Method for evaluating influence of digital literacy on agricultural socialization service adoption of grain farmers

Abstract

The invention provides a method for evaluating the influence of digital literacy on agricultural socialization service adoption of grain farmers, and the method comprises the steps: constructing a grain farmer digital literacy comprehensive evaluation system through employing an entropy method; establishing a variable descriptive statistical framework; constructing a reference regression model; an ordered Probit model, a tendency score matching method and a strategy of removing elderly grain farmer samples are adopted to verify that reference model evaluation has high robustness and reliability; constructing an intermediary effect model, and proposing two key paths of long-term production decision optimization and facultative degree improvement; through heterogeneity analysis, the promotion effect of digital attainment on the adoption behaviors of the grain planting farmers who have accepted agricultural technical training, the middle education level and the low-risk preference is revealed. According to the method, the blank in the field of digital literacy influence evaluation is filled, and a scientific basis and powerful support are provided for formulating related policies for promoting agricultural socialization service popularization and optimization.

Description

Technical Field

[0001] The present invention relates to the field of agriculture, and mainly relates to a method for evaluating the impact of digital literacy on the adoption behavior of agricultural socialized services by grain-growing farmers. Background Art

[0002] For a long time, great importance has been attached to the important functions of agricultural socialized services in alleviating the resource constraints of grain-growing farmers and achieving grain yield increase and income growth, regarding it as the only way to realize the organic connection between small farmers and modern agriculture. However, in the process of agricultural socialized service practice, problems such as small operation scale and fragmented land have weakened the adoption willingness of grain-growing farmers for agricultural socialized services. Existing research has confirmed that the current adoption level of farmers for agricultural socialized services in some links is relatively low. How to improve the adoption degree of grain-growing farmers for agricultural socialized services has become an urgent task for promoting the healthy development of agricultural socialized services in China. In recent years, the digital economy has developed rapidly. While promoting the integration of agricultural socialized services with digital technologies and digital elements, it has significantly improved the digital literacy of rural residents, providing a new possible way for promoting grain-growing farmers to adopt agricultural socialized services. Therefore, it is necessary to explore the relationship between digital literacy and the adoption of agricultural socialized services by grain-growing farmers, so as to effectively promote the adoption level of grain-growing farmers for agricultural socialized services in each link, and expand the application of the behavior theory of grain-growing farmers in the fields of digital economy and agricultural socialized services.

[0003] In recent years, digital literacy has gradually become a hot topic of concern in the industry. However, domestic achievements on the relationship between digital literacy and the adoption behavior of agricultural socialized services by grain-growing farmers are relatively scarce. Scholars such as Ma Ronghui and Hu Nanyan, when analyzing the impact of digital literacy on the behaviors of grain-growing farmers such as green production and food conservation and loss reduction, implied the potential mechanism of the relationship between digital literacy and the adoption behavior of agricultural socialized services by grain-growing farmers. Scholars such as Sun Ziye believe that digital literacy can affect the behaviors of grain-growing farmers such as adopting green technologies and reducing the use of chemical fertilizers and pesticides through channels such as helping grain-growing farmers obtain and understand relevant information and knowledge, expanding credit access, and promoting the implementation of long-term production decisions, which provides guidance for the present invention to propose a method for evaluating the impact of digital literacy on the adoption behavior of agricultural socialized services by grain-growing farmers. Summary of the Invention

[0004] Object of the Invention: Aiming at the problems existing in the above background art, the present invention provides a method for evaluating the impact of digital literacy on the adoption behavior of agricultural socialized services by grain-growing farmers, which not only fills the gap in the field of digital literacy impact evaluation, but also provides a scientific basis and strong support for formulating relevant policies to promote the popularization and optimization of agricultural socialized services.

