Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

The impacts of observational learning and word-of-mouth learning on farmers’ use of biogas in rural Hubei, China: does interpersonal trust play a role?

The impacts of observational learning and word-of-mouth learning on farmers’ use of biogas in... Background: Residue-based biogas is considered as a renewable energy that should be used to improve energy security and household livelihoods in rural areas. Observational learning and word-of-mouth learning are critical in the dissemination of knowledge about agricultural technologies. Yet, scholars have little understanding of the impacts of these learning methods on farmers’ use of residue-based biogas. Using survey data from rural areas of Hubei China, this study estimates the impacts of observational learning and word-of-mouth learning from different subjects (i.e., relatives, neighbors, cadres, cooperative members, and technical instructors) on the use of residue-based biogas by farmers. Additionally, the moderating role of interpersonal trust in these relationships is explored. Results: Results from logistic regression models show that observational learning from technical instructors signifi- cantly increases farmers’ use of biogas. Furthermore, interpersonal trust significantly and positively influences the impact of observational learning on farmers’ decisions to use biogas. Similarly, interpersonal trust significantly and positively moderates the influence of positive word-of-mouth learning on farmers’ decision to use biogas. In contrast, a negative moderating role exists in the relationship between negative word-of-mouth learning and farmers’ decision to use biogas. These impacts are further affirmed by robustness checks. Conclusions: The results presented here show that enhancing farmers’ interpersonal trust promotes the use of resi- due-based biogas by farmers. One important implication is that the government might promote the use of residue- based biogas by organizing technology demonstration activities, providing communication platforms, and enhancing mutual trust between farmers and relevant groups. et  al. [4] revealed that just 29% and 11% of the sampled Background households in Tanzania and Uganda, respectively, use Residue-based biogas is becoming increasingly attrac- biogas exclusively. According to Rahman et  al. [5], only tive as a means to improve energy security and house- 32.5% of the surveyed households adopted the biogas hold livelihoods in rural areas [1–3]. However, it is still technology in 2021 based on data from four districts in underutilized in rural areas around the world, especially rural Bangladesh. Therefore, identifying the obstacles to in developing countries [4, 5]. For example, Clemens biogas technology adoption is a formidable challenge for policymakers. *Correspondence: zhangjb513@126.com Empirical evidence suggests that the lack of rele- College of Economics and Management, Huazhong Agricultural vant information is a major obstacle to the diffusion of University, 1 Shizishan Street, Hongshan, Wuhan 430070, Hubei, China biogas technology in many developing countries [6, 7]. Full list of author information is available at the end of the article © The Author(s) 2022. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. The Creative Commons Public Domain Dedication waiver (http:// creat iveco mmons. org/ publi cdoma in/ zero/1. 0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Zeng et al. Energy, Sustainability and Society (2022) 12:21 Page 2 of 16 Generally, transportation is usually underdeveloped in is confident in the words or actions of another [28]. As an rural areas in developing countries, and people live in essential ingredient of interpersonal ties, interpersonal congested geographical spaces [8, 9]. Accordingly, social trust can influence the extent to which farmers put the learning among farmers has become prominent in the information acquired from others into practice, such as diffusion of agricultural technology [10–13]. For exam - pesticides use [29], agricultural cultivation technology ple, learning from neighbors [10, 11], extension agents adoption [30], and the use of irrigation technology [31]. [12], and “progressive” peers [13] all help in accelerat- To address these research gaps, this study aims at ing the transfer of information and increasing the use of accomplishing two objectives: (1) to examine the impacts technology by farmers. of OB and WOM on farmers’ use of residue-based The existing literature indicates that farmers often learn biogas; and (2) to investigate whether and to what extent through two prominent social learning mechanisms: interpersonal trust can mediate the effects of OB and observational learning (OB) and word-of-mouth learn- WOM. We contribute to the literature of technology ing (WOM) [14–16]. In OB, individuals infer relevant adoption and social learning by providing the first field information from the actions of other people indirectly evidence of the various roles played by OB and WOM through visual observations [14, 17]. In WOM, individu- in the usage of residue-based biogas. We also contrib- als extract relevant information from the opinion of other ute to the trust literature by identifying interpersonal people directly through verbal communication [18]. trust as a mediating factor in the effects of social learn - These two types of learning occur in different ways. OB ings. China is an interesting and relevant case study for occurs when relevant subjects are facing similar shocks empirically illustrating technology adoption concerns. in making decisions, whereas WOM occurs when tempo- In addition to the semi-closed geographical conditions ral, spatial, and social proximity among individuals exists in rural areas, the distinctive Chinese rural acquaintance [14]. Furthermore, WOM can be divided into positive society makes technology information to spread through WOM, meaning that it is explicitly from a subject with observation and communication more conveniently [32]. positive experiences, and negative WOM, which comes Consequently, this study used a unique survey data set from a subject with negative experiences [19]. collected in rural areas of Hubei province in China. The The effects of OB and WOM have been separately outcomes of this study will help us better understand assessed in the extant empirical publications. For exam- adoption decisions and develop strategies that leverage ple, Jones et al. [20] showed that OB from other farmers interpersonal trust to encourage the adoption of new has a causal effect on the uptake of the novel pigeon pea technologies. variety in the semi-arid areas of Mwea Division in the The remainder of this paper is organized as follows: Eastern Province of Kenya. Conley and Udry [11] con- The next section describes the overview of residue-based firmed that farmers in Ghana choose fertilizer by observ - biogas in rural China. “Methods” section presents the ing the inputs and outputs of their neighbors. Based on data and methodology. “Results” section discusses the sample potato farmers in Ecuador, Mauceri et  al. [21] estimation results. “Discussion” section discusses and found that the diffusion of integrated pest management deals with policy implications and also identifies limita - techniques is influenced by WOM. Similarly, Zilberman tions and gives an outlook for future research. “Conclu- et al. [22] suggested that WOM could induce technology sions” section offers the conclusions. adoption. Although OB and WOM may coincide [23], few studies have disentangled the distinct effects of the Overview of residue‑based biogas in rural China two learning mechanisms against the same backdrop. Residue-based biogas is a mixture of methane, carbon We are unaware of any research regarding the role of dioxide, and other gases generated from agricultural interpersonal trust in the investigation of the impacts of residues through anaerobic processes in liquid–state/ OB and WOM. Existing literature suggests that the extent solid–state/liquid–solid two-phase digestions [33, 34]. to which social learning may speed up adoption is closely This mixture is commonly used for cooking, generating related to the connections among individuals [24]. Mean- power, and fueling vehicles among others [35, 36]. ingful social connections create value typically in the In China, residue-based biogas is increasingly gain- form of interpersonal trust that helps individuals to learn ing value, especially as a renewable and clean alternative information from others to improve different behaviors cooking fuel in the rural areas [33]. As the largest devel- [25]. For example, farmers adopting aquaculture technol- oping country in the world, China has an abundance ogies and practices in the Mekong Delta, Vietnam [26], of agricultural residues. According to the Ministry of and behaviors of farmers toward individual and collec- Agriculture and Rural Areas [37], more than 4.7 billion tive measures of controlling the western corn rootworm tons of agricultural residues have been produced annu- [27]. Interpersonal trust is the extent to which a person ally in recent years, with 29.87% of available agricultural Z eng et al. Energy, Sustainability and Society (2022) 12:21 Page 3 of 16 residues used to generate biogas that is mainly consumed questions, modified the language, and added some by farmers according to the National 13th Five-Year Plan interview questions to obtain the final version of the for Rural Biogas Development [38]. questionnaire, which had six parts. To promote the use of residue-based biogas in rural The first part captured the details of infrastructure areas, effective policies such as the Pilot Scheme of Pro - and living conditions in the villages, including the ter- moting the Resource Utilization of Agricultural Wastes rain where the households were located and the closest [37] and Guiding Opinions on the Construction of Straw town or market. The second part entailed personal and Gasification Clean Energy Utilization Project [39] have household characteristics such as gender, age, educa- been implemented in the past decade. These top-down tion, and total household income. The third part cov - national policies not only promote biogas expansion by ered rural biogas utilization such as social learnings, strengthening scientific–technological innovation and interpersonal trust, and the adoption of biogas utiliza- providing service guarantee, but they also set national tion. The fourth part evaluated the extension of agri - short- and long-term targets for achieving sustainable use cultural technologies, whereas the fifth part collected of residue-based biogas and call for efforts and actions information on farmers’ participation in social activi- from provincial governments. Notably, the national and ties and their waste disposals. The final part mainly local governments have already provided financial sup - measured self-happiness of farmers and their percep- port for the use of residue-based biogas. During the 12th tions of different situations. Additional file  1 is part of Five-Year Plan period (2011–2015), 14.2 billion RMB the survey questionnaire translated from Chinese into (equivalent to 2.25 billion USD) were invested in the English. construction of rural biogas projects for farmers [38]. In 2018, investment in small biogas projects for farmers in the sample province Hubei was 86.4 million RMB (equiv- Data collection alent to 12.55 million USD) [40]. Furthermore, each The data used in this study were collected from a house - household in Hubei province that has a biogas digester hold survey conducted in rural areas of Hubei province, has been subsidized with 1000 RMB (equivalent to 132.98 China, in August 2018 using a multistage systematic USD) since 2007 [41]. random sampling procedure. First, four cities were ran- Despite the efforts by the national and provincial gov - domly selected: Huanggang city, Wuhan city, Ezhou city ernments, farmers’ use of biogas in rural China remains and Jingmen city. Second, one to two counties or districts unsatisfactory [42]. It is estimated that the proportion of in each city were randomly selected. The selected coun - biogas farmers consumed in 2014 was only 3.31% [42]. ties or districts cover the main landforms in Hubei prov- Recently, numerous studies have identified predictions of ince, which range from hills to mountains and plain. The residue-based biogas use in rural China [43–46]. The pre - reason for selecting counties or districts at random with dictions include demographic factors (such as age, labor respect to the main landforms was that the use of biogas availability, and total household income), village basic varies with household location in different landforms infrastructure [43], personality traits [45], and energy- [2, 44]. Third, three to five towns in each county or dis - related command and control policies [44, 46]. However, trict were randomly selected. Finally, one village in each no research has investigated the impacts of different town per county or district was randomly selected. In types of social learning. this stage, we obtained a household roster for each village from the local government and randomly selected house- holds for interview. Interviews were conducted by team Methods members from the School of Economics and Manage- Data sources ment, Huazhong Agricultural University. All the mem- Questionnaire design bers had rich experience in rural investigation and were This study was based on surveys conducted in rural professionally trained before the survey was conducted. areas of Hubei province, China. To obtain the data, Supervised face-to-face interviews were conducted with we designed a detailed questionnaire, which was an adult member of the sample households. then modified by the relevant experts in agricultural A total of 1084 observations were collected. The final resource economy from Hubei Rural Development sample for this study comprised 913 observations after Research Center. The experts considerably improved excluding those with missing values. Observations from the logic and language. A pre-test was conducted Huanggang city, Wuhan city, Ezhou city, and Jingmen among 20 farmers in our targeted area of study to city were 163, 328, 203, and 219, respectively, which boost the validity, accuracy, and credibility of our data. accounted for 17.85%, 35.93%, 22.23%, and 23.99%, Based on the pre-test results, we deleted few invalid respectively. Zeng et al. Energy, Sustainability and Society (2022) 12:21 Page 4 of 16 Methodology VIF scores range from 1.04 to 3.24. Campbell et al. [49] Model selection demonstrated that a multicollinearity problem exists In this research, the dependent variable was dichoto- when VIF is greater than 5. Thus, the multicollinearity mous, with 1 indicating that the ith farmer had used res- test result was acceptable in our study. Furthermore, the idue-based biogas (y = 1), and 0 indicating otherwise (y robust standard error procedure was used in this paper in i i = 0). The binary logit model is one of the most widely order to obtain unbiased standard errors under hetero- used statistical models for dealing with the relationship scedasticity. We also used OLS and probit approaches to between dichotomous dependent variable and multiple check the robustness of the binary logistic estimates. continuous or categorical independent variables [47]. According to previous studies, coefficients in non - Therefore, this paper used a binary logistic model to deal linear regressions (e.g., logit) cannot be used to infer with discrete outcomes, assuming that the ith farmer had the statistical significance of an interaction term and its or had not used a distribution Pr(y ). Pr(y ) is believed to underlying variables [50, 51]. The marginal effect of the i i be affected by a vector of key explanatory variables (X ) interaction term varies not only in sign, but also in mag- (e.g., OB, and positive and negative WOM), interpersonal nitude and statistical significance with the values of other trust (I ), and the interaction terms between the learnings explanatory variables [51]. The interpretation of