives, and agricultural enterprises, combining both online and offline approaches. This collaboration should aim to strengthen farmers’ information and knowledge training, improving their ability to use modern information tools such as mobile payments, online marketing, internet searching, and social media applications. These efforts aim to reduce the difficulty farmers face in adopting green agricultural technologies. Additionally, the government can utilize village collective platforms to promote the importance of agricultural information resources, thereby motivating farmers to improve their information literacy. Moreover, the government should invest in building information infrastructure and create information-sharing platforms to broaden farmers’ access to information and enhance their ability to analyze and apply agricultural data. On this foundation, the government should actively guide farmers to adopt green production technologies and encourage leading farmers as role models, thus promoting the widespread adoption of green agricultural technologies, improving agricultural productivity, and contributing to the sustainability of agriculture.

Second, a sound agricultural production support system should be established to increase farmers’ confidence in agricultural investment. Insurance institutions should expand the scope and types of agricultural production insurance, as well as improve coverage, to alleviate farmers’ risk aversion when adopting new technologies. Furthermore, improving the rural social security system and introducing risk-sharing mechanisms through policy guarantees or agricultural subsidies will help build farmers’ risk resilience and foster a greater tolerance for the risks associated with adopting innovative technologies. The government can encourage cooperatives, enterprises, family farms, and other new agricultural business entities to form closer ties with smallholder farmers, facilitating the sharing of new technological resources. Agricultural enterprises can reduce farmers’ production risks by using strategies such as contract farming and regional branding, which will increase farmers’ risk preference.

Third, the government and agricultural enterprises can develop diversified technical training and differentiated information literacy improvement strategies based on farmers’ preferences for receiving information. For example, encouraging farmers to adopt green food certification and soil testing techniques not only improves product quality and standardizes green production, but also increases the added value of agricultural products, enhancing farmers’ economic benefits. The government should strongly promote the adoption of other green production technologies to facilitate the green transformation of agricultural production methods. For farmers with smaller land holdings or higher agricultural incomes, the government can improve information service platforms to provide timely and accurate market and policy information and offer financial and technical support when necessary to help these farmers make decisions about adopting green production technologies, thereby deepening their adoption. For farmers with larger land holdings, agricultural enterprises can promote large-scale and modern agricultural technologies to help these farmers explore more suitable combinations of green production technologies. For farmers with lower agricultural income shares, agricultural enterprises should prioritize promoting green production technologies that are low in labor intensity, low in risk, and cost-effective, emphasizing the applicability and practicality of the technologies.

This study acknowledges several limitations. First, the research was geographically confined to Shanxi and Hebei provinces, with a specific focus on pear growers. This geographic and crop-specific emphasis may introduce biases that limit the generalizability of the findings to other agricultural sectors or regions. To enhance the broader applicability of the results, future research should aim to include a more diverse and representative sample. Second, while the study examined five types of green production technologies, this selection was relatively narrow and may not capture the full range of green technologies employed in agricultural practices. Expanding the scope to incorporate a broader spectrum of green production technologies could offer a more comprehensive understanding of their application and impact in agriculture. Third, although the study focused on the moderating role of risk attitude, it did not account for other psychological factors influencing farmers’ behavior. Future studies could benefit from exploring a wider array of psychological drivers to more fully understand the complex mechanisms shaping farmers’ adoption behaviors.