[0005] To achieve the above object, the technical solution adopted by the present invention is as follows:

[0006] An evaluation method for the impact of digital literacy on the adoption behavior of agricultural socialized services by grain farmers, comprising the following steps:

[0007] Step S1: Select the China Land Economy Survey (CLES) conducted by Nanjing Agricultural University in 2022 as the data source;

[0008] Step S2: Use the entropy method to construct a comprehensive evaluation system for the digital literacy of grain farmers;

[0009] Step S3: Establish a descriptive statistical framework for variables, including explained variables, explanatory variables, control variables, and mechanism variables;

[0010] Step S4: Construct a benchmark regression model to quantify the positive promotion effect of digital literacy on the adoption behavior of agricultural socialized services by grain farmers;

[0011] Step S5: Adopt an ordered Probit model, propensity score matching method (PSM), and the strategy of excluding samples of elderly grain farmers to verify that the estimation of the benchmark model is highly robust and reliable;

[0012] Step S6: Construct a mediating effect model and propose two key paths: optimizing long-term production decisions and enhancing the degree of part-time farming;

[0013] Step S7: Through heterogeneity analysis, reveal the promotion effect of digital literacy on the adoption behavior of grain farmers who have received agricultural technical training, have a secondary education level, and have a low risk preference.

[0014] Furthermore, in step S1, the CLES data covers 16 categories of agricultural socialized services involved in the pre-production (improved seed services, soil testing and fertilization, crop cultivation management), mid-production (pest control technology, mechanized production technology, energy-efficient and high-efficiency facility agriculture technology, water-saving irrigation technology, disaster prevention and mitigation technology), and post-production services (agricultural product processing, packaging, preservation technology, crop straw comprehensive utilization technology, agricultural clean and renewable energy technology, preferential agricultural policies information service, agricultural product market information service, agricultural credit fund service) of the agricultural production process of grain farmers.

[0015] Furthermore, in step S2, the digital literacy of grain farmers reflects the ability of farmers as individuals to safely and effectively acquire, use, communicate, manage, evaluate, and create information or data using digital technologies or through digital devices. It is a form of human capital with digital technology as the carrier and using digital resources. The digital literacy level of grain farmers is measured by indicators such as the accessibility of digital technology applications and the depth of digital technology use, and a comprehensive evaluation system for digital literacy (including multiple key indicators) is constructed based on the entropy method, as shown in Table 1 specifically.

[0016] Table 1 Comprehensive evaluation system table for digital literacy

[0017]

[0018]

[0019] The comprehensive evaluation score of each index of the digital literacy of grain-growing farmers = index value * weight.

[0020] Furthermore, in the step S3, the definitions of the four types of variables are as follows:

[0021] The explained variable is the adoption behavior of agricultural socialized services by grain-growing farmers, which is represented by the adoption degree (ASS) of agricultural socialized services by grain-growing farmers. The proportion of the number of the above 16 types of agricultural socialized service categories adopted by grain-growing farmers is selected to quantify the adoption degree of agricultural socialized services by grain-growing farmers, and it is specifically expressed as:

[0022]

[0023] Among them, ASS i is the adoption degree of agricultural socialized services of a single grain-growing farmer, S ij is the number of types of 16 agricultural socialized services adopted by the i-th grain-growing farmer, and the average weight is counted as

[0024] The core explanatory variable is the digital literacy (DL) of grain-growing farmers.

[0025] The control variables comprehensively consider the relevant characteristics at the levels of the household heads, families, land management, and villages of grain-growing farmers, including variables such as gender (Gender), age (Age), education level (Edu), health status (Health), total number of family members (Rsi), cadre status (Cadre), operation scale (Caa), degree of land fragmentation (Ncp), terrain characteristics (Tf), and traffic characteristics (Vnhe).

[0026] The mechanism variables mainly analyze the influence mechanism of digital literacy on the adoption behavior of agricultural socialized services by grain-growing farmers from two perspectives: long-term production decision-making (Pd) and the degree of farmer sideline occupation (Dpte).

[0027] The specific meanings and assignments of the variables are shown in Table 2.

[0028] Table 2 Descriptive statistics table of relevant variables

[0029]

[0030]

[0031] Furthermore, the benchmark regression model in the step S4 is as follows:

[0032] ASS i = α0 + α1DL i + α2X i + ε i

[0033] where DL i represents the digital literacy of the i-th grain-growing farmer; X i represents the relevant control variables such as the individual characteristics, family characteristics, management characteristics of the grain-growing farmers, and village characteristics variables; α0 is the constant term; α1 and α2 are the parameters to be estimated; ε i is the random error term.