logistic and interpersonal trust (X I ). To avoid potential endoge- model coefficients has certain pitfalls and is less intuitive i i neity due to omitted variable biases, we added other fac- than linear model estimation. Hence, in this paper, the tors to the model as a vector of control variables (Z ) that marginal effects were calculated using the “delta” method may explain the differences in the biogas use probability in Stata. As the relationship between the probability of among the farmers. These factors included demographic social learning and residue-based biogas use is nonlinear, variables, personal perceptions, and geographic loca- the marginal change could be presented by the tangent to tions [34, 48]. Furthermore, it is assumed that Pr(y ) is the probability curve: influenced by an error term μ . The corresponding set of ∂ Pr(y ) ∂ Pr(y ) ∂y i i parameters is {β , β , β , β }, where β is the intercept. It 1 2 3 4 0 = · ∂X ∂y ∂X i i is considered that Pr(y ) is determined by the aforemen- (5) ∗ ∗ tioned factors through a nonlinear link function F that =[�(y )(1 − �(y ))]· (β + β I ) 1 3 i i i maps the unbounded index. Since only dummy variables were involved in the inter- y = β + β X + β I + β X I + β Z + µ 0 1 i 2 i 3 i i 4 i i (1) action terms, the marginal change in this paper was pre- sented as follows: into bounded probability space [0,1]: �y = Pr(y |X, X = 1) − Pr(y |X, X = 0) i i i i (6) Pr(y ) = F (y ) i (2) For the main explanatory variables in this paper, OB where F is the logistic cumulative density function ( ) refers to the process of observing as other individuals that produces the logit model. Thus, use residue-based biogas. Observing others is a behav- ∗ ior-based social interaction that farmers can use as a 1 exp(y ) ∗ ∗ i Pr(y ) = F(y ) = �(y ) = = i reference when making choices from an overwhelming i i ∗ ∗ 1 + exp(y ) 1 + exp(y ) i i number of options. OB may also update personal experi- (3) ence of farmers and beliefs about the profitability of the Then, with (1) incorporated into (3), we get: technology based on the profitable signal from the adop - ∗ ∗ tion behavior of others [52]. Consequently, although the Pr(y ) = F(y ) = �(y ) i i profits and utility of others’ use are unknown, OB may increase the probability of farmers using residue-based 1 + exp[−(β + β X + β I + β X I + β Z + µ ) 0 1 i 2 i 3 i i 4 i i biogas. exp(β + β X + β I + β X I + β Z + µ ) 0 1 i 2 i 3 i i 4 i i Farmers in rural areas rely on various cues, including 1 + exp(β + β X + β I + β X I + β Z + µ ) 0 1 i 2 i 3 i i 4 i i WOM, to acquire technology information in their imme- (4) diate social circles. In this paper, positive WOM refers To investigate potential multicollinearity affecting the explicitly to the process of learning via communication result, we calculated the variance inflation factor (VIF) from the positive experiences of others who have used for the binary logistic model with all variables included in the regression estimation except the interaction terms, As noted by Brambor et  al. [51], any multicollinearity problem cannot be and the calculated results are shown in Appendix 1. The solved by centering the relevant variables. Therefore, the variables in this paper were involved in the interaction terms and were not centered in the logit estimation by subtracting their means. Z eng et al. Energy, Sustainability and Society (2022) 12:21 Page 5 of 16 Fig. 1 Theoretical framework of this paper residue-based biogas. This learning method may inform The relationships between the key independent vari - the beliefs of farmers about the strengths, expected qual- ables, the trust variables, and the dependent variable are ity or profit, and other positive effects of the technology. presented in summary in Fig. 1. It could also increase the exante confidence of the farm - ers in the expected utility of the technology. Therefore, Variables and descriptive statistics positive WOM may increase farmer use rates. On the As aforementioned, the dependent variable was farmers’ contrary, negative WOM refers to the process of learn- use of residue-based biogas, while the key independent ing via communication from the negative experiences of variables were OB, positive and negative WOMs and the farmers’ use of residue-based biogas. This type of learn - interaction terms between the learning types and inter- ing may enhance the perceptions of farmers on the disad- personal trust. Several factors serve as control variables vantages and potential problems of residue-based biogas. to rule out alternative explanations. Gender, age and It may also constitute a noisy signal about the profitability education of the respondents were controlled because of this technology, thus increasing use uncertainty among the characteristics of farmers have been shown to inevi- the farmers. Therefore, negative WOM may reduce the tably correlate with the uptake of biogas technology [6, probability of farmers using the residue-based biogas. 34]. Furthermore, household labor and total household In this paper, interpersonal trust is defined explicitly income were used as control variables because labor as farmers trust the opinion or decisions of others on [43] and family income [48] are all determinants of clean residue-based biogas. According to Sol et  al. [25], inter- energy consumption by households. Sun et al. [44] found personal trust facilitates the relationships between types that biogas subsidy is critical in biogas utilization, hence of social learning and individual behaviors in the face of we included a dummy variable subsidy into the regres- ambiguity and unstructured nature of decision-making sion analysis. Moreover, risk and personal perceptions problems. Embedded within cohesive groups marked served as control variables because they promote biogas with closure, farmers are likely to learn about residue- dissemination [53]. Meeks et  al. [2] reported that biogas based biogas from others. This, in turn, can indicate that projects are generally unsuited for mountain regions due they have trust in other farmers, and the more individuals to temperature requirements to operate them, whereas trust the information provider, the more likely they are Sun et  al. [44] revealed that biogas users are more likely to transform the knowledge obtained from the provider to be located in hilly areas than plains. Accordingly, in into practice. Consequently, the use behaviors of farmers this study, we controlled household location heterogenei- tend to be consistent. Therefore, interpersonal trust may ties to reduce the error in the regression analysis caused mediate the relationships between types of social learn- by landform factor disunity. ing and the use of residue-based biogas. Variable definitions and descriptive statistics are pre - sented in Table  1. According to Table  1, 22% of sur- veyed farm households had used residue-based biogas. Zeng et al. Energy, Sustainability and Society (2022) 12:21 Page 6 of 16 Table 1 Variable definitions and descriptive statistics Variables Definition Mean (S.D.) Adoption 1 if respondent has used residue-based biogas, 0 otherwise 0.22 (0.41) OB from relatives (OB_r) 1 if respondent obtain observational learning from relatives’ adoption of residue- 0.51 (0.50) based biogas, 0 otherwise OB from neighbors (OB_n) 1 if respondent obtain observational learning from neighbors’ adoption of 0.59 (0.49) residue-based biogas, 0 otherwise OB from cadres (OB_c) 1 if respondent obtain observational learning from cadres’ adoption of residue- 0.54 (0.50) based biogas, 0 otherwise OB from cooperative members (OB_o) 1 if respondent obtain observational learning from cooperative members’ adop- 0.31 (0.46) tion of residue-based biogas, 0 otherwise OB from technical instructors (OB_t) 1 if respondent obtain observational learning from technical instructors’ adop- 0.32 (0.47) tion of residue-based biogas, 0 otherwise Positive WOM from relatives (WOM_rp) 1 if respondent obtain positive word-of-mouth learning from relatives, 0 0.24 (0.43) otherwise Positive WOM from neighbors (WOM_np) 1 if respondent obtain positive word-of-mouth learning from neighbors, 0 0.27 (0.44) otherwise Positive WOM from cadres (WOM_cp) 1 if respondent obtain positive word-of-mouth learning from cadres , 0 0.13 (0.34) otherwise Positive WOM from cooperative members (WOM_op) 1 if respondent obtain positive word-of-mouth learning from cooperative 0.04 (0.19) members, 0 otherwise Positive WOM from technical instructors (WOM_tp) 1 if respondent obtain positive word-of-mouth learning from technical instruc- 0.07 (0.25) tors, 0 otherwise Negative WOM from relatives (WOM_rn) 1 if respondent obtain negative word-of-mouth learning from relatives, 0 0.21 (0.40) otherwise Negative WOM from neighbors (WOM_nn) 1 if respondent obtain negative word-of-mouth learning from neighbors, 0 0.22 (0.42) otherwise Negative WOM from cadres (WOM_cn) 1 if respondent obtain negative word-of-mouth learning from cadres , 0 0.11 (0.31) otherwise 0.03 (0.18) Negative WOM from cooperative members (WOM_on) 1 if respondent obtain negative word-of-mouth learning from cooperative members, 0 otherwise Negative WOM from technical instructors (WOM_tn) 1 if respondent obtain negative word-of-mouth learning from technical instruc- 0.09 (0.29) tors, 0 otherwise Trust in relatives (Tr) 1 if respondent trusts relatives, 0 otherwise 0.41 (0.49) Trust in neighbors (Tn) 1 if respondent trusts neighbors, 0 otherwise 0.73 (0.45) 1 b Trust in cadres (Tc) 1 if respondent trusts cadres , 0 otherwise 0.62 (0.49) Trust in cooperative members (To) 1 if respondent trusts cooperative members, 0 otherwise 0.65 (0.47) Trust in technical instructors (Tt) 1 if respondent trusts technical instructors, 0 otherwise 0.58 (0.49) Gender 1 if respondent is male; 0 if female 0.56 (0.50) Age Respondent’s age 57.52 (10.94) Education Schooling of respondents (in years) 6.43 (3.75) Labor Number of individuals in the household that are aged 16 or more but below 3.13 (1.41) 65 years old Household income Total household income in 2017 (10 000 Yuan) 6.96 (7.92) Subsidy 1if subsidy is provided for those who use residue-based biogas 0.23 (0.42) Risk perception Risk perception of the adoption of residue-based biogas 2.39 (1.01) Cost-effective perception Cost-effective perception of the adoption of residue-based biogas 3.95 (0.89) Plain 1 if household is located in plain, 0 otherwise 0.26 (0.44) Hill 1 if household is located in hill, 0 otherwise 0.69 (0.46) Mountains 1 if household is located in mountains, 0 otherwise 0.05 (0.22) N = 913. Following Ziegler [54], if the respondent has high or very high frequency of obtaining positive/negative word-of-mouth learning from these objects, we take value one; otherwise, we take value zero. We take value one if the respondent have high or very high interpersonal trust in these objects, and otherwise, we take c d 1 value zero. Yuan is Chinese currency (1$ = 6.62 Yuan in 2018). related to a 5-point-Likert scale, 1-very low; 5-very high. Cadres refers to local governors who hold certain positions in the village’s political organization, exercise local power, manage local affairs and provide local services, etc Z eng et al. Energy, Sustainability and Society (2022) 12:21 Page 7 of 16 Moreover, farmers who obtained OBs from relatives, OBs from relatives and neighbors and the use behavior neighbors, cadres, cooperative members, and technical all become insignificant when the interaction terms were instructors accounted for 51%, 59%, 54%, 31% and 32%, included. This emphasized the importance of interper - respectively. In addition, 24%, 27%, and 13% of the sample sonal trust in mediating the relationships between OBs farmers obtained positive WOMs from relatives, neigh- and adoption behavior, implying that focusing solely on bors, and cadres, respectively. However, much fewer the estimation results of model 4 is misleading. There - farmers obtain this kind of learning from cooperative fore, the conclusions of the following analyses are mainly members (4%) and technical instructors (7%). Similarly, based on the results of model 5. farmers who obtained negative WOMs from relatives, neighbors and cadres accounted for 21%, 22%, and 11% The impacts of OBs and WOMs on use of biogas by farmers accordingly, which were all much higher than those of The coefficient for OB from technical instructors was negative WOMs from cooperative members (3%) and positive and statistically significant at the 5% level, indi - technical instructors (9%). Table  1 shows that 73% and cating that it significantly influenced the use behavior. 65% of farmers trusted their neighbors and coopera- However, the coefficient of the corresponding interaction tive members, respectively. Farmers who trusted cadres, term was not statistically significant, which suggested that technical instructors, and relatives accounted for 62%, OB from technical instructors significantly and directly 58%, and 41%, respectively. Notably, during the inter- increased the likelihood of the use by farmers. This find - views, interviewers clarified to the respondents that “rel - ing is consistent with the study of Krishnan and Patnam atives”, “neighbors”, “cadres”, “cooperative members”, and [12], which demonstrated that OB from extension agents “technical instructors” refer to different types of generic can effectively predict technology adoption, which is and group concepts rather than specific individuals. probably because technical instructors have professional technical skills. Therefore, obtaining OB from these Results instructors’ adoption of residue-based biogas could be an Main results efficient and reliable source that indicates the expected To better understand the direct influences of OBs and utility of this technology. Therefore, this type of learn - WOMs on farmers’ use of residue-based biogas and the ing greatly increases the probability of farmers using the moderating role of interpersonal trust in these influences, technologies. we constructed 5 models in which key independent vari- ables and interaction terms into the models step by step. The moderating role of interpersonal trust The results of the 5 binary logistic regression estimates Interpersonal trust not only significantly strengthened are presented in Table 2, while all the estimation findings the impact of OB from relatives on the use behavior, but are given in Appendix 2. it also statistically and significantly strengthened the In Table 2, model 1 exclusively tests the effects of inter - relationships between positive WOMs from neighbors, personal trust and control variables on farmers’ use of cooperative members, technical instructors, and the use residue-based biogas. Model 2 adds the key independ- behavior. In contrast, interpersonal trust statistically ent variables to explore the direct effects of OBs and and significantly weakened the links between negative WOMs on farmers’ use of residue-based biogas. Model WOMs from cadres, cooperative members, and the use 3 estimates the interaction between OBs and interper- behavior. As Brambor et  al. [51] suggested, constructing sonal trust, while model 4 tests the interaction between marginal effect plots is an effective way to show how the WOMs and interpersonal trust. Model 5, which was the estimated marginal effect of a variable on the probabil - preferred model, includes all the key explanatory vari- ity varies with another. Furthermore, there is a need to ables, interaction terms, interpersonal trust, and control plot the marginal effect of the interaction terms at the variables. Likelihood-ratio tests, AUC, and AIC results all mean value [51] with all dichotomous variables involved suggested that models 2–5 were significantly better than in the interaction terms. Therefore, following the study of the baseline model 1. When the pseudo-R s, AUC, and Franken et  al. [55], we used figures to graphically depict AIC in the models 1–5 were compared, it was noted that the magnitude and significance of the interaction effects both models 4 and 5 had the most explanatory power. in model 5. The plots are presented in (a) ~ (f ) in Fig. 2. According to model 4, the effects of OB from relatives and neighbors on farmers’ use were statistically signifi - Moderating role of interpersonal trust on the relationships cant, although the corresponding interaction terms were between OBs and adoption excluded. However, in model 5, the positive link between The interaction coefficient of OB from relatives and trust in relatives was positive and statistically significant at The direct effects of different types of interpersonal trust on the use behav - the 5% level. A graph of this moderating effect, which is ior were out of scope of this research. Zeng et al. Energy, Sustainability and Society (2022) 12:21 Page 8 of 16 Table 2 Evaluation statistics of the logistic model Variables Model 1 Model 2 Model 3 Model 4 Model 5 ** ** OB_r 0.64 (0.27) 0.22 (0.30) 0.67 (0.28) 0.26 (0.31) ** ** OB_n 0.83 (0.36) 0.56 (0.67) 0.86 (0.37) 0.81 (0.69) OB_c − 0.39 (0.35) − 0.30 (0.44) − 0.45 (0.37) − 0.40 (0.45) OB_o − 0.22 (0.32) − 0.05 (0.44) − 0.28 (0.33) − 0.05 (0.44) *** ** *** ** OB_t 0.94 (0.33) 0.79 (0.39) 1.09 (0.34) 0.97 (0.40) WOM_rp 0.25 (0.29) 0.23(0.29) − 0.01 (0.36) 0.02 (0.35) *** *** WOM_np 1.14 (0.30) 1.19 (0.31) 0.02 (0.65) 0.06 (0.65) *** *** ** *** WOM_cp 0.99 (0.34) 1.00 (0.33) 1.27 (0.49) 1.29 (0.49) * * WOM_op 1.11 (0.60) 1.08 (0.59) − 0.97 (1.23) − 1.19 (1.29) *** *** WOM_tp − 0.78 (0.58) − 0.73 (0.56) − 2.34 (0.80) − 2.20 (0.79) WOM_rn 0.02 (0.34) 0.04 (0.34) 0.13 (0.38) 0.15 (0.38) *** *** WOM_nn − 1.01 (0.34) − 1.06 (0.34) − 0.25 (0.76) − 0.27 (0.75) ** ** WOM_cn − 0.87 (0.38) − 0.91 (0.38) − 0.76 (0.53) − 0.74 (0.53) *** *** WOM_on 0.72 (0.62) 0.62 (0.62) 3.10 (1.00) 3.12 (1.03) WOM_tn − 0.25 (0.41) − 0.19 (0.41) 0.14 (0.47) 0.17 (0.46) ** ** OB_r × Tr 1.11 (0.44) 1.12 (0.45) OB_n × Tn 0.29 (0.63) 0.03 (0.65) OB_c × Tc 0.03(0.44) 0.07 (0.45) OB_o × To − 0.23 (0.42) − 0.31 (0.43) OB_t × Tt 0.19 (0.40) 0.11 (0.41) WOM_rp × Tr 0.56 (0.60) 0.42 (0.60) ** ** WOM_np × Tn 1.58 (0.70) 1.56 (0.69) WOM_cp × Tc − 0.43 (0.70) − 0.48 (0.68) ** ** WOM_op × To 3.33 (1.43) 3.48 (1.47) *** *** WOM_tp × Tt 3.60 (1.18) 3.36 (1.14) WOM_rn × Tr − 0.42 (0.63) − 0.38 (0.64) WOM_nn × Tn − 1.06 (0.79) − 1.09 (0.78) * * WOM_cn × Tc − 1.47 (0.82) − 1.54 (0.79) *** *** WOM_on × To − 3.22 (1.21) − 3.34 (1.22) WOM_tn × Tt − 1.06 (0.88) − 0.92 (0.85) Interpersonal trust variables Yes Yes Yes Yes Yes Control variables Yes Yes Yes Yes Yes * *** *** *** ** Constant − 1.39 (0.83) − 3.43 (0.89) − 2.78 (1.01) − 3.05 (0.93) − 2.66 (1.06) Log likelihood − 426.72 − 369.44 − 377.00 − 356.81 − 353.11 Pseudo-r 0.11 0.23 0.23 0.25 0.26 Prob > chi 0.00 0.00 0.00 0.00 0.00 *** *** *** *** Likelihood-ratio test 114.57 122.41 139.82 147.23 AUC 0.73 0.82 0.82 0.83 0.84 AIC 885.45 800.87 803.04 795.63 798.21 *** ** * N = 913; p < 0.01, p < 0.05, p < 0.1; standard errors are in parentheses; landform dummies are included in models 1, 2, 3, 4 and 5 with plains as the reference shown in Fig. 2a, allows further investigation. Specifically, As a result, personal ambiguity about the technology can there was a 15% increase in the use probability (from 0.11 be greatly reduced if farmers trusted their relatives. Con- to 0.26) by farmers who trusted and obtained OB from sequently, having trust in relatives may strengthen the relatives. However, for farmers who did not trust their relationship between OB from relatives and biogas use. relatives, the increase in use probability was only 4% Williams [56] noted that when trust is present, positive (from 0.21 to 0.25). One explanation for this phenom - information may facilitate consistent behaviors like tech- enon is that trusting relatives can make the information nology adoption, which requires little time and cognitive obtained via OB more salient, reliable, and persuasive. resources. Z eng et al. Energy, Sustainability and Society (2022) 12:21 Page 9 of 16 (a) Moderating effect of interpersonal truston the (b) Moderating effect of interpersonal truston the relationship between OB from relatives and adoption relationship between positive WOM from neighbors and adoption (c) Moderating effect of interpersonal truston the (d) Moderating effect of interpersonal truston the relationship between positive WOM from relationship between positive WOM from technical instructors and adoption cooperative members and adoption (e) Moderating effect of interpersonal trust on the (f) Moderating effect of interpersonal trust on the relationship between negative WOM from cadres relationship between negative WOM from cooperative members and adoption and adoption Fig. 2 Moderating effect of interpersonal trust on the relationships between WOMs/OBs and adoption Zeng et al. Energy, Sustainability and Society (2022) 12:21 Page 10 of 16 Moderating role of interpersonal trust on the relationships was positive and statistically significant at the 1% level. between WOMs and adoption A graph illustrating this moderating effect is shown in The interaction coefficient of positive WOM from neigh - Fig.  2d. Farmers who trusted technical instructors and bors with trust in neighbors was positive and statistically acquired positive WOM from them had a 16% higher significant at the 5% level (Fig.  2b). Farmers who trusted chance (from 0.21 to 0.37) of using residue-based biogas their neighbors and acquired positive WOM from them compared to those who trusted technical instructors but had a 25% higher probability (from 0.19 to 0.44) of using failed to obtain this type of learning. However, among the residue-based biogas than those who did not trust their farmers who did not trust the technical instructors, those neighbors had only 1% (from 0.11 to 0.12) increase in use who obtained positive WOM from them had an 18% probability, which was in line with the study of Abrams lower probability (from 0.24 to 0.06) of using residue- et al. [57], who stated that interpersonal trust is a driver based biogas than those who did not. Technical instruc- for the knowledge and experience sharing in networks, tors are expected to have a significant degree of technical and makes knowledge exchanges less costly. The trans - information and to be skilled in agricultural technolo- formation of knowledge into actions could be greatly gies. Thus, trusting them can greatly help the transfor - promoted if interpersonal trust exists. Farmers in rural mation of positive WOM from technical instructors to China are closely interconnected with their neighbors, real actions. However, technical instructors’ work for the resulting in familiarity and frequent interaction with government in rural areas [12], and farmers distrust local each other. Therefore, positive WOM from neighbors can government representatives, could exacerbate a negative enhance the perceptions of farmers on the advantages of influence. Therefore, the probability of biogas uses for residue-based biogas and neighbors’ satisfaction with the farmers who do not trust technical instructors yet obtain adoption. Trust in neighbors helps to make these percep- positive WOM from them may be reduced. tions more profound, inducing their transformation into Model 5 indicated that the interaction coefficient action. between the negative WOM from cadres and trust in The interaction coefficient of positive WOM from cadres was negative and statistically significant at the 10% cooperative members with trust in cooperative mem- level. This moderating effect is shown in a graph (Fig.  2e). bers was also positive and statistically significant at the Farmers who trusted the cadres and as well obtained 5% level. This moderating effect is highlighted in Fig.  2c. negative WOM from them had an 18% lower chance Farmers who trusted cooperative members and acquired (from 0.24 to 0.06) of using residue-based biogas than positive WOM from them had a 36% higher chance (from those who trusted the cadres but did not obtain this type 0.18 to 0.54) of using residue-based biogas than those of learning. However, among farmers who did not trust who trusted cooperative members but failed to obtain the cadres, those who obtained negative WOM from the this type of learning. However, among farmers who did cadres had an 8% lower probability (from 0.22 to 0.14) not trust cooperative members, those who obtained posi- of using residue-based biogas than those who failed to tive WOM from cooperative members had a 13% lower obtain this type of learning. This observation could be probability (from 0.26 to 0.13) of using residue-based because village cadres in rural China represent power biogas than those who did not obtain this type of learn- and authority, and are respected by rural farmers [60]. ing. This finding has its particularity and rationality. Trust in cadres can distinctively increase farmers’ nega- Cooperative members in rural China are assumed to have tive perceptions of residue-based biogas if they obtain better access to technical information because the spe- this kind of learning from cadres, resulting in a low prob- cialized cooperative organizations provide their mem- ability of use. bers with excellent technical services [58]. Therefore, Model 5 suggests that the interaction coefficient of farmers who trust these members can have confidence in negative WOM from cooperative members with trust the positive information accumulated by positive WOM in cooperative members is negative and statistically sig- from these members. On the contrary, farmers are not nificant at the 1% level. This moderating effect is shown emotionally close to cooperative members because of in Fig.  2f. Specifically, farmers who trusted cooperative different social identities [59], which could lead to a dis - members and obtained negative WOM from them had trust in cooperative members. Against this backdrop, the a 3% lower probability (from 0.20 to 0.17) of using res- side effect of total distrust in cooperative members may idue-based biogas than those who did not obtain this be greatly strengthened, resulting in a negative impact type of learning. In comparison, among farmers who on the relationship between positive WOM and the use did not trust cooperative members, those who obtained behavior. negative WOM from cooperative members had a 49% The interaction coefficient of positive WOM from higher probability (from 0.24 to 0.73) of using residue- technical instructors with trust in technical instructors based biogas than those who did not obtain this type of Z eng et al. Energy, Sustainability and Society (2022) 12:21 Page 11 of 16 learning. This seemingly contradictory result is not sur - Table 3 Robustness check results with OLS and probit model employed prising for rural China. As stated before, cooperative members are assumed to have good technical informa- Variables OLS Probit tion. Negative WOM may violate farmers’ previous on OB_r 0.04 (0.05) 0.14 (0.18) the utility of this biogas given the asymmetric informa- OB_n 0.02 (0.06) 0.39 (0.33) tion in rural China [61]. Therefore, farmers who trust OB_c − 0.04 (0.06) − 0.20 (0.25) cooperative members have a lower use probability since OB_o 0.02 (0.07) − 0.03 (0.24) cooperative members are not emotionally attentive to ** ** OB_t 0.15 (0.06) 0.55 (0.23) farmers. The lack of trust among farmers in cooperative WOM_rp 0.05 (0.06) 0.04 (0.20) members may have a reverse effect, i.e., the farmers’ WOM_np − 0.00 (0.10) 0.02 (0.35) use probability will increase even though they receive ** ** WOM_cp 0.18 (0.07) 0.71 (0.28) negative WOM from cooperative members. WOM_op − 0.17 (0.21) − 0.69 (0.74) *** *** WOM_tp − 0.28 (0.09) − 1.16 (0.43) WOM_rn − 0.01 (0.06) 0.09 (0.22) Robustness checks WOM_nn − 0.04 (0.10) − 0.19 (0.41) To check the robustness of binary logistic estimates, WOM_cn − 0.11 (0.08) − 0.38 (0.30) OLS and probit methods were employed to exam- ** *** WOM_on 0.41 (0.19) 1.79 (0.60) ine the impact of OBs and WOMs on farmers’ use of WOM_tn 0.02 (0.06) 0.11 (0.26) biogas, and the moderating role of interpersonal trust. ** OB_r × Tr 0.09 (0.05) 0.61 (0.24) The results of the obtained robustness are shown in OB_n × Tn 0.08 (0.05) 0.12 (0.32) Table 3. They indicate that the OB obtained from tech - OB_c × Tc 0.02(0.06) 0.04(0.25) nical instructors significantly influences the use behav - OB_o × To − 0.08(0.07) − 0.17(0.24) ior and that interpersonal trust does moderate the OB_t × Tt − 0.00(0.06) 0.04(0.23) relationships between OBs and use of biogas by farm- WOM_rp × Tr 0.00(0.10) 0.20 (0.34) ers, and between WOMs and use of biogas by farmers. ** ** WOM_np × Tn 0.24 (0.11) 0.94 (0.38) These results further support credibility of the binary WOM_cp × Tc − 0.12(0.10) − 0.24 (0.39) logistic estimates present in Table 2. ** ** WOM_op × To 0.51 (0.23) 1.98 (0.84) *** *** WOM_tp × Tt 0.42 (0.13) 1.85 (0.62) WOM_rn × Tr − 0.04 (0.10) − 0.24 (0.36) Discussion WOM_nn × Tn − 0.14 (0.11) − 0.61 (0.43) As a final discussion, after trying to explore the role of ** WOM_cn × Tc − 0.07 (0.09) − 0.91 (0.45) interpersonal trust in the effects of social learning on * *** WOM_on × To − 0.41 (0.21) − 1.96 (0.72) farmers’ use of residue-based biogas, this paper obtains WOM_tn × Tt − 0.07 (0.09) − 0.54 (0.48) some general and main findings. The general find - Interpersonal trust variables Yes Yes ings are drawn as follows. Analysis of this survey data Control variables Yes Yes reveals that only 22% of the farmers surveyed used ** Constant 0.17 (0.11) − 1.46 (0.58) residue-based biogas, implying that the use rate is rel- Prob > F/ Prob > chi 0.00 0.00 atively low in rural Hubei, China. A large number of farmers receive OB from neighbors. Compared to the *** ** * N = 913; p < 0.01, p < 0.05, p < 0.1; standard errors are in parentheses; landform dummies are included in OLS and Probit estimates with plains as the number of farmers who had access to OB, the number reference of farmers who obtained positive or negative WOMs is much less. Comparatively, interpersonal trust is gen- especially from technical instructors, they are more likely erally high, and more than 70% of farmers trust their to utilize biogas, and the demonstration activities are neighbors. much more effective. Thus, it is necessary for local gov - ernments and their development partners to design and Main findings and policy implications implement strategies aimed at encouraging farmers to First, the empirical results presented here show that OB obtain OB from technical instructors. from technical instructors positively influences farmers’ Second, the results reveal that interpersonal trust use of residue-based biogas. This implies that acquisi - strengthens the relationship between OB from relatives tion of OB from technical instructors can help promote and use behavior by farmers. That is, without farmers’ adoption of residue-based biogas. Farmers may be uncer- trust in relatives, the transfer of technology from relatives tain about the utility and return effects of using residue- through OB may be ineffective. Farmers are generally risk based biogas. However, if they obtain OB from others, Zeng et al. Energy, Sustainability and Society (2022) 12:21 Page 12 of 16 averse [62]. Thus, if they do not trust others, they are less on biogas use from different subjects (i.e., relatives, likely to use residue-based biogas based on knowledge neighbors, cadres, cooperative members and technical acquired via OB. Therefore, implementation of policies instructors). It reveals that OB from technical instruc- meant to increase farmers’ use of residue-based biogas tors positively influences farmers’ use of residue-based based on knowledge acquired through OB, policymakers biogas, and that interpersonal trust strengthens the rela- and practitioners should first pay attention to the inter - tionships between OB from relatives and use behavior by personal trust levels among farmers. Mechanisms aimed farmers, and between positive WOM and use behavior, at promoting mutual trust between farmers and relatives but it weakens the relationship between negative WOM should be developed and implemented. and the use behavior. These findings of this paper con - Third, this empirical study shows that interpersonal trust tribute to our understanding in the moderating role of improves the relationship between positive WOM and use interpersonal trust in the investigation of the impact of behavior. However, it weakens the relationship