[0034] Furthermore, the model involved in step S5 is as follows:

[0035] Due to the possible endogeneity problem, ordinary least squares regression may have estimation bias. To make the research results more robust, the following ordered Probit model is constructed:

[0036] ASS i = α0 + α1DL i + α2X i + ε i

[0037] Assuming that μ ~ N(0,1) distribution, the ordered Probit model can be expressed as:

[0038] ......

[0040]

[0041] In equation (4), r0, r1…r 15 are the parameters to be estimated; the values of ASS are 0, 1, 2, …, 16, respectively, indicating that the grain-growing farmers "do not adopt" to "adopt 16 kinds of agricultural socialized services", is the cumulative density function of the standard normal distribution. By constructing the likelihood function of the adoption of agricultural socialized services by each grain-growing farmer, the model parameters are estimated by the maximum likelihood method.

[0042] On this basis, the propensity score matching method is used to solve the self-selection problem, that is, through "counterfactual" inference, comparing with the observed group of grain-growing farmers with digital literacy, and then obtaining the average treatment effect (ATT) of consistent estimation. The specific calculation formula of ATT is:

[0043] ATT = E(Y 1i |I i = 1) - E(Y 0i |I i = 1)

[0044] Among them, Y 1i represents the adoption of agricultural socialized services by grain-growing farmers when they have digital literacy, and Y 0i represents the adoption of agricultural socialized services by grain-growing farmers who are assumed to have digital literacy when they actually do not have digital literacy.

[0045] Furthermore, the mediation effect model in step S6 is as follows:

[0046] Since the traditional mediation effect model is difficult to ensure the exogeneity of the mediating variable, and there are problems such as endogeneity bias and low statistical test power, the present invention constructs linear regression equations for the explanatory variable digital literacy (DL) on the mechanism variable long-term production decision (Pd) and the degree of part-time farming of grain-growing farmers (Dpte) respectively:

[0047] Pd i = β0 + β1DL i + β2X i + ε i

[0048] Dpte i = λ0 + λ1DL i + λ2X i + ε i

[0049] Among them, β0 and λ0 are constant terms; β1, β2, λ1, and λ2 are parameters to be estimated; ε i is a random error term.

[0050] Beneficial effects:

[0051] The present invention proposes a method for evaluating the impact of digital literacy on the adoption behavior of agricultural socialized services by grain-growing farmers, analyzes and tests the mechanism and effect of digital literacy on the adoption behavior of agricultural socialized services by farmers, and makes marginal contributions in the following three aspects: First, explore the internal mechanism of the impact of digital literacy on the adoption behavior of agricultural socialized services by farmers, and test whether digital literacy can significantly promote the adoption of agricultural socialized services by grain-growing farmers. Second, from the two perspectives of promoting the implementation of long-term production decisions and improving the degree of part-time farming, test and analyze the impact mechanism of digital literacy on the adoption behavior of agricultural socialized services by grain-growing farmers. Third, based on perspectives such as whether they have received agricultural technical training, educational level, and risk preference, test and analyze the heterogeneity of the impact of digital literacy on the adoption behavior of agricultural socialized services by grain-growing farmers. Description of the drawings

[0052] Figure 1 is a schematic flow chart of a method for evaluating the impact of digital literacy on the adoption behavior of agricultural socialized services by grain-growing farmers according to the present invention. Detailed implementation manners

[0053] The following further describes the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention.

[0054] The method for evaluating the impact of digital literacy on the adoption behavior of agricultural socialized services by grain-growing farmers provided by the present invention includes the following steps:

[0055] Step S1: The research data in this embodiment comes from the China Land Economy Survey (CLES) carried out by Nanjing Agricultural University in 2022;

[0056] In this survey, 12 research counties were randomly selected from 6 prefecture-level cities in Jiangsu Province, 2 towns were randomly selected from each research county, and 1 administrative village was selected from each town. A total of 1203 questionnaires from grain-growing farmers were collected. The CLES data covers information such as the basic family characteristics of grain-growing farmers, the production situation of grain-growing farmers, and rural infrastructure. The present invention finally obtains 1050 valid questionnaires by horizontally merging village questionnaires and questionnaires of grain-growing farmers and deleting outliers.

[0057] The CLES data covers 16 categories of agricultural socialized services involved in the pre-production (improved seed services, soil testing and fertilization, crop cultivation management), mid-production (pest control technology, mechanized production technology, energy-saving and efficient facility agriculture technology, water-saving irrigation technology, disaster prevention and mitigation technology), and post-production services (agricultural product processing, packaging, preservation technology, crop straw comprehensive utilization technology, agricultural clean and renewable energy technology, preferential agricultural policy information service, agricultural product market information service, agricultural credit fund service) of the agricultural production process of grain-growing farmers.

[0058] Step S2: Use the entropy method to construct a comprehensive evaluation system for the digital literacy of grain-growing farmers;

[0059] In this embodiment, the digital literacy of grain-growing farmers reflects the ability of individual grain-growing farmers to safely and effectively acquire, use, communicate, manage, evaluate, and create information or data by adopting digital technologies or through digital devices. It is a form of human capital with digital technologies as the carrier and using digital resources. The digital literacy level of grain-growing farmers is measured by indicators such as the accessibility of digital technology applications and the depth of digital technology use, and a comprehensive evaluation system for digital literacy is constructed based on the entropy method, as shown in Table 1 specifically.

[0060] Table 1 Comprehensive Evaluation System for Digital Literacy

[0061]

[0062] The comprehensive evaluation score of each index of digital literacy of grain-growing farmers = index value * weight.

[0063] Step S3: Construct a detailed descriptive statistical framework for variables, including the explained variable, explanatory variable, control variable, and mechanism variable;

[0064] In this embodiment, the definitions of the four types of variables are as follows:

[0065] The explained variable is the adoption behavior of agricultural socialized services by grain-growing farmers, which is represented by the adoption degree of agricultural socialized services by grain-growing farmers (ASS). The proportion of the number of the above 16 types of agricultural socialized service categories adopted by grain-growing farmers is selected to quantify the adoption degree of agricultural socialized services by grain-growing farmers, and it is specifically expressed as:

[0066]

[0067] Among them, ASS i is the adoption degree of agricultural socialized services of a single grain-growing farmer, S ij is the number of categories of 16 types of agricultural socialized services adopted by the i-th grain-growing farmer, and the average weight is counted as

[0068] The core explanatory variable is the digital literacy (DL) of grain-growing farmers.

[0069] The control variables comprehensively consider the relevant characteristics at the levels of the household head, family, land management, and the village where the grain-growing farmers are located, including variables such as gender (Gender), age (Age), education level (Edu), health status (Health), total number of family members (Rsi), cadre status (Cadre), operation scale (Caa), degree of land fragmentation (Ncp), terrain characteristics (Tf), traffic characteristics (Vnhe), etc.

[0070] The mechanism variables mainly analyze the influence mechanism of digital literacy on the adoption behavior of agricultural socialized services by grain-growing farmers from two perspectives: long-term production decision-making (Pd) and the degree of part-time farming of grain-growing farmers (Dpte).

[0071] The specific meanings and assignments of the variables are shown in Table 2.

[0072] Table 2 Descriptive statistical table of relevant variables

[0073]

[0074] Step S4: Construct a benchmark regression model to quantify the positive promotion effect of digital literacy on the behavior of grain-growing farmers to adopt agricultural socialized services;

[0075] The benchmark regression model in this embodiment is as follows:

[0076] ASS i = α0 + α1DL i + α2X i + ε i

[0077] In Equation (2), DL i represents the digital literacy of the i-th grain-growing farmer; X i represents relevant control variables such as the individual characteristics, family characteristics, business characteristics of grain-growing farmers, and village characteristic variables; α0 is a constant term; α1 and α2 are parameters to be estimated; ε i is a random error term.