between social learnings and to enriching the research field of negative WOM and the use behavior. When farmers trust farmers’ use of biogas. others, dissemination of knowledge related to diffuse residue-based biogas via positive WOM becomes more effective. In contrast, when farmers trust others, acquisi - Appendix 1 tion of knowledge about residue-based biogas via negative See Table 4. WOM may fail. Therefore, policy practitioners should be cautious when implementing measures to promote biogas Table 4 Multicollinearity test results adoption through WOMs combined with interpersonal Variables VIF 1/VIF trust. Communication platforms should be built to inspire farmers who trust in others to actively participate in posi- OB_r 2.04 0.49 tive WOM. In addition, policy practitioners should advice OB_n 3.06 0.33 farmers who trust others to critically value the information OB_c 3.24 0.31 obtained from negative WOM. OB_o 2.56 0.39 OB_t 2.59 0.40 WOM_rp 2.43 0.41 Limitations and outlook for future research WOM_np 2.47 0.41 Despite the interesting results presented in this study, we WOM_cp 1.74 0.58 acknowledge that the generalizability of these results to WOM_op 1.83 0.55 the national level should be carefully considered since WOM_tp 1.95 0.51 the study sample was only derived from of rural Hubei, WOM_rn 2.80 0.36 China. Nevertheless, this study provides foundational WOM_nn 2.68 0.37 data that can be used to establish national data. In addi- WOM_cn 1.75 0.57 tion, owing to the lack of time series data, this study WOM_on 1.64 0.61 was based on one-year data which do not capture the WOM_tn 1.65 0.61 dynamic impacts of OB and WOM on technology adop- Tr 1.14 0.87 tion. Therefore, future empirical studies should aim to Tn 1.04 0.96 analyzing panel data to examine these dynamic effects. Tc 1.28 0.78 Furthermore, with data unavailable, we could not include To 1.26 0.79 the potential factors influencing residue-based biogas Tt 1.29 0.78 use such as the volume of agricultural waste production Gender 1.24 0.81 in our study. Future studies may explore the impacts of Age 1.29 0.78 these potential factors on residue-based biogas use with Education 1.46 0.68 relevant data at hand. Labor 1.14 0.88 Household income 1.17 0.86 Conclusions Subsidy 1.11 0.90 This paper is the first to incorporate the moderating role Risk perception 1.19 0.84 of interpersonal trust into the effects of social learning Cost-effective perception 1.16 0.86 on farmers’ use of residue-based biogas. Using data from Hill 1.24 0.81 representative household-based surveys in rural Hubei Mountains 1.24 0.81 China comprising 913 farmers, we empirically exam- N = 913; the VIF scores range from 1.04 to 3.24, and the multicollinearity test ined and distinguished the impacts of OBs and WOMs result was acceptable in our study Z eng et al. Energy, Sustainability and Society (2022) 12:21 Page 13 of 16 Appendix 2 See Table 5. Table 5 Evaluation statistics of the logistic model Variables Model 1 Model 2 Model 3 Model 4 Model 5 ** ** OB_r 0.64 (0.27) 0.22 (0.30) 0.67 (0.28) 0.26 (0.31) ** ** OB_n 0.83 (0.36) 0.56 (0.67) 0.86 (0.37) 0.81 (0.69) OB_c − 0.39 (0.35) − 0.30 (0.44) − 0.45 (0.37) − 0.40 (0.45) OB_o − 0.22 (0.32) − 0.05 (0.44) − 0.28 (0.33) − 0.05 (0.44) *** ** *** ** OB_t 0.94 (0.33) 0.79 (0.39) 1.09 (0.34) 0.97 (0.40) WOM_rp 0.25 (0.29) 0.23(0.29) − 0.01 (0.36) 0.02 (0.35) *** *** WOM_np 1.14 (0.30) 1.19 (0.31) 0.02 (0.65) 0.06 (0.65) *** *** ** *** WOM_cp 0.99 (0.34) 1.00 (0.33) 1.27 (0.49) 1.29 (0.49) * * WOM_op 1.11 (0.60) 1.08 (0.59) − 0.97 (1.23) − 1.19 (1.29) *** *** WOM_tp − 0.78 (0.58) − 0.73 (0.56) − 2.34 (0.80) − 2.20 (0.79) WOM_rn 0.02 (0.34) 0.04 (0.34) 0.13 (0.38) 0.15 (0.38) *** *** WOM_nn − 1.01 (0.34) − 1.06 (0.34) − 0.25 (0.76) − 0.27 (0.75) ** ** WOM_cn − 0.87 (0.38) − 0.91 (0.38) − 0.76 (0.53) − 0.74 (0.53) *** *** WOM_on 0.72 (0.62) 0.62 (0.62) 3.10 (1.00) 3.12 (1.03) WOM_tn − 0.25 (0.41) − 0.19 (0.41) 0.14 (0.47) 0.17 (0.46) ** ** OB_r × Tr 1.11 (0.44) 1.12 (0.45) OB_n × Tn 0.29 (0.63) 0.03 (0.65) OB_c × Tc 0.03 (0.44) 0.07 (0.45) OB_o × To − 0.23 (0.42) − 0.31 (0.43) OB_t × Tt 0.19 (0.40) 0.11 (0.41) WOM_rp × Tr 0.56 (0.60) 0.42 (0.60) ** ** WOM_np × Tn 1.58 (0.70) 1.56 (0.69) WOM_cp × Tc − 0.43 (0.70) − 0.48 (0.68) ** ** WOM_op × To 3.33 (1.43) 3.48 (1.47) *** *** WOM_tp × Tt 3.60 (1.18) 3.36 (1.14) WOM_rn × Tr − 0.42 (0.63) − 0.38 (0.64) WOM_nn × Tn − 1.06 (0.79) − 1.09 (0.78) * * WOM_cn × Tc − 1.47 (0.82) − 1.54 (0.79) *** *** WOM_on × To − 3.22 (1.21) − 3.34 (1.22) WOM_tn × Tt − 1.06 (0.88) − 0.92 (0.85) * *** *** Tr − 0.33 (0.19) − 0.23 (0.20) − 0.98 (0.37) − 0.36 (0.25) − 1.08 (0.31) *** *** * *** * Tn 1.07 (0.22) 1.25 (0.27) 1.02 (0.55) 0.96 (0.31) 0.95 (0.57) Tc − 0.03 (0.19) 0.12 (0.21) 0.11 (0.37) 0.23 (0.24) 0.22 (0.38) To − 0.24 (0.19) − 0.32 (0.21) − 0.24 (0.29) − 0.44 (0.23) − 0.33 (0.30) Tt − 0.16 (0.19) − 0.06 (0.21) − 0.14 (0.281) − 0.19 (0.23) − 0.24 (0.29) Gender − 0.16 (0.19) − 0.10 (0.21) − 0.08 (0.21) − 0.19 (0.21) − 0.17 (0.22) Age − 0.00 (0.01) 0.01 (0.01) 0.00 (0.01) 0.01 (0.01) 0.00 (0.01) *** ** ** ** ** Education 0.08 (0.03) 0.08 (0.03) 0.07 (0.03) 0.09 (0.03) 0.08 (0.03) Labor 0.04 (0.06) 0.06 (0.07) 0.06 (0.07) 0.07 (0.07) 0.07 (0.07) Household income − 0.00 (0.01) − 0.01 (0.01) − 0.01 (0.01) − 0.01 (0.01) − 0.01 (0.01) ** Subsidy 0.44 (0.19) 0.25 (0.22) 0.29 (0.22) 0.23 (0.23) 0.27 (0.23) *** *** *** *** *** Risk perception − 0.46 (0.09) − 0.29 (0.10) − 0.29 (0.10) − 0.31 (0.10) − 0.32 (0.10) Cost-effective perception 0.18 (0.10) 0.09 (0.11) 0.08 (0.12) 0.09 (0.12) 0.09 (0.12) *** *** *** *** *** Hill − 0.69 (0.19) − 0.79 (0.22) − 0.76 (0.22) − 0.789 (0.23) − 0.79 (0.23) Mountains − 0.60 (0.40) − 0.57 (0.49) 0.64 (0.49) − 0.65 (0.53) − 0.72 (0.52) * *** *** *** ** Constant − 1.39 (0.83) − 3.43 (0.89) − 2.78 (1.01) − 3.05 (0.93) − 2.66 (1.06) Log likelihood − 426.72 − 369.44 − 377.00 − 356.81 − 353.11 Zeng et al. Energy, Sustainability and Society (2022) 12:21 Page 14 of 16 Table 5 (continued) Variables Model 1 Model 2 Model 3 Model 4 Model 5 Pseudo-r 0.11 0.23 0.23 0.25 0.26 Prob > chi 0.00 0.00 0.00 0.00 0.00 *** *** *** *** Likelihood-ratio test 114.57 122.41 139.82 147.23 AUC 0.73 0.82 0.82 0.83 0.84 AIC 885.45 800.87 803.04 795.63 798.21 *** ** * N = 913; p < 0.01, p < 0.05, p < 0.1; standard errors are in parentheses; terrain dummies are included in all models with plains as the reference Received: 28 May 2021 Accepted: 5 May 2022 Supplementary Information The online version contains supplementary material available at https:// doi. org/ 10. 1186/ s13705- 022- 00350-8. Additional file 1. Part of the survey questionnaire. References 1. Kabyanga M, Balana BB, Mugisha J, Walekhwa PN, Smith J, Glenk K (2018) Are smallholder farmers willing to pay for a flexible balloon biogas Acknowledgements digester? Evidence from a case study in Uganda. Energy Sustain Dev The authors gratefully acknowledge financial support from the National 43:123–129. https:// doi. org/ 10. 1016/j. esd. 2018. 01. 008 Natural Science Foundation of China (72103055). This paper is also supported 2. Meeks R, Sims KRE, Thompson H (2019) Waste not: can household biogas by the scholarship granted by the China Scholarship Council (CSC) and by deliver sustainable development? Environ Resour Econ 72(3):763–794. the General Project of Hubei Social Science Foundation, China (2021191). The https:// doi. org/ 10. 1007/ s10640- 018- 0224-1 authors are very grateful to the editors and the anonymous reviewers of this 3. Kumar JCR, Majid MA (2020) Renewable energy for sustainable develop- journal who provided very valuable feedback on this paper. In addition, the ment in India: current status, future prospects, challenges, employment, authors thank the Home for Researchers Editorial Team for providing language and investment opportunities. Energy Sustain Soc 10:2. https:// doi. org/ editing service, which improved the readability and clarity of this manuscript. 10. 1186/ s13705- 019- 0232-1 4. Clemens H, Bailis R, Nyambane A, Ndung’u V (2018) Africa biogas partner- Author contributions ship program: a review of clean cooking implementation through market YZ was the main author responsible for the design, data collection, analysis, development in East Africa. Energy Sustain Dev 46:23–31. https:// doi. org/ and interpretation of the results of this study. FQ complemented the discus- 10. 1016/j. esd. 2018. 05. 012 sion related to the cultural theory and provided relevant feedback for the 5. Rahman MS, Majumder MK, Sujan MHK (2021) Adoption determinants of improvement of the article’s clarity. JZ complemented the design and data biogas and its impact on poverty in Bangladesh. Energy Rep 7:5026– collection. All authors read and approved the final manuscript. 5033. https:// doi. org/ 10. 1016/j. egyr. 2021. 08. 027 6. Berhe M, Hoag D, Tesfay G, Keske C (2017) Factors influencing the adop - Funding tion of biogas digesters in rural Ethiopia. Energy Sustain Soc 7:10. https:// This study is financially support by the National Natural Science Foundation doi. org/ 10. 1186/ s13705- 017- 0112-5 of China (72103055), by the scholarship granted by the China Scholarship 7. Pestalozz J, Bieling C, Scheer D, Kropp C (2019) Integrating power-to-gas Council (CSC) and by the General Project of Hubei Social Science Foundation, in the biogas value chain: analysis of stakeholder perception and risk gov- China (2021191). ernance requirements. Energy Sustain Soc 9:38. https:// doi. org/ 10. 1186/ s13705- 019- 0220-5 Availability of data and materials 8. Fletschner D, Mesbah D (2011) Gender disparity in access to information: The datasets generated and analyzed in this paper are not publicly available do spouses share what they know? World Dev 39(8):1422–1433. https:// due to confidentiality reasons. However, the transcripts of the anonymous doi. org/ 10. 1016/j. world dev. 2010. 12. 014 interviews are available on reasonable request. 9. Kashif M, Awan MB, Nawaz S, Amjad M, Talib B, Farooq M, Nizami AS, Rehan M (2020) Untapped renewable energy potential of crop residues in Pakistan: challenges and future directions. J Environ Manag Declarations 256:109924. https:// doi. org/ 10. 1016/j. jenvm an. 2019. 109924 10. Munshi K (2004) Social learning in a heterogeneous population: technol- Ethics approval and consent to participate ogy diffusion in the Indian Green Revolution. J Dev Econ 73:185–215. Not applicable. https:// doi. org/ 10. 1016/j. jdeve co. 2003. 03. 003 11. Conley TG, Udry CR (2010) Learning about a new technology: pineapple Consent for publication in Ghana. Am Econ Rev 100(1):35–69. http://www.aeaweb.org/articles. Not applicable. php?doi=https:// doi. org/ 10. 1257/ aer. 100.1. 35 12. Krishnan P, Patnam M (2013) Neighbors and extension agents in Competing interests Ethiopia: who matters more for technology adoption? Am J Agric Econ The authors declare that they have no competing interests. 96(1):308–327. https:// doi. org/ 10. 1093/ ajae/ aat017 13. Maertens A (2017) Who cares what others think (or do)? Social learn- Author details 1 ing and social pressures in cotton farming in India. Am J Agric Econ School of Economics and Management, Hubei University of Technol- 2 99(4):988–1007. https:// doi. org/ 10. 1093/ ajae/ aaw098 ogy, 28 Nanlilu Street, Hongshan, Wuhan 430068, Hubei, China. F acult y 14. Cai H, Chen Y, Fang H (2009) Observational learning: evidence from a ran- of Agricultural, Life and Environmental Science, University of Alberta, General 3 domized natural field experiment. Am Econ Rev 99(3):864–882. https:// Services Building, 9007 116 St NW, Edmonton, AB T6G 2H1, Canada. College doi. org/ 10. 1257/ aer. 99.3. 864 of Economics and Management, Huazhong Agricultural University, 1 Shizishan Street, Hongshan, Wuhan 430070, Hubei, China. Z eng et al. Energy, Sustainability and Society (2022) 12:21 Page 15 of 16 15. Campbell A (2013) Word-of-mouth communication and percolation in Tigray, Ethiopia. Energy Econ 66:272–278. https:// doi. org/ 10. 1016/j. eneco. social networks. Am Econ Rev 103(6):2466–2498. https:// doi. org/ 10. 1257/ 2017. 06. 022 aer. 103.6. 2466 35. Reise C, Musshoff O, Granoszewsk K, Spiller A (2012) Which factors influ- 16. Ameri M, Honka E, Xie Y (2016) Word-of-mouth, observational learning, ence the expansion of bioenergy? An empirical study of the investment and product adoption: evidence from an anime platform. SSRN Electron behaviours of German farmers. Ecol Econ 73:133–141. https:// doi. org/ 10. J. https:// doi. org/ 10. 2139/ ssrn. 27989 621016/j. ecole con. 2011. 10. 008 17. Bandura A (1977) Social learning theory. Prentice Hall, Englewood Cliffs, 36. Muvhiiwa R, Hildebrandt D, Chimwani N, Ngubevana L, Matambo T NJ (2017) The impact and challenges of sustainable biogas implementation: 18. Berger J (2014) Word of mouth and interpersonal communication: a moving towards a bio-based economy. Energy Sustain Soc 7:20. https:// review and directions for future research. J Consum Psychol 24(4):586–doi. org/ 10. 1186/ s13705- 017- 0122-3 607. https:// doi. org/ 10. 1016/j. jcps. 2014. 05. 002 37. Ministry of Agriculture and Rural Areas of the People’s Republic of China. 19. Wang Y, Yu C (2017) Social interaction-based consumer decision-making 2016-08-11. Pilot Scheme of Promoting the Resource Utilization of Agri- model in social commerce: the role of word of mouth and observational cultural Wastes. http:// www. moa. gov. cn/ govpu blic/ FZJHS/ 201609/ t2016 learning. Int J Inf Manag 37:179–189. https:// doi. org/ 10. 1016/j. ijinf omgt. 0919_ 52778 46. htm 2015. 11. 005 38. National Development and Reform Commission. 2017a-01-25. National 20. Jones R, Audi PA, Tripp R (2001) The role of informal seed systems in dis- 13th five-year plan for rural biogas development. https:// www. ndrc. gov. seminating modern varieties: the example of pigeonpea from a semi-arid cn/ fzggw/ jgsj/ njs/ sjdt/ 201702/ t2017 0210_ 11948 08. html area of Kenya. Exp Agric 37:539–548. https:// doi. org/ 10. 1017/ s0014 47970 39. National Development and Reform Commission. 2017b-12–28. Guiding 10004 61 Opinions on the Construction of Straw Gasification Clean Energy Utiliza- 21. Mauceri M, Alwang J, Norton G, Barrera VH (2007) Eec ff tiveness of tion Project. https:// www. ndrc. gov. cn/ xxgk/ zcfb/ tz/ 201801/ t2018 0102_ integrated pest management dissemination techniques: a case study 962631. html of potato farmers in Carchi, Ecuador. J Agric Appl Econ 39(3):765–780. 40. General Office of Hubei Provincial People’s Government. 2019-06-25. https:// doi. org/ 10. 1017/ s1074 07080 00234 03 Evaluation Report on the Performance of “Special Project for Modern 22. Zilberman D, Zhao J, Heiman A (2012) Adoption versus adaptation, with Agricultural and Rural Energy Development” in 2018. http:// nyt. hubei. emphasis on climate change. Annu Rev Econ 4:27–53. https://www.gov. cn/ fbjd/ xxgkml/ czgk/ ysjx/ 201906/ t2019 0625_ 123499. shtml annualreviews.org/doi/pdf/https:// doi. org/ 10. 1146/ annur ev- resou 41. General Office of Hubei Provincial People’s Government. 2007–4–17. rce- 083110- 115954 Opinions on vigorously promoting biogas construction in rural areas. 23. Mobius M, Rosenblat T (2014) Social learning in economics. Annu http:// www. hubei. gov. cn/ zfwj/ ezbf/ 201112/ t2011 1210_ 17125 55. shtml Rev Econ 6:827–847. https:// doi. org/ 10. 1146/ annur ev- econo 42. Han H, Wu S (2018) Rural residential energy transition and energy con- mics- 120213- 012609 sumption intensity in China. Energy Econ 74:523–534. https:// doi. org/ 10. 24. Oreszczyn S, Lane A, Carr S (2010) The role of networks of practice and 1016/j. eneco. 2018. 04. 033 webs of influencers on farmers’ engagement with and learning about 43. Christiaensen L, Heltberg R (2014) Greening China’s rural energy: new agricultural innovations. J Rural Stud 26(4):404–417. https:// doi. org/ 10. insights on the potential of smallholder biogas. Environ Dev Econ 1016/j. jrurs tud. 2010. 03. 003 19(1):8–29. https:// doi. org/ 10. 1017/ s1355 770x1 30003 75 25. Sol J, Beers PJ, Wals AEJ (2013) Social learning in regional innovation 44. Sun D, Bai J, Qiu H, Cai Y (2014) Impact of government subsidies on networks: trust, commitment and reframing as emergent properties of household biogas use in rural China. Energy Policy 73:748–756. https:// interaction. J Clean Prod 49:35–43. https:// doi. org/ 10. 1016/j. jclep ro. 2012. doi. org/ 10. 1016/j. enpol. 2014. 06. 009 07. 041 45. He P, Veronesi M (2017) Personality traits and renewable energy technol- 26. Joffre OM, De Vries JR, Klerkx L, Poortvliet PM (2020) Why are cluster ogy adoption: a policy case study from China. Energy Policy 107:472–479. farmers adopting more aquaculture technologies and practices? The role https:// doi. org/ 10. 1016/j. enpol. 2017. 05. 017 of trust and interaction within shrimp farmers’ networks in the Mekong 46. Si S, Lyu M, Lawell CYC, Chen S (2018) The effects of energy-related poli- Delta, Vietnam. Aquaculture 523:735181. https:// doi. org/ 10. 1016/j. aquac cies on energy consumption in China. Energy Econ 76:202–227. https:// ulture. 2020. 735181doi. org/ 10. 1016/j. eneco. 2018. 10. 013 27. Kropf B, Schmid E, Schönhart M, Mitter H (2020) Exploring farmers’ behav- 47. Senaviratna NAMR, Cooray TMJA (2019) Diagnosing multicollinearity of ior toward individual and collective measures of Western Corn Rootworm logistic regression model. Asian J Prob Stat 5(2):1–9. https:// doi. org/ 10. control—a case study in south-east Austria. J Environ Manag 264:110431. 9734/ ajpas/ 2019/ v5i23 0132 https:// doi. org/ 10. 1016/j. jenvm an. 2020. 110431 48. Mengistu MG, Simane B, Eshete G, Workneh TS (2016) Factors affect - 28. Mahajan A, Bishop JW, Scott D (2012) Does trust in top management ing households’ decisions in biogas technology adoption, the case of mediate top management communication, employee involvement and Ofla and Mecha Districts, northern Ethiopia. Renew Energy 93:215–227. organizational commitment relationships. J Manag Issues 24(2):173–190. https:// doi. org/ 10. 1016/j. renene. 2016. 02. 066 https:// www. jstor. org/ stable/ 43488 132 49. Campbell RJ, Jeong SH, Graffin SD (2019) Born to take risk? The effect of 29. Jin S, Bluemling B, Mol APJ (2015) Information, trust and pesticide ceo birth order on strategic risk taking. Acad Manag J 62(4):1278–1306. overuse: interactions between retailers and cotton farmers in China. https:// doi. org/ 10. 5465/ ambpp. 2018. 186 NJAS-Wagen J Life Sc 72:23–32. https:// doi. org/ 10. 1016/j. njas. 2014. 10. 003 50. Hoetker G (2007) The use of logit and probit models in strategic manage- 30. Buck S, Alwang J (2011) Agricultural extension, trust, and learning: results ment research: critical issues. Strat Manag J 28:331–343. https:// doi. org/ from economic experiments in Ecuador. Agric Econ 42(6):685–699. 10. 1002/ smj. 582 https:// doi. org/ 10. 1111/j. 1574- 0862. 2011. 00547.x 51. Brambor T, Clark WR, Golder M (2006) Understanding interaction models: 31. Hunecke C, Engler A, Jara-Rojas R, Poortvliet PM (2017) Understanding improving empirical analyses. Polit Anal 14:63–82. https:// doi. org/ 10. the role of social capital in adoption decisions: an application to irrigation 1093/ pan/ mpi014 technology. Agric Syst 153:221–231. https:// doi. org/ 10. 1016/j. agsy. 2017. 52. Zeng Y, Zhang J, He K (2019) Eec ff ts of conformity tendencies on house - 02. 002 holds’ willingness to adopt energy utilization of crop straw: evidence 32. Wang H, Yu F, Reardon T, Huang J, Roelle S (2013) Social learning and from biogas in rural China. Renew Energy 138:573–584. https:// doi. org/ parameter uncertainty in irreversible investments: evidence from green-10. 1016/j. renene. 2019. 02. 003 house adoption in northern China. China Econ Rev 27:104–120. https:// 53. Mittal S, Ahlgren EO, Shukla PR (2018) Barriers to biogas dissemination doi. org/ 10. 1016/j. chieco. 2013. 09. 003 in India: a review. Energy Policy 112:361–370. https:// doi. org/ 10. 1016/j. 33. Chen L, Zhao L, Ren C, Wang F (2012) The progress and prospects of rural enpol. 2017. 10. 027 biogas production in China. Energy Policy 51:58–63. https:// doi. org/ 10. 54. Ziegler A (2017) Political orientation, environmental values, and climate 1016/j. enpol. 2012. 05. 052 change beliefs and attitudes: an empirical cross country analysis. Energy 34. Kelebe HE, Ayimut KM, Berhe GH, Hintsa K (2017) Determinants for adop- Econ 63:144–153. https:// doi. org/ 10. 1016/j. eneco. 2017. 01. 022 tion decision of small scale biogas technology by rural households in Zeng et al. Energy, Sustainability and Society (2022) 12:21 Page 16 of 16 55. Franken JRV, Pennings JME, Garcia P (2018) Graphical illustration of inter- action effects in binary choice models: a note. J Agric Econ 69(3):852– 858. https:// doi. org/ 10. 1111/ 1477- 9552. 12257 56. Williams M (2001) In whom we trust: Group membership as an affective context for trust development. Acad Manage Rev 26(3):377–396. https:// doi. org/ 10. 2307/ 259183 57. Abrams LC, Cross R, Lesser E, Levin DZ (2003) Nurturing interpersonal trust in knowledge-sharing networks. Acad Manag Perspect 17:64–77. https:// www. jstor. org/ stable/ 41660 07 58. Ma W, Renwick A, Yuan P, Ratna N (2018) Agricultural cooperative mem- bership and technical efficiency of apple farmers in China: an analysis accounting for selectivity bias. Food Policy 81:122–132. https:// doi. org/ 10. 1016/j. foodp ol. 2018. 10. 009 59. Liang Q, Huang Z, Lu H, Wang X, Ifamr I (2015) Social capital, member par- ticipation, and cooperative performance: Evidence from China’s Zhejiang. Int Food Agribus Man Assoc 18(1):49–77. https:// www. resea rchga te. net/ publi cation/ 28176 2869 60. Oi JC, Rozelle S (2000) Elections and power: the locus of decision-making in Chinese villages. China Q 162:513–539. https:// doi. org/ 10. 1017/ S0305 74100 00082 37 61. Yip W, Hsiao W (2009) China’s health care reform: a tentative assessment. China Econ Rev 20(4):613–619. https:// doi. org/ 10. 1016/j. chieco. 2009. 08. 62. Wang HH, Zhang YP, Wu L (2011) Is contract farming a risk management instrument for Chinese farmers? China Agric Econ Rev 3(4):489–504. https:// doi. org/ 10. 1108/ 17561 37111 11923 47 Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in pub- lished maps and institutional affiliations. Re Read ady y to to submit y submit your our re researc search h ? Choose BMC and benefit fr ? Choose BMC and benefit from om: : fast, convenient online submission thorough peer review by experienced researchers in your field rapid publication on acceptance support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year At BMC, research is always in progress. Learn more biomedcentral.com/submissions http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png "Energy, Sustainability and Society" Springer Journals

The impacts of observational learning and word-of-mouth learning on farmers’ use of biogas in rural Hubei, China: does interpersonal trust play a role?

Download PDF
16 pages

Loading next page...
 
/lp/springer-journals/the-impacts-of-observational-learning-and-word-of-mouth-learning-on-vF40tcXeW9

References (51)

Publisher
Springer Journals
Copyright
Copyright © The Author(s) 2022
eISSN
2192-0567
DOI
10.1186/s13705-022-00350-8
Publisher site
See Article on Publisher Site

Abstract

Background: Residue-based biogas is considered as a renewable energy that should be used to improve energy security and household livelihoods in rural areas. Observational learning and word-of-mouth learning are critical in the dissemination of knowledge about agricultural technologies. Yet, scholars have little understanding of the impacts of these learning methods on farmers’ use of residue-based biogas. Using survey data from rural areas of Hubei China, this study estimates the impacts of observational learning and word-of-mouth learning from different subjects (i.e., relatives, neighbors, cadres, cooperative members, and technical instructors) on the use of residue-based biogas by farmers. Additionally, the moderating role of interpersonal trust in these relationships is explored. Results: Results from logistic regression models show that observational learning from technical instructors signifi- cantly increases farmers’ use of biogas. Furthermore, interpersonal trust significantly and positively influences the impact of observational learning on farmers’ decisions to use biogas. Similarly, interpersonal trust significantly and positively moderates the influence of positive word-of-mouth learning on farmers’ decision to use biogas. In contrast, a negative moderating role exists in the relationship between negative word-of-mouth learning and farmers’ decision to use biogas. These impacts are further affirmed by robustness checks. Conclusions: The results presented here show that enhancing farmers’ interpersonal trust promotes the use of resi- due-based biogas by farmers. One important implication is that the government might promote the use of residue- based biogas by organizing technology demonstration activities, providing communication platforms, and enhancing mutual trust between farmers and relevant groups. et  al. [4] revealed that just 29% and 11% of the sampled Background households in Tanzania and Uganda, respectively, use Residue-based biogas is becoming increasingly attrac- biogas exclusively. According to Rahman et  al. [5], only tive as a means to improve energy security and house- 32.5% of the surveyed households adopted the biogas hold livelihoods in rural areas [1–3]. However, it is still technology in 2021 based on data from four districts in underutilized in rural areas around the world, especially rural Bangladesh. Therefore, identifying the obstacles to in developing countries [4, 5]. For example, Clemens biogas technology adoption is a formidable challenge for policymakers. *Correspondence: zhangjb513@126.com Empirical evidence suggests that the lack of rele- College of Economics and Management, Huazhong Agricultural vant information is a major obstacle to the diffusion of University, 1 Shizishan Street, Hongshan, Wuhan 430070, Hubei, China biogas technology in many developing countries [6, 7]. Full list of author information is available at the end of the article © The Author(s) 2022. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. The Creative Commons Public Domain Dedication waiver (http:// creat iveco mmons. org/ publi cdoma in/ zero/1. 0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Zeng et al. Energy, Sustainability and Society (2022) 12:21 Page 2 of 16 Generally, transportation is usually underdeveloped in is confident in the words or actions of another [28]. As an rural areas in developing countries, and people live in essential ingredient of interpersonal ties, interpersonal congested geographical spaces [8, 9]. Accordingly, social trust can influence the extent to which farmers put the learning among farmers has become prominent in the information acquired from others into practice, such as diffusion of agricultural technology [10–13]. For exam - pesticides use [29], agricultural cultivation technology ple, learning from neighbors [10, 11], extension agents adoption [30], and the use of irrigation technology [31]. [12], and “progressive” peers [13] all help in accelerat- To address these research gaps, this study aims at ing the transfer of information and increasing the use of accomplishing two objectives: (1) to examine the impacts technology by farmers. of OB and WOM on farmers’ use of residue-based The existing literature indicates that farmers often learn biogas; and (2) to investigate whether and to what extent through two prominent social learning mechanisms: interpersonal trust can mediate the effects of OB and observational learning (OB) and word-of-mouth learn- WOM. We contribute to the literature of technology ing (WOM) [14–16]. In OB, individuals infer relevant adoption and social learning by providing the first field information from the actions of other people indirectly evidence of the various roles played by OB and WOM through visual observations [14, 17]. In WOM, individu- in the usage of residue-based biogas. We also contrib- als extract relevant information from the opinion of other ute to the trust literature by identifying interpersonal people directly through verbal communication [18]. trust as a mediating factor in the effects of social learn - These two types of learning occur in different ways. OB ings. China is an interesting and relevant case study for occurs when relevant subjects are facing similar shocks empirically illustrating technology adoption concerns. in making decisions, whereas WOM occurs when tempo- In addition to the semi-closed geographical conditions ral, spatial, and social proximity among individuals exists in rural areas, the distinctive Chinese rural acquaintance [14]. Furthermore, WOM can be divided into positive society makes technology information to spread through WOM, meaning that it is explicitly from a subject with observation and communication more conveniently [32]. positive experiences, and negative WOM, which comes Consequently, this study used a unique survey data set from a subject with negative experiences [19]. collected in rural areas of Hubei province in China. The The effects of OB and WOM have been separately outcomes of this study will help us better understand assessed in the extant empirical publications. For exam- adoption decisions and develop strategies that leverage ple, Jones et al. [20] showed that OB from other farmers interpersonal trust to encourage the adoption of new has a causal effect on the uptake of the novel pigeon pea technologies. variety in the semi-arid areas of Mwea Division in the The remainder of this paper is organized as follows: Eastern Province of Kenya. Conley and Udry [11] con- The next section describes the overview of residue-based firmed that farmers in Ghana choose fertilizer by observ - biogas in rural China. “Methods” section presents the ing the inputs and outputs of their neighbors. Based on data and methodology. “Results” section discusses the sample potato farmers in Ecuador, Mauceri et  al. [21] estimation results. “Discussion” section discusses and found that the diffusion of integrated pest management deals with policy implications and also identifies limita - techniques is influenced by WOM. Similarly, Zilberman tions and gives an outlook for future research. “Conclu- et al. [22] suggested that WOM could induce technology sions” section offers the conclusions. adoption. Although OB and WOM may coincide [23], few studies have disentangled the distinct effects of the Overview of residue‑based biogas in rural China two learning mechanisms against the same backdrop. Residue-based biogas is a mixture of methane, carbon We are unaware of any research regarding the role of dioxide, and other gases generated from agricultural interpersonal trust in the investigation of the impacts of residues through anaerobic processes in liquid–state/ OB and WOM. Existing literature suggests that the extent solid–state/liquid–solid two-phase digestions [33, 34]. to which social learning may speed up adoption is closely This mixture is commonly used for cooking, generating related to the connections among individuals [24]. Mean- power, and fueling vehicles among others [35, 36]. ingful social connections create value typically in the In China, residue-based biogas is increasingly gain- form of interpersonal trust that helps individuals to learn ing value, especially as a renewable and clean alternative information from others to improve different behaviors cooking fuel in the rural areas [33]. As the largest devel- [25]. For example, farmers adopting aquaculture technol- oping country in the world, China has an abundance ogies and practices in the Mekong Delta, Vietnam [26], of agricultural residues. According to the Ministry of and behaviors of farmers toward individual and collec- Agriculture and Rural Areas [37], more than 4.7 billion tive measures of controlling the western corn rootworm tons of agricultural residues have been produced annu- [27]. Interpersonal trust is the extent to which a person ally in recent years, with 29.87% of available agricultural Z eng et al. Energy, Sustainability and Society (2022) 12:21 Page 3 of 16 residues used to generate biogas that is mainly consumed questions, modified the language, and added some by farmers according to the National 13th Five-Year Plan interview questions to obtain the final version of