[0078] The results of the benchmark regression are shown in Table 3. Whether control variables are included or not, the regression coefficients of the core explanatory variable digital literacy are positive and significant, indicating that digital literacy can promote the adoption of agricultural socialized services by grain-growing farmers. In regression (1) of Table 3, after controlling for the region, the estimated coefficient of the impact of digital literacy on the adoption of agricultural socialized services by grain-growing farmers is 0.0726 and is significant at the 1% level; regression (2) of Table 3 shows that on the basis of further controlling for the individual characteristics, family characteristics, business characteristics, and village characteristic variables of grain-growing farmers, the estimated coefficient of the impact of digital literacy on the adoption of agricultural socialized services by grain-growing farmers is 0.0562 and is significant at the 1% level. A possible explanation is that digital literacy can reduce information asymmetry and transaction costs, broaden the entrepreneurial and employment channels of grain-growing farmers, increase the income and purchasing power of grain-growing farmers, and thus is conducive to grain-growing farmers adopting agricultural socialized services.

[0079] Table 3 Impact of Digital Literacy on the Adoption Behavior of Agricultural Socialized Services

[0080]

[0081] Note: *, **, and *** indicate significance at the 10%, 5%, and 1% levels respectively, and the standard errors are in parentheses. The Chinese names corresponding to the English variables in the table are shown in Table 2.

[0082] Step S5: Use the ordered Probit model, propensity score matching method (PSM), and the strategy of excluding samples of elderly grain-growing farmers to verify that the estimation of the benchmark model is highly robust and reliable;

[0083] Due to possible endogeneity problems, ordinary least squares regression may have estimation biases. To make the research results more robust, the following ordered Probit model is constructed:

[0084] ASS i = α0 + α1DL i + α2X i + ε i

[0085] Assume that μ follows an N(0,1) distribution, then the ordered Probit model can be expressed as:

[0086] ......

[0088]

[0089] where r0, r1…r 15 are parameters to be estimated; the values of ASS are 0, 1, 2, …, 16, representing that the grain-growing farmers "do not adopt" to "adopt 16 kinds of agricultural socialized services" respectively, is the cumulative density function of the standard normal distribution. By constructing the likelihood function of the adoption of agricultural socialized services by each grain-growing farmer, the maximum likelihood method is then used to estimate the model parameters.

[0090] On this basis, the propensity score matching method is used to solve the self-selection problem, that is, through "counterfactual" inference, it is compared with the observed group of grain-growing farmers with digital literacy, and then the average treatment effect (ATT) with consistent estimation is obtained. The specific calculation formula of ATT is:

[0091] ATT = E(Y 1i |I i = 1) - E(Y 0i |I i = 1)

[0092] where Y 1i represents the adoption situation of agricultural socialized services when the grain-growing farmer has digital literacy, and Y 0i represents the adoption situation of agricultural socialized services when it is assumed that the grain-growing farmer without digital literacy has digital literacy.

[0093] The following three methods are used for robustness test and endogeneity analysis. First, using the ordered Probit model, Tables 3(3) and (4) show that whether control variables are included or not, the estimated coefficients of the impact of digital literacy on the adoption of agricultural socialized services by grain-growing farmers are both positive and significant. Second, using the propensity score matching method (PSM), Table 4 shows that digital literacy significantly promotes the adoption of agricultural socialized services by grain-growing farmers, and the estimation results of different matching methods have high consistency. Third, using the sample robustness test, the results of re-regression after excluding the samples of grain-growing farmers aged 75 and above show (see Table 5(2)) that the impact of digital literacy on the adoption behavior of agricultural socialized services by grain-growing farmers is positive and significant. The above tests show that the estimation results of the present invention have a certain degree of robustness.

[0094] Table 4 Estimation Results of the Propensity Score Matching Method

[0095]

[0096] So far, it has been verified that digital literacy can promote the adoption of agricultural socialized services by grain farmers.