the for Rural Biogas Development [38]. questionnaire, which had six parts. To promote the use of residue-based biogas in rural The first part captured the details of infrastructure areas, effective policies such as the Pilot Scheme of Pro - and living conditions in the villages, including the ter- moting the Resource Utilization of Agricultural Wastes rain where the households were located and the closest [37] and Guiding Opinions on the Construction of Straw town or market. The second part entailed personal and Gasification Clean Energy Utilization Project [39] have household characteristics such as gender, age, educa- been implemented in the past decade. These top-down tion, and total household income. The third part cov - national policies not only promote biogas expansion by ered rural biogas utilization such as social learnings, strengthening scientific–technological innovation and interpersonal trust, and the adoption of biogas utiliza- providing service guarantee, but they also set national tion. The fourth part evaluated the extension of agri - short- and long-term targets for achieving sustainable use cultural technologies, whereas the fifth part collected of residue-based biogas and call for efforts and actions information on farmers’ participation in social activi- from provincial governments. Notably, the national and ties and their waste disposals. The final part mainly local governments have already provided financial sup - measured self-happiness of farmers and their percep- port for the use of residue-based biogas. During the 12th tions of different situations. Additional file  1 is part of Five-Year Plan period (2011–2015), 14.2 billion RMB the survey questionnaire translated from Chinese into (equivalent to 2.25 billion USD) were invested in the English. construction of rural biogas projects for farmers [38]. In 2018, investment in small biogas projects for farmers in the sample province Hubei was 86.4 million RMB (equiv- Data collection alent to 12.55 million USD) [40]. Furthermore, each The data used in this study were collected from a house - household in Hubei province that has a biogas digester hold survey conducted in rural areas of Hubei province, has been subsidized with 1000 RMB (equivalent to 132.98 China, in August 2018 using a multistage systematic USD) since 2007 [41]. random sampling procedure. First, four cities were ran- Despite the efforts by the national and provincial gov - domly selected: Huanggang city, Wuhan city, Ezhou city ernments, farmers’ use of biogas in rural China remains and Jingmen city. Second, one to two counties or districts unsatisfactory [42]. It is estimated that the proportion of in each city were randomly selected. The selected coun - biogas farmers consumed in 2014 was only 3.31% [42]. ties or districts cover the main landforms in Hubei prov- Recently, numerous studies have identified predictions of ince, which range from hills to mountains and plain. The residue-based biogas use in rural China [43–46]. The pre - reason for selecting counties or districts at random with dictions include demographic factors (such as age, labor respect to the main landforms was that the use of biogas availability, and total household income), village basic varies with household location in different landforms infrastructure [43], personality traits [45], and energy- [2, 44]. Third, three to five towns in each county or dis - related command and control policies [44, 46]. However, trict were randomly selected. Finally, one village in each no research has investigated the impacts of different town per county or district was randomly selected. In types of social learning. this stage, we obtained a household roster for each village from the local government and randomly selected house- holds for interview. Interviews were conducted by team Methods members from the School of Economics and Manage- Data sources ment, Huazhong Agricultural University. All the mem- Questionnaire design bers had rich experience in rural investigation and were This study was based on surveys conducted in rural professionally trained before the survey was conducted. areas of Hubei province, China. To obtain the data, Supervised face-to-face interviews were conducted with we designed a detailed questionnaire, which was an adult member of the sample households. then modified by the relevant experts in agricultural A total of 1084 observations were collected. The final resource economy from Hubei Rural Development sample for this study comprised 913 observations after Research Center. The experts considerably improved excluding those with missing values. Observations from the logic and language. A pre-test was conducted Huanggang city, Wuhan city, Ezhou city, and Jingmen among 20 farmers in our targeted area of study to city were 163, 328, 203, and 219, respectively, which boost the validity, accuracy, and credibility of our data. accounted for 17.85%, 35.93%, 22.23%, and 23.99%, Based on the pre-test results, we deleted few invalid respectively. Zeng et al. Energy, Sustainability and Society (2022) 12:21 Page 4 of 16 Methodology VIF scores range from 1.04 to 3.24. Campbell et al. [49] Model selection demonstrated that a multicollinearity problem exists In this research, the dependent variable was dichoto- when VIF is greater than 5. Thus, the multicollinearity mous, with 1 indicating that the ith farmer had used res- test result was acceptable in our study. Furthermore, the idue-based biogas (y = 1), and 0 indicating otherwise (y robust standard error procedure was used in this paper in i i = 0). The binary logit model is one of the most widely order to obtain unbiased standard errors under hetero- used statistical models for dealing with the relationship scedasticity. We also used OLS and probit approaches to between dichotomous dependent variable and multiple check the robustness of the binary logistic estimates. continuous or categorical independent variables [47]. According to previous studies, coefficients in non - Therefore, this paper used a binary logistic model to deal linear regressions (e.g., logit) cannot be used to infer with discrete outcomes, assuming that the ith farmer had the statistical significance of an interaction term and its or had not used a distribution Pr(y ). Pr(y ) is believed to underlying variables [50, 51]. The marginal effect of the i i be affected by a vector of key explanatory variables (X ) interaction term varies not only in sign, but also in mag- (e.g., OB, and positive and negative WOM), interpersonal nitude and statistical significance with the values of other trust (I ), and the interaction terms between the learnings explanatory variables [51]. The interpretation of logistic and interpersonal trust (X I ). To avoid potential endoge- model coefficients has certain pitfalls and is less intuitive i i neity due to omitted variable biases, we added other fac- than linear model estimation. Hence, in this paper, the tors to the model as a vector of control variables (Z ) that marginal effects were calculated using the “delta” method may explain the differences in the biogas use probability in Stata. As the relationship between the probability of among the farmers. These factors included demographic social learning and residue-based biogas use is nonlinear, variables, personal perceptions, and geographic loca- the marginal change could be presented by the tangent to tions [34, 48]. Furthermore, it is assumed that Pr(y ) is the probability curve: influenced by an error term μ . The corresponding set of ∂ Pr(y ) ∂ Pr(y ) ∂y i i parameters is {β , β , β , β }, where β is the intercept. It 1 2 3 4 0 = · ∂X ∂y ∂X i i is considered that Pr(y ) is determined by the aforemen- (5) ∗ ∗ tioned factors through a nonlinear link function F that =[�(y )(1 − �(y ))]· (β + β I ) 1 3 i i i maps the unbounded index. Since only dummy variables were involved in the inter- y = β + β X + β I + β X I + β Z + µ 0 1 i 2 i 3 i i 4 i i (1) action terms, the marginal change in this paper was pre- sented as follows: into bounded probability space [0,1]: �y = Pr(y |X, X = 1) − Pr(y |X, X = 0) i i i i (6) Pr(y ) = F (y ) i (2) For the main explanatory variables in this paper, OB where F is the logistic cumulative density function ( ) refers to the process of observing as other individuals that produces the logit model. Thus, use residue-based biogas. Observing others is a behav- ∗ ior-based social interaction that farmers can use as a 1 exp(y ) ∗ ∗ i Pr(y ) = F(y ) = �(y ) = = i reference when making choices from an overwhelming i i ∗ ∗ 1 + exp(y ) 1 + exp(y ) i i number of options. OB may also update personal experi- (3) ence of farmers and beliefs about the profitability of the Then, with (1) incorporated into (3), we get: technology based on the profitable signal from the adop - ∗ ∗ tion behavior of others [52]. Consequently, although the Pr(y ) = F(y ) = �(y ) i i profits and utility of others’ use are unknown, OB may increase the probability of farmers using residue-based 1 + exp[−(β + β X + β I + β X I + β Z + µ ) 0 1 i 2 i 3 i i 4 i i biogas. exp(β + β X + β I + β X I + β Z + µ ) 0 1 i 2 i 3 i i 4 i i Farmers in rural areas rely on various cues, including 1 + exp(β + β X + β I + β X I + β Z + µ ) 0 1 i 2 i 3 i i 4 i i WOM, to acquire technology information in their imme- (4) diate social circles. In this paper, positive WOM refers To investigate potential multicollinearity affecting the explicitly to the process of learning via communication result, we calculated the variance inflation factor (VIF) from the positive experiences of others who have used for the binary logistic model with all variables included in the regression estimation except the interaction terms, As noted by Brambor et  al. [51], any multicollinearity problem cannot be and the calculated results are shown in Appendix 1. The solved by centering the relevant variables. Therefore, the variables in this paper were involved in the interaction terms and were not centered in the logit estimation by subtracting their means. Z eng et al. Energy, Sustainability and Society (2022) 12:21 Page 5 of 16 Fig. 1 Theoretical framework of this paper residue-based biogas. This learning method may inform The relationships between the key independent vari - the beliefs of farmers about the strengths, expected qual- ables, the trust variables, and the dependent variable are ity or profit, and other positive effects of the technology. presented in summary in Fig. 1. It could also increase the exante confidence of the farm - ers in the expected utility of the technology. Therefore, Variables and descriptive statistics positive WOM may increase farmer use rates. On the As aforementioned, the dependent variable was farmers’ contrary, negative WOM refers to the process of learn- use of residue-based biogas, while the key independent ing via communication from the negative experiences of variables were OB, positive and negative WOMs and the farmers’ use of residue-based biogas. This type of learn - interaction terms between the learning types and inter- ing may enhance the perceptions of farmers on the disad- personal trust. Several factors serve as control variables vantages and potential problems of residue-based biogas. to rule out alternative explanations. Gender, age and It may also constitute a noisy signal about the profitability education of the respondents were controlled because of this technology, thus increasing use uncertainty among the characteristics of farmers have been shown to inevi- the farmers. Therefore, negative WOM may reduce the tably correlate with the uptake of biogas technology [6, probability of farmers using the residue-based biogas. 34]. Furthermore, household labor and total household In this paper, interpersonal trust is defined explicitly income were used as control variables because labor as farmers trust the opinion or decisions of others on [43] and family income [48] are all determinants of clean residue-based biogas. According to Sol et  al. [25], inter- energy consumption by households. Sun et al. [44] found personal trust facilitates the relationships between types that biogas subsidy is critical in biogas utilization, hence of social learning and individual behaviors in the face of we included a dummy variable subsidy into the regres- ambiguity and unstructured nature of decision-making sion analysis. Moreover, risk and personal perceptions problems. Embedded within cohesive groups marked served as control variables because they promote biogas with closure, farmers are likely to learn about residue- dissemination [53]. Meeks et  al. [2] reported that biogas based biogas from others. This, in turn, can indicate that projects are generally unsuited for mountain regions due they have trust in other farmers, and the more individuals to temperature requirements to operate them, whereas trust the information provider, the more likely they are Sun et  al. [44] revealed that biogas users are more likely to transform the knowledge obtained from the provider to be located in hilly areas than plains. Accordingly, in into practice. Consequently, the use behaviors of farmers this study, we controlled household location heterogenei- tend to be consistent. Therefore, interpersonal trust may ties to reduce the error in the regression analysis caused mediate the relationships between types of social learn- by landform factor disunity. ing and the use of residue-based biogas. Variable definitions and descriptive statistics are pre - sented in Table  1. According to Table  1, 22% of sur- veyed farm households had used residue-based biogas. Zeng et al. Energy, Sustainability and Society (2022) 12:21 Page 6 of 16 Table 1 Variable definitions and descriptive statistics Variables Definition Mean (S.D.) Adoption 1 if respondent has used residue-based biogas, 0 otherwise 0.22 (0.41) OB from relatives (OB_r) 1 if respondent obtain observational learning from relatives’ adoption of residue- 0.51 (0.50) based biogas, 0 otherwise OB from neighbors (OB_n) 1 if respondent obtain observational learning from neighbors’ adoption of 0.59 (0.49) residue-based biogas, 0 otherwise OB from cadres (OB_c) 1 if respondent obtain observational learning from cadres’ adoption of residue- 0.54 (0.50) based biogas, 0 otherwise OB from cooperative members (OB_o) 1 if respondent obtain observational learning from cooperative members’ adop- 0.31 (0.46) tion of residue-based biogas, 0 otherwise OB from technical instructors (OB_t) 1 if respondent obtain observational learning from technical instructors’ adop- 0.32 (0.47) tion of residue-based biogas, 0 otherwise Positive WOM from relatives (WOM_rp) 1 if respondent obtain positive word-of-mouth learning from relatives, 0 0.24 (0.43) otherwise Positive WOM from neighbors (WOM_np) 1 if respondent obtain positive word-of-mouth learning from neighbors, 0 0.27 (0.44) otherwise Positive WOM from cadres (WOM_cp) 1 if respondent obtain positive word-of-mouth learning from cadres , 0 0.13 (0.34) otherwise Positive WOM from cooperative members (WOM_op) 1 if respondent obtain positive word-of-mouth learning from cooperative 0.04 (0.19) members, 0 otherwise Positive WOM from technical instructors (WOM_tp) 1 if respondent obtain positive word-of-mouth learning from technical instruc- 0.07 (0.25) tors, 0 otherwise Negative WOM from relatives (WOM_rn) 1 if respondent obtain negative word-of-mouth learning from relatives, 0 0.21 (0.40) otherwise Negative WOM from neighbors (WOM_nn) 1 if respondent obtain negative word-of-mouth learning from neighbors, 0 0.22 (0.42) otherwise Negative WOM from cadres (WOM_cn) 1 if respondent obtain negative word-of-mouth learning from cadres , 0 0.11 (0.31) otherwise 0.03 (0.18) Negative WOM from cooperative members (WOM_on) 1 if respondent obtain negative word-of-mouth learning from cooperative members, 0 otherwise Negative WOM from technical instructors (WOM_tn) 1 if respondent obtain negative word-of-mouth learning from technical instruc- 0.09 (0.29) tors, 0 otherwise Trust in relatives (Tr) 1 if respondent trusts relatives, 0 otherwise 0.41 (0.49) Trust in neighbors (Tn) 1 if respondent trusts neighbors, 0 otherwise 0.73 (0.45) 1 b Trust in cadres (Tc) 1 if respondent trusts cadres , 0 otherwise 0.62 (0.49) Trust in cooperative members (To) 1 if respondent trusts cooperative members, 0 otherwise 0.65 (0.47) Trust in technical instructors (Tt) 1 if respondent trusts technical instructors, 0 otherwise 0.58 (0.49) Gender 1 if respondent is male; 0 if female 0.56 (0.50) Age Respondent’s age 57.52 (10.94) Education Schooling of respondents (in years) 6.43 (3.75) Labor Number of individuals in the household that are aged 16 or more but below 3.13 (1.41) 65 years old Household income Total household income in 2017 (10 000 Yuan) 6.96 (7.92) Subsidy 1if subsidy is provided for those who use residue-based biogas 0.23 (0.42) Risk perception Risk perception of the adoption of residue-based biogas 2.39 (1.01) Cost-effective perception Cost-effective perception of the adoption of residue-based biogas 3.95 (0.89) Plain 1 if household is located in plain, 0 otherwise 0.26 (0.44) Hill 1 if household is located in hill, 0 otherwise 0.69 (0.46) Mountains 1 if household is located in mountains, 0 otherwise 0.05 (0.22) N = 913. Following Ziegler [54], if the respondent has high or very high frequency of obtaining positive/negative word-of-mouth learning from these objects, we take value one; otherwise, we take value zero. We take value one if the respondent have high or very high interpersonal trust in these objects, and otherwise, we take c d 1 value zero. Yuan is Chinese currency (1$ = 6.62 Yuan in 2018). related to a 5-point-Likert scale, 1-very low; 5-very high. Cadres refers to local governors who hold certain positions in the village’s political organization, exercise local power, manage local affairs and provide local services, etc Z eng et al. Energy, Sustainability and Society (2022) 12:21 Page 7 of 16 Moreover, farmers who obtained OBs from relatives, OBs from relatives and neighbors and the use behavior neighbors, cadres, cooperative members, and technical all become insignificant when the interaction terms were instructors accounted for 51%, 59%, 54%, 31% and 32%, included. This emphasized the importance of interper - respectively. In addition, 24%, 27%, and 13% of the sample sonal trust in mediating the relationships between OBs farmers obtained positive WOMs from relatives, neigh- and adoption behavior, implying that focusing solely on bors, and cadres, respectively. However, much fewer the estimation results of model 4 is misleading. There - farmers obtain this kind of learning from cooperative fore, the conclusions of the following analyses are mainly members (4%) and technical instructors (7%). Similarly, based on the results of model 5. farmers who obtained negative WOMs from relatives, neighbors and cadres accounted for 21%, 22%, and 11% The impacts of OBs and WOMs on use of biogas by farmers accordingly, which were all much higher than those of The coefficient for OB from technical instructors was negative WOMs from cooperative members (3%) and positive and statistically significant at the 5% level, indi - technical instructors (9%). Table  1 shows that 73% and cating that it significantly influenced the use behavior. 65% of farmers trusted their neighbors and coopera- However, the coefficient of the corresponding interaction tive members, respectively. Farmers who trusted cadres, term was not statistically significant, which suggested that technical instructors, and relatives accounted for 62%, OB from technical instructors significantly and directly 58%, and 41%, respectively. Notably, during the inter- increased the likelihood of the use by farmers. This find - views, interviewers clarified to the respondents that “rel - ing is consistent with the study of Krishnan and Patnam atives”, “neighbors”, “cadres”, “cooperative members”, and [12], which demonstrated that OB from extension agents “technical instructors” refer to different types of generic can effectively predict technology adoption, which is and group concepts rather than specific individuals. probably because technical instructors have professional technical skills. Therefore, obtaining OB from these Results instructors’ adoption of residue-based biogas could be an Main results efficient and reliable source that indicates the expected To better understand the direct influences of OBs and utility of this technology. Therefore, this type of learn - WOMs on farmers’ use of residue-based biogas and the ing greatly increases the probability of farmers using the moderating role of interpersonal trust in these influences, technologies. we constructed 5 models in which key independent vari- ables and interaction terms into the models step by step. The moderating role of interpersonal trust The results of the 5 binary logistic regression estimates Interpersonal trust not only significantly strengthened are presented in Table 2, while all the estimation findings the impact of OB from relatives on the use behavior, but are given in Appendix 2. it also statistically and significantly strengthened the In Table 2, model 1 exclusively tests the effects of inter - relationships between positive WOMs from neighbors, personal trust and control variables on farmers’ use of cooperative members, technical instructors, and the use residue-based biogas. Model 2 adds the key independ- behavior. In contrast, interpersonal trust statistically ent variables to explore the direct effects of OBs and and significantly weakened the links between negative WOMs on farmers’ use of residue-based biogas. Model WOMs from cadres, cooperative members, and the use 3 estimates the interaction between OBs and interper- behavior. As Brambor et  al. [51] suggested, constructing sonal trust, while model 4 tests the interaction between marginal effect plots is an effective way to show how the WOMs and interpersonal trust. Model 5, which was the estimated marginal effect of a variable on the probabil - preferred model, includes all the key explanatory vari- ity varies with another. Furthermore, there is a need to ables, interaction terms, interpersonal trust, and control plot the marginal effect of the interaction terms at the variables. Likelihood-ratio tests, AUC, and AIC results all mean value [51] with all dichotomous variables involved suggested that models 2–5 were significantly better than in the interaction terms. Therefore, following the study of the baseline model 1. When the pseudo-R s, AUC, and Franken et  al. [55], we used figures to graphically depict AIC in the models 1–5 were compared, it was noted that the magnitude and significance of the interaction effects both models 4 and 5 had the most explanatory power. in model 5. The plots are presented in (a) ~ (f ) in Fig. 2. According to model 4, the effects of OB from relatives and neighbors on farmers’ use were statistically signifi - Moderating role of interpersonal trust on the relationships cant, although the corresponding interaction terms were between OBs and adoption excluded. However, in model 5, the positive link between The interaction coefficient of OB from relatives and trust in relatives was positive and statistically significant at The direct effects of different types of interpersonal trust on the use behav - the 5% level. A graph of this moderating effect, which is ior were out of scope of this research. Zeng et al. Energy, Sustainability and Society (2022) 12:21 Page 8 of 16 Table 2 Evaluation statistics of the logistic model Variables Model 1 Model 2 Model 3 Model 4 Model 5 ** ** OB_r 0.64 (0.27) 0.22 (0.30) 0.67 (0.28) 0.26 (0.31) ** ** OB_n 0.83 (0.36) 0.56 (0.67) 0.86 (0.37) 0.81 (0.69) OB_c − 0.39 (0.35) − 0.30 (0.44) − 0.45 (0.37) − 0.40 (0.45) OB_o − 0.22 (0.32) − 0.05 (0.44) − 0.28 (0.33) − 0.05 (0.44) *** ** *** ** OB_t 0.94 (0.33) 0.79 (0.39) 1.09 (0.34) 0.97 (0.40) WOM_rp 0.25 (0.29) 0.23(0.29) − 0.01 (0.36) 0.02 (0.35) *** *** WOM_np 1.14 (0.30) 1.19 (0.31) 0.02 (0.65) 0.06 (0.65) *** *** ** *** WOM_cp 0.99 (0.34) 1.00 (0.33) 1.27 (0.49) 1.29 (0.49) * * WOM_op 1.11 (0.60) 1.08 (0.59) − 0.97 (1.23) − 1.19 (1.29) *** *** WOM_tp − 0.78 (0.58) − 0.73 (0.56) − 2.34 (0.80) − 2.20 (0.79) WOM_rn 0.02 (0.34) 0.04 (0.34) 0.13 (0.38) 0.15 (0.38) *** *** WOM_nn − 1.01 (0.34) − 1.06 (0.34) − 0.25 (0.76) − 0.27 (0.75) ** ** WOM_cn − 0.87 (0.38) − 0.91 (0.38) − 0.76 (0.53) − 0.74 (0.53) *** *** WOM_on 0.72 (0.62) 0.62 (0.62) 3.10 (1.00) 3.12 (1.03) WOM_tn − 0.25 (0.41) − 0.19 (0.41) 0.14 (0.47) 0.17 (0.46) ** ** OB_r × Tr 1.11 (0.44) 1.12 (0.45) OB_n × Tn 0.29 (0.63) 0.03 (0.65) OB_c × Tc 0.03(0.44) 0.07 (0.45) OB_o × To − 0.23 (0.42) − 0.31 (0.43) OB_t × Tt 0.19 (0.40) 0.11 (0.41) WOM_rp × Tr 0.56 (0.60) 0.42 (0.60) ** ** WOM_np × Tn 1.58 (0.70) 1.56 (0.69) WOM_cp × Tc − 0.43 (0.70) − 0.48 (0.68) ** ** WOM_op × To 3.33 (1.43) 3.48 (1.47) *** *** WOM_tp × Tt 3.60 (1.18) 3.36 (1.14) WOM_rn × Tr − 0.42 (0.63) − 0.38 (0.64) WOM_nn × Tn − 1.06 (0.79) − 1.09 (0.78) * * WOM_cn × Tc − 1.47 (0.82) − 1.54 (0.79) *** *** WOM_on × To − 3.22 (1.21) − 3.34 (1.22) WOM_tn × Tt − 1.06 (0.88) − 0.92 (0.85) Interpersonal trust variables Yes Yes Yes Yes Yes Control variables Yes Yes Yes Yes Yes * *** *** *** ** Constant − 1.39 (0.83) − 3.43 (0.89) − 2.78 (1.01) − 3.05 (0.93) − 2.66 (1.06) Log likelihood − 426.72 − 369.44 − 377.00 − 356.81 − 353.11 Pseudo-r 0.11 0.23 0.23 0.25 0.26 Prob > chi 0.00 0.00 0.00 0.00 0.00 *** *** *** *** Likelihood-ratio test 114.57 122.41 139.82 147.23 AUC 0.73 0.82 0.82 0.83 0.84 AIC 885.45 800.87 803.04 795.63 798.21 *** ** * N = 913; p < 0.01, p < 0.05, p < 0.1; standard errors are in parentheses; landform dummies are included in models 1, 2, 3, 4 and 5 with plains as the reference shown in Fig. 2a, allows further investigation. Specifically, As a result, personal ambiguity about the technology can there was a 15% increase in the use probability (from 0.11 be greatly reduced if farmers trusted their relatives. Con- to 0.26) by farmers who trusted and obtained OB from sequently, having trust in relatives may strengthen the relatives. However, for farmers who did not trust their relationship between OB from relatives and biogas use. relatives, the increase in use probability was only 4% Williams [56] noted that when trust is present, positive (from 0.21 to 0.25). One explanation for this phenom - information may facilitate consistent behaviors like tech- enon is that trusting relatives can make the information nology adoption, which requires little time and cognitive obtained via OB more salient, reliable, and persuasive. resources. Z eng et al. Energy, Sustainability and Society (2022) 12:21 Page 9 of 16 (a) Moderating effect of interpersonal truston the (b) Moderating effect of interpersonal truston the relationship between OB from relatives and adoption relationship between positive WOM from neighbors and adoption (c) Moderating effect of interpersonal truston the (d) Moderating effect of interpersonal truston the relationship between positive WOM from relationship between positive WOM from technical instructors and adoption cooperative members and adoption (e) Moderating effect of interpersonal trust on the (f) Moderating effect of interpersonal trust on the relationship between negative WOM from cadres relationship between negative WOM from cooperative members and adoption and adoption Fig. 2 Moderating effect of interpersonal trust on the relationships between WOMs/OBs and adoption Zeng et al. Energy, Sustainability and Society (2022) 12:21 Page 10 of 16 Moderating role of interpersonal trust on the relationships was positive and statistically significant at the 1% level. between WOMs and adoption A graph illustrating this moderating effect is shown in The interaction coefficient of positive WOM from neigh - Fig.  2d. Farmers who trusted technical instructors and bors with trust in neighbors was positive and statistically acquired positive WOM from them had a 16% higher significant at the 5% level (Fig.  2b). Farmers who trusted chance (from 0.21 to 0.37) of using residue-based biogas their neighbors and acquired positive WOM from them compared to those who trusted technical instructors but had a 25% higher probability (from 0.19 to 0.44) of using failed to obtain this type of learning. However, among the residue-based biogas than those who did not trust their farmers who did not trust the technical instructors, those neighbors had only 1% (from 0.11 to 0.12) increase in use who obtained positive WOM from them had an 18% probability, which was in line with the study of Abrams lower probability (from 0.24 to 0.06) of using residue- et al. [57], who stated that interpersonal trust is a driver based biogas than those who did not. Technical instruc- for the knowledge and experience sharing in networks, tors are expected to have a significant degree of technical and makes knowledge exchanges less costly. The trans - information and to be skilled in agricultural technolo- formation of knowledge into actions could be greatly gies. Thus, trusting them can greatly help the transfor - promoted if interpersonal trust exists. Farmers in rural mation of positive WOM from technical instructors to China are closely interconnected with their neighbors, real actions. However, technical instructors’ work for the resulting in familiarity and frequent interaction with government in rural areas [12], and farmers distrust local each other. Therefore, positive WOM from neighbors can government representatives, could exacerbate a negative enhance the perceptions of farmers on the advantages of influence. Therefore, the probability of biogas uses for residue-based biogas and neighbors’ satisfaction with the farmers who do not trust technical instructors yet obtain adoption. Trust in neighbors helps to make these percep- positive WOM from them may be reduced. tions more profound, inducing their transformation into Model 5 indicated that the interaction coefficient action. between the negative WOM from cadres and trust in The interaction coefficient of positive WOM from cadres was negative and statistically significant at the 10% cooperative members with trust in cooperative mem- level. This moderating effect is shown in a graph (Fig.  2e). bers was also positive and statistically significant at the Farmers who trusted the cadres and as well obtained 5% level. This moderating effect is highlighted in Fig.  2c. negative WOM from them had an 18% lower chance Farmers who trusted cooperative members and acquired (from 0.24 to 0.06) of using residue-based biogas than positive WOM from them had a 36% higher chance (from those who trusted the cadres but did not obtain this type 0.18 to 0.54) of using residue-based biogas than those of learning. However, among farmers who did not trust who trusted cooperative members but failed to obtain the cadres, those who obtained negative WOM from the this type of learning. However, among farmers who did cadres had an 8% lower probability (from 0.22 to 0.14) not trust cooperative members, those who obtained posi- of using residue-based biogas than those who failed to tive WOM from cooperative members had a 13% lower obtain this type of learning. This observation could be probability (from 0.26 to 0.13) of using residue-based because village cadres in rural China represent power biogas than those who did not obtain this type of learn- and authority, and are respected by rural farmers [60]. ing. This finding has its particularity and rationality. Trust in cadres can distinctively increase farmers’ nega- Cooperative members in rural China are assumed to have tive perceptions of residue-based biogas if they obtain better access to technical information because the spe- this kind of learning from cadres, resulting in a low prob- cialized cooperative organizations provide their mem- ability of use. bers with excellent technical services [58]. Therefore, Model 5 suggests that the interaction coefficient of farmers who trust these members can have confidence in negative WOM from cooperative members with trust the positive information accumulated by positive WOM in cooperative members is negative and statistically sig- from these members. On the contrary, farmers are not nificant at the 1% level. This moderating effect is shown emotionally close to cooperative members because of in Fig.  2f. Specifically, farmers who trusted cooperative different social