[0097] Table 5 Estimation results of robustness test based on sample deletion

[0098]

[0099]

[0100] Note: *, **, and *** indicate significance at the 10%, 5%, and 1% levels respectively, and the standard errors are in parentheses.

[0101] Step S6: Construct a mediating effect model to reveal two key paths of optimizing long-term production decisions and improving the degree of part-time farming;

[0102] In this embodiment, the constructed mediating effect model is as follows:

[0103] Since the traditional mediating effect model is difficult to ensure the exogeneity of the mediating variable, there are problems such as endogeneity bias and low statistical test power. The present invention respectively constructs linear regression equations of the explanatory variable digital literacy (DL) on the mechanism variable long-term production decision (Pd) and the degree of part-time farming of grain farmers (Dpte):

[0104] Pd i = β0 + β1DL i + β2X i + ε i

[0105] Dpte i = λ0 + λ1DL i + λ2X i + ε i

[0106] where β0 and λ0 are constant terms; β1, β2, λ1, and λ2 are parameters to be estimated; ε i is a random error term.

[0107] Table 6 (1) and (2) show that, regardless of whether the control variables are included, digital literacy has a significant positive impact on the implementation of long-term production decisions by grain-growing farmers. After controlling for the region, the regression (1) of Table 6 shows that the estimated coefficient of the impact of digital literacy on the implementation of long-term production decisions by grain-growing farmers is 0.4343, and it is significant at the 1% level; the regression (2) of Table 6 shows that on the basis of further controlling for the individual characteristics, family characteristics, business characteristics and village characteristics of grain-growing farmers, the estimated coefficient of the impact of digital literacy on the implementation of long-term production decisions by grain-growing farmers is 0.2405, and it is significant at the 5% level. The possible reason is that the improvement of digital literacy is conducive to the collection and use of information by grain-growing farmers, helping them to think more comprehensively and make rational long-term decisions. Table 6 (3) and (4) show that, regardless of whether the control variables are included, the impact of digital literacy on improving the degree of concurrent employment of grain-growing farmers is positive and significant at the 1% level. The possible reason is that the improvement of digital literacy has expanded the information channels and social resources of grain farmers, thereby increasing their degree of part-time employment. So far, the two influencing mechanisms proposed in this paper have been preliminarily confirmed.

[0108] Table 6 Mechanism test results

[0109]

[0110] Note: *, **, *** indicate significance at the 10%, 5% and 1% levels respectively, and the values ​​in brackets are standard errors.

[0111] The premise for the establishment of the above mechanism is that the mechanism variables long-term production decision and the degree of part-time employment have an impact on the adoption of agricultural social services. For the sake of robustness, the adoption of agricultural social services is further used as the explained variable and the mechanism variable as the explanatory variable for regression analysis. Table 7(1) shows that the impact of implementing long-term production decisions on the adoption of agricultural social services by grain-growing farmers is positive and significant at the 5% level. A possible explanation is that the implementation of long-term production decisions allows grain-growing farmers to measure the cost-benefit of their production and management activities in a relatively ample time, and to more fully experience the welfare effects of adopting agricultural social services, thereby increasing their adoption of agricultural social services. Table 7(2) shows that the impact of increasing the degree of part-time employment on the adoption of agricultural social services by grain-growing farmers is positive and significant at the 5% level. A possible explanation is that farmers with a high degree of part-time employment have relatively higher non-agricultural income, and considering the opportunity cost and travel cost of going back and forth during the busy farming season, they are more willing to adopt agricultural social services.

[0112] Table 7 Estimation results of the impact of mechanism variables on the adoption of agricultural social services

[0113]

[0114] Note: *, **, and *** indicate significance at the 10%, 5%, and 1% levels respectively, and the standard errors are in parentheses.

[0115] So far, it has been verified that digital literacy can promote the adoption of agricultural socialized services by helping grain farmers implement long-term production decisions, and digital literacy can promote the adoption of agricultural socialized services by improving the degree of part-time farming of farmers.

[0116] Step S7: Through heterogeneity analysis, reveal the promoting effect of digital literacy on the adoption behavior of grain farmers who have received agricultural technical training, have a secondary education level, and have a low risk preference.

[0117] First, the heterogeneity analysis is based on whether grain farmers have received agricultural technical training. Table 8(A) shows that the impact of digital literacy on the adoption behavior of grain farmers who have received agricultural technical training is positive and significant at the 1% level, and the impact on grain farmers who have not received agricultural technical training is positive and significant at the 5% level. The possible reason is that grain farmers who have received agricultural technical training have a stronger awareness of understanding and accepting new technologies and new models in the agricultural field, and thus may be able to more effectively play the promoting role of digital literacy in the adoption of agricultural socialized services by grain farmers.

[0118] Second, the heterogeneity analysis is based on the educational level of grain farmers. Grain farmers are divided into three categories: low educational level (primary school or below), medium educational level (junior high school), and high educational level (senior high school or above), and the control variable of the educational level of grain farmers is removed in the regression. Table 8(B) shows that the impact of digital literacy on the adoption behavior of grain farmers with low and medium educational levels is positive and significant, but the impact on grain farmers with high educational levels is not significant. The possible reason is that the digital technology application ability of grain farmers with low and medium educational levels is insufficient, and the improvement of digital literacy can significantly broaden their information acquisition channels, thus significantly promoting their adoption of agricultural socialized services.

[0119] Table 8 Results of Heterogeneity Test

[0120]

[0121] Note: *, **, and *** indicate significance at the 10%, 5%, and 1% levels respectively, and the standard errors are in parentheses.

[0122] Thirdly, it is the heterogeneity analysis based on the risk preferences of grain-growing farmers. The risk preferences of grain-growing farmers are divided into three categories. Those who tend to choose riskier investments have high risk preferences, those who tend to choose medium-risk investments have medium risk preferences, and those who tend to choose less risky investments have low risk preferences. Table 8(C) shows that digital literacy has a significantly positive impact on the adoption behavior of agricultural socialized services by grain-growing farmers with low risk preferences, but has no significant impact on the adoption behavior of agricultural socialized services by grain-growing farmers with medium and high risk preferences. The possible reason is that farmers with low risk preferences are more inclined to keep agricultural production in their own hands, and the improvement of digital literacy reduces their concerns about transaction costs such as negotiation and supervision that they may face, thereby significantly promoting their adoption of agricultural socialized services.

Claims

1. An evaluation method for the impact of digital literacy on the adoption of agricultural socialized services by grain farmers, characterized in that, It includes the following steps: Step S1: Obtain the data source; Step S2: Use the entropy method to construct a comprehensive evaluation system for the digital literacy of grain-growing farmers; Step S3: Establish a variable descriptive statistical framework, including the explained variable, explanatory variable, control variable, and mechanism variable; Step S4: Construct a benchmark regression model to quantify the positive promotion effect of digital literacy on the behavior of grain-growing farmers in adopting agricultural socialized services; Step S5: Adopt an ordered Probit model, propensity score matching method PSM, and the strategy of excluding samples of elderly grain-growing farmers to verify that the evaluation of the benchmark model is highly robust and reliable; Step S6: Construct a mediating effect model and propose two key paths: optimizing long-term production decisions and enhancing the degree of part-time farming; Step S7: Through heterogeneity analysis, reveal the promotion effect of digital literacy on the adoption behavior of grain-growing farmers who have received agricultural technical training, have a secondary education level, and have a low risk preference.

2. The impact assessment method of digital literacy on the adoption of agricultural socialized services by grain farmers according to claim 1, characterized in that: Select the China Land Economy Survey CLES conducted by Nanjing Agricultural University in 2022 as the data source.

3. The impact assessment method of digital literacy on the adoption of agricultural socialized services by grain farmers according to claim 2, characterized in that: The CLES data in Step S1 cover the categories of agricultural socialized services involved in the agricultural production process of grain-growing farmers, including improved seed services, soil testing and fertilization, crop cultivation management, pest and disease control technology, mechanized production technology, energy-saving and efficient facility agriculture technology, water-saving irrigation technology, disaster prevention and mitigation technology, agricultural product processing, packaging, preservation technology, comprehensive utilization technology of crop straw, agricultural clean renewable energy technology, preferential agricultural policy information services, agricultural product market information services, and agricultural credit fund services.

4. The impact assessment method of digital literacy on the adoption of agricultural socialized services by grain farmers according to claim 1, characterized in that: In Step S2, based on the entropy method, calculate the weights of different indicators and construct a comprehensive evaluation system for the digital literacy of grain-growing farmers, as shown in Table 1: Table 1 Comprehensive Evaluation System Table for Digital Literacy Among them, W j represents the weights of each index of the j-th index, and x' ij represents the j-th index value of the i-th province.

5. An evaluation method for the impact of digital literacy on the adoption of agricultural socialized services by grain farmers according to claim 1, characterized in that: The definitions of the four types of variables in Step S3 are as follows: The explained variable is specifically expressed as: Among them, ASS i is the adoption degree of agricultural socialized services for a single grain farmer, S ij is the number of categories of 16 agricultural socialized services adopted by the i-th grain farmer, and the average weight is counted as The core explanatory variable is the digital literacy DL of grain-growing farmers; The control variables include gender Gender, age Age, education level Edu, health status Health, total household population Rsi, cadre status Cadre, operation scale Caa, degree of land fragmentation Ncp, terrain feature Tf, and traffic feature Vnhe; The mechanism variables include long-term production decision Pd and the degree of part-time farming Dpte of farmers.

6. The impact assessment method of digital literacy on the adoption of agricultural socialized services by grain farmers according to claim 5, characterized in that: The specific meanings and assignments of the variables are shown in Table 2: Table 2 Descriptive Statistical Table of Related Variables 7. An evaluation method for the impact of digital literacy on the adoption of agricultural socialized services by grain farmers according to claim 1, characterized in that: The benchmark regression model in Step S4 is as follows: ASS i = α0 + α1DL i + α2X i + ε i Among them, DL i represents the digital literacy of the i-th grain-growing farmer; X i represents relevant control variables such as the individual characteristics, family characteristics, management characteristics, and village characteristics variables of grain-growing farmers; α0 is a constant term; α1 and α2 are parameters to be estimated; ε i is a random error term.

8. An evaluation method for the impact of digital literacy on the adoption of agricultural socialized services by grain farmers according to claim 1, characterized in that: The models involved in Step S5 are as follows: Construct the following ordered Probit model: ASS i = α0 + α1DL i + α2X i + ε i Assume that μ~N(0,1) distribution, then the ordered Probit model is expressed as: ...... Among them, r0, r1…r 15 are parameters to be estimated; the values of ASS are 0, 1, 2, …, 16, representing that the grain farmers "do not adopt" to "adopt 16 kinds of agricultural socialized services" respectively, is the cumulative density function of the standard normal distribution; by constructing the likelihood function of each grain farmer's adoption of agricultural socialized services, the maximum likelihood method is then used to estimate the model parameters; Through "counterfactual" inference, compare with the observed group of grain-growing farmers with digital literacy, and then obtain the average treatment effect ATT with consistent estimation. The specific calculation formula of ATT is: ATT = E(Y 1i | I i = 1) - E(Y 0i | I i = 1) Among them, Y 1i represents the adoption of agricultural socialized services when grain-growing farmers have digital literacy, and Y 0i represents the adoption of agricultural socialized services when grain-growing farmers who are assumed not to have digital literacy are assumed to have digital literacy.

9. The impact assessment method of digital literacy on the adoption of agricultural socialized services by grain farmers according to claim 1, characterized in that: The mediating effect model in Step S6 is as follows: Construct a linear regression equation of the explanatory variable digital literacy DL on the mechanism variables long-term production decision Pd and the degree of part-time farming Dpte of grain-growing farmers: Pd i = β0 + β1DL i + β2X i + ε i Dpte i = λ0 + λ1DL i + λ2X i + ε i where: β0, λ0 are constant terms; β1, β2, λ1, λ2 are parameters to be estimated; ε i is a random error term.

Images