identities [59], which could lead to a dis - members and obtained negative WOM from them had trust in cooperative members. Against this backdrop, the a 3% lower probability (from 0.20 to 0.17) of using res- side effect of total distrust in cooperative members may idue-based biogas than those who did not obtain this be greatly strengthened, resulting in a negative impact type of learning. In comparison, among farmers who on the relationship between positive WOM and the use did not trust cooperative members, those who obtained behavior. negative WOM from cooperative members had a 49% The interaction coefficient of positive WOM from higher probability (from 0.24 to 0.73) of using residue- technical instructors with trust in technical instructors based biogas than those who did not obtain this type of Z eng et al. Energy, Sustainability and Society (2022) 12:21 Page 11 of 16 learning. This seemingly contradictory result is not sur - Table 3 Robustness check results with OLS and probit model employed prising for rural China. As stated before, cooperative members are assumed to have good technical informa- Variables OLS Probit tion. Negative WOM may violate farmers’ previous on OB_r 0.04 (0.05) 0.14 (0.18) the utility of this biogas given the asymmetric informa- OB_n 0.02 (0.06) 0.39 (0.33) tion in rural China [61]. Therefore, farmers who trust OB_c − 0.04 (0.06) − 0.20 (0.25) cooperative members have a lower use probability since OB_o 0.02 (0.07) − 0.03 (0.24) cooperative members are not emotionally attentive to ** ** OB_t 0.15 (0.06) 0.55 (0.23) farmers. The lack of trust among farmers in cooperative WOM_rp 0.05 (0.06) 0.04 (0.20) members may have a reverse effect, i.e., the farmers’ WOM_np − 0.00 (0.10) 0.02 (0.35) use probability will increase even though they receive ** ** WOM_cp 0.18 (0.07) 0.71 (0.28) negative WOM from cooperative members. WOM_op − 0.17 (0.21) − 0.69 (0.74) *** *** WOM_tp − 0.28 (0.09) − 1.16 (0.43) WOM_rn − 0.01 (0.06) 0.09 (0.22) Robustness checks WOM_nn − 0.04 (0.10) − 0.19 (0.41) To check the robustness of binary logistic estimates, WOM_cn − 0.11 (0.08) − 0.38 (0.30) OLS and probit methods were employed to exam- ** *** WOM_on 0.41 (0.19) 1.79 (0.60) ine the impact of OBs and WOMs on farmers’ use of WOM_tn 0.02 (0.06) 0.11 (0.26) biogas, and the moderating role of interpersonal trust. ** OB_r × Tr 0.09 (0.05) 0.61 (0.24) The results of the obtained robustness are shown in OB_n × Tn 0.08 (0.05) 0.12 (0.32) Table 3. They indicate that the OB obtained from tech - OB_c × Tc 0.02(0.06) 0.04(0.25) nical instructors significantly influences the use behav - OB_o × To − 0.08(0.07) − 0.17(0.24) ior and that interpersonal trust does moderate the OB_t × Tt − 0.00(0.06) 0.04(0.23) relationships between OBs and use of biogas by farm- WOM_rp × Tr 0.00(0.10) 0.20 (0.34) ers, and between WOMs and use of biogas by farmers. ** ** WOM_np × Tn 0.24 (0.11) 0.94 (0.38) These results further support credibility of the binary WOM_cp × Tc − 0.12(0.10) − 0.24 (0.39) logistic estimates present in Table 2. ** ** WOM_op × To 0.51 (0.23) 1.98 (0.84) *** *** WOM_tp × Tt 0.42 (0.13) 1.85 (0.62) WOM_rn × Tr − 0.04 (0.10) − 0.24 (0.36) Discussion WOM_nn × Tn − 0.14 (0.11) − 0.61 (0.43) As a final discussion, after trying to explore the role of ** WOM_cn × Tc − 0.07 (0.09) − 0.91 (0.45) interpersonal trust in the effects of social learning on * *** WOM_on × To − 0.41 (0.21) − 1.96 (0.72) farmers’ use of residue-based biogas, this paper obtains WOM_tn × Tt − 0.07 (0.09) − 0.54 (0.48) some general and main findings. The general find - Interpersonal trust variables Yes Yes ings are drawn as follows. Analysis of this survey data Control variables Yes Yes reveals that only 22% of the farmers surveyed used ** Constant 0.17 (0.11) − 1.46 (0.58) residue-based biogas, implying that the use rate is rel- Prob > F/ Prob > chi 0.00 0.00 atively low in rural Hubei, China. A large number of farmers receive OB from neighbors. Compared to the *** ** * N = 913; p < 0.01, p < 0.05, p < 0.1; standard errors are in parentheses; landform dummies are included in OLS and Probit estimates with plains as the number of farmers who had access to OB, the number reference of farmers who obtained positive or negative WOMs is much less. Comparatively, interpersonal trust is gen- especially from technical instructors, they are more likely erally high, and more than 70% of farmers trust their to utilize biogas, and the demonstration activities are neighbors. much more effective. Thus, it is necessary for local gov - ernments and their development partners to design and Main findings and policy implications implement strategies aimed at encouraging farmers to First, the empirical results presented here show that OB obtain OB from technical instructors. from technical instructors positively influences farmers’ Second, the results reveal that interpersonal trust use of residue-based biogas. This implies that acquisi - strengthens the relationship between OB from relatives tion of OB from technical instructors can help promote and use behavior by farmers. That is, without farmers’ adoption of residue-based biogas. Farmers may be uncer- trust in relatives, the transfer of technology from relatives tain about the utility and return effects of using residue- through OB may be ineffective. Farmers are generally risk based biogas. However, if they obtain OB from others, Zeng et al. Energy, Sustainability and Society (2022) 12:21 Page 12 of 16 averse [62]. Thus, if they do not trust others, they are less on biogas use from different subjects (i.e., relatives, likely to use residue-based biogas based on knowledge neighbors, cadres, cooperative members and technical acquired via OB. Therefore, implementation of policies instructors). It reveals that OB from technical instruc- meant to increase farmers’ use of residue-based biogas tors positively influences farmers’ use of residue-based based on knowledge acquired through OB, policymakers biogas, and that interpersonal trust strengthens the rela- and practitioners should first pay attention to the inter - tionships between OB from relatives and use behavior by personal trust levels among farmers. Mechanisms aimed farmers, and between positive WOM and use behavior, at promoting mutual trust between farmers and relatives but it weakens the relationship between negative WOM should be developed and implemented. and the use behavior. These findings of this paper con - Third, this empirical study shows that interpersonal trust tribute to our understanding in the moderating role of improves the relationship between positive WOM and use interpersonal trust in the investigation of the impact of behavior. However, it weakens the relationship between social learnings and to enriching the research field of negative WOM and the use behavior. When farmers trust farmers’ use of biogas. others, dissemination of knowledge related to diffuse residue-based biogas via positive WOM becomes more effective. In contrast, when farmers trust others, acquisi - Appendix 1 tion of knowledge about residue-based biogas via negative See Table 4. WOM may fail. Therefore, policy practitioners should be cautious when implementing measures to promote biogas Table 4 Multicollinearity test results adoption through WOMs combined with interpersonal Variables VIF 1/VIF trust. Communication platforms should be built to inspire farmers who trust in others to actively participate in posi- OB_r 2.04 0.49 tive WOM. In addition, policy practitioners should advice OB_n 3.06 0.33 farmers who trust others to critically value the information OB_c 3.24 0.31 obtained from negative WOM. OB_o 2.56 0.39 OB_t 2.59 0.40 WOM_rp 2.43 0.41 Limitations and outlook for future research WOM_np 2.47 0.41 Despite the interesting results presented in this study, we WOM_cp 1.74 0.58 acknowledge that the generalizability of these results to WOM_op 1.83 0.55 the national level should be carefully considered since WOM_tp 1.95 0.51 the study sample was only derived from of rural Hubei, WOM_rn 2.80 0.36 China. Nevertheless, this study provides foundational WOM_nn 2.68 0.37 data that can be used to establish national data. In addi- WOM_cn 1.75 0.57 tion, owing to the lack of time series data, this study WOM_on 1.64 0.61 was based on one-year data which do not capture the WOM_tn 1.65 0.61 dynamic impacts of OB and WOM on technology adop- Tr 1.14 0.87 tion. Therefore, future empirical studies should aim to Tn 1.04 0.96 analyzing panel data to examine these dynamic effects. Tc 1.28 0.78 Furthermore, with data unavailable, we could not include To 1.26 0.79 the potential factors influencing residue-based biogas Tt 1.29 0.78 use such as the volume of agricultural waste production Gender 1.24 0.81 in our study. Future studies may explore the impacts of Age 1.29 0.78 these potential factors on residue-based biogas use with Education 1.46 0.68 relevant data at hand. Labor 1.14 0.88 Household income 1.17 0.86 Conclusions Subsidy 1.11 0.90 This paper is the first to incorporate the moderating role Risk perception 1.19 0.84 of interpersonal trust into the effects of social learning Cost-effective perception 1.16 0.86 on farmers’ use of residue-based biogas. Using data from Hill 1.24 0.81 representative household-based surveys in rural Hubei Mountains 1.24 0.81 China comprising 913 farmers, we empirically exam- N = 913; the VIF scores range from 1.04 to 3.24, and the multicollinearity test ined and distinguished the impacts of OBs and WOMs result was acceptable in our study Z eng et al. Energy, Sustainability and Society (2022) 12:21 Page 13 of 16 Appendix 2 See Table 5. Table 5 Evaluation statistics of the logistic model Variables Model 1 Model 2 Model 3 Model 4 Model 5 ** ** OB_r 0.64 (0.27) 0.22 (0.30) 0.67 (0.28) 0.26 (0.31) ** ** OB_n 0.83 (0.36) 0.56 (0.67) 0.86 (0.37) 0.81 (0.69) OB_c − 0.39 (0.35) − 0.30 (0.44) − 0.45 (0.37) − 0.40 (0.45) OB_o − 0.22 (0.32) − 0.05 (0.44) − 0.28 (0.33) − 0.05 (0.44) *** ** *** ** OB_t 0.94 (0.33) 0.79 (0.39) 1.09 (0.34) 0.97 (0.40) WOM_rp 0.25 (0.29) 0.23(0.29) − 0.01 (0.36) 0.02 (0.35) *** *** WOM_np 1.14 (0.30) 1.19 (0.31) 0.02 (0.65) 0.06 (0.65) *** *** ** *** WOM_cp 0.99 (0.34) 1.00 (0.33) 1.27 (0.49) 1.29 (0.49) * * WOM_op 1.11 (0.60) 1.08 (0.59) − 0.97 (1.23) − 1.19 (1.29) *** *** WOM_tp − 0.78 (0.58) − 0.73 (0.56) − 2.34 (0.80) − 2.20 (0.79) WOM_rn 0.02 (0.34) 0.04 (0.34) 0.13 (0.38) 0.15 (0.38) *** *** WOM_nn − 1.01 (0.34) − 1.06 (0.34) − 0.25 (0.76) − 0.27 (0.75) ** ** WOM_cn − 0.87 (0.38) − 0.91 (0.38) − 0.76 (0.53) − 0.74 (0.53) *** *** WOM_on 0.72 (0.62) 0.62 (0.62) 3.10 (1.00) 3.12 (1.03) WOM_tn − 0.25 (0.41) − 0.19 (0.41) 0.14 (0.47) 0.17 (0.46) ** ** OB_r × Tr 1.11 (0.44) 1.12 (0.45) OB_n × Tn 0.29 (0.63) 0.03 (0.65) OB_c × Tc 0.03 (0.44) 0.07 (0.45) OB_o × To − 0.23 (0.42) − 0.31 (0.43) OB_t × Tt 0.19 (0.40) 0.11 (0.41) WOM_rp × Tr 0.56 (0.60) 0.42 (0.60) ** ** WOM_np × Tn 1.58 (0.70) 1.56 (0.69) WOM_cp × Tc − 0.43 (0.70) − 0.48 (0.68) ** ** WOM_op × To 3.33 (1.43) 3.48 (1.47) *** *** WOM_tp × Tt 3.60 (1.18) 3.36 (1.14) WOM_rn × Tr − 0.42 (0.63) − 0.38 (0.64) WOM_nn × Tn − 1.06 (0.79) − 1.09 (0.78) * * WOM_cn × Tc − 1.47 (0.82) − 1.54 (0.79) *** *** WOM_on × To − 3.22 (1.21) − 3.34 (1.22) WOM_tn × Tt − 1.06 (0.88) − 0.92 (0.85) * *** *** Tr − 0.33 (0.19) − 0.23 (0.20) − 0.98 (0.37) − 0.36 (0.25) − 1.08 (0.31) *** *** * *** * Tn 1.07 (0.22) 1.25 (0.27) 1.02 (0.55) 0.96 (0.31) 0.95 (0.57) Tc − 0.03 (0.19) 0.12 (0.21) 0.11 (0.37) 0.23 (0.24) 0.22 (0.38) To − 0.24 (0.19) − 0.32 (0.21) − 0.24 (0.29) − 0.44 (0.23) − 0.33 (0.30) Tt − 0.16 (0.19) − 0.06 (0.21) − 0.14 (0.281) − 0.19 (0.23) − 0.24 (0.29) Gender − 0.16 (0.19) − 0.10 (0.21) − 0.08 (0.21) − 0.19 (0.21) − 0.17 (0.22) Age − 0.00 (0.01) 0.01 (0.01) 0.00 (0.01) 0.01 (0.01) 0.00 (0.01) *** ** ** ** ** Education 0.08 (0.03) 0.08 (0.03) 0.07 (0.03) 0.09 (0.03) 0.08 (0.03) Labor 0.04 (0.06) 0.06 (0.07) 0.06 (0.07) 0.07 (0.07) 0.07 (0.07) Household income − 0.00 (0.01) − 0.01 (0.01) − 0.01 (0.01) − 0.01 (0.01) − 0.01 (0.01) ** Subsidy 0.44 (0.19) 0.25 (0.22) 0.29 (0.22) 0.23 (0.23) 0.27 (0.23) *** *** *** *** *** Risk perception − 0.46 (0.09) − 0.29 (0.10) − 0.29 (0.10) − 0.31 (0.10) − 0.32 (0.10) Cost-effective perception 0.18 (0.10) 0.09 (0.11) 0.08 (0.12) 0.09 (0.12) 0.09 (0.12) *** *** *** *** *** Hill − 0.69 (0.19) − 0.79 (0.22) − 0.76 (0.22) − 0.789 (0.23) − 0.79 (0.23) Mountains − 0.60 (0.40) − 0.57 (0.49) 0.64 (0.49) − 0.65 (0.53) − 0.72 (0.52) * *** *** *** ** Constant − 1.39 (0.83) − 3.43 (0.89) − 2.78 (1.01) − 3.05 (0.93) − 2.66 (1.06) Log likelihood − 426.72 − 369.44 − 377.00 − 356.81 − 353.11 Zeng et al. Energy, Sustainability and Society (2022) 12:21 Page 14 of 16 Table 5 (continued) Variables Model 1 Model 2 Model 3 Model 4 Model 5 Pseudo-r 0.11 0.23 0.23 0.25 0.26 Prob > chi 0.00 0.00 0.00 0.00 0.00 *** *** *** *** Likelihood-ratio test 114.57 122.41 139.82 147.23 AUC 0.73 0.82 0.82 0.83 0.84 AIC 885.45 800.87 803.04 795.63 798.21 *** ** * N = 913; p < 0.01, p < 0.05, p < 0.1; standard errors are in parentheses; terrain dummies are included in all models with plains as the reference Received: 28 May 2021 Accepted: 5 May 2022 Supplementary Information The online version contains supplementary material available at https:// doi. org/ 10. 1186/ s13705- 022- 00350-8. Additional file 1. Part of the survey questionnaire. References 1. Kabyanga M, Balana BB, Mugisha J, Walekhwa PN, Smith J, Glenk K (2018) Are smallholder farmers willing to pay for a flexible balloon biogas Acknowledgements digester? Evidence from a case study in Uganda. Energy Sustain Dev The authors gratefully acknowledge financial support from the National 43:123–129. https:// doi. org/ 10. 1016/j. esd. 2018. 01. 008 Natural Science Foundation of China (72103055). This paper is also supported 2. Meeks R, Sims KRE, Thompson H (2019) Waste not: can household biogas by the scholarship granted by the China Scholarship Council (CSC) and by deliver sustainable development? Environ Resour Econ 72(3):763–794. the General Project of Hubei Social Science Foundation, China (2021191). The https:// doi. org/ 10. 1007/ s10640- 018- 0224-1 authors are very grateful to the editors and the anonymous reviewers of this 3. Kumar JCR, Majid MA (2020) Renewable energy for sustainable develop- journal who provided very valuable feedback on this paper. In addition, the ment in India: current status, future prospects, challenges, employment, authors thank the Home for Researchers Editorial Team for providing language and investment opportunities. Energy Sustain Soc 10:2. https:// doi. org/ editing service, which improved the readability and clarity of this manuscript. 10. 1186/ s13705- 019- 0232-1 4. Clemens H, Bailis R, Nyambane A, Ndung’u V (2018) Africa biogas partner- Author contributions ship program: a review of clean cooking implementation through market YZ was the main author responsible for the design, data collection, analysis, development in East Africa. Energy Sustain Dev 46:23–31. https:// doi. org/ and interpretation of the results of this study. FQ complemented the discus- 10. 1016/j. esd. 2018. 05. 012 sion related to the cultural theory and provided relevant feedback for the 5. Rahman MS, Majumder MK, Sujan MHK (2021) Adoption determinants of improvement of the article’s clarity. JZ complemented the design and data biogas and its impact on poverty in Bangladesh. Energy Rep 7:5026– collection. All authors read and approved the final manuscript. 5033. https:// doi. org/ 10. 1016/j. egyr. 2021. 08. 027 6. Berhe M, Hoag D, Tesfay G, Keske C (2017) Factors influencing the adop - Funding tion of biogas digesters in rural Ethiopia. Energy Sustain Soc 7:10. https:// This study is financially support by the National Natural Science Foundation doi. org/ 10. 1186/ s13705- 017- 0112-5 of China (72103055), by the scholarship granted by the China Scholarship 7. Pestalozz J, Bieling C, Scheer D, Kropp C (2019) Integrating power-to-gas Council (CSC) and by the General Project of Hubei Social Science Foundation, in the biogas value chain: analysis of stakeholder perception and risk gov- China (2021191). ernance requirements. Energy Sustain Soc 9:38. https:// doi. org/ 10. 1186/ s13705- 019- 0220-5 Availability of data and materials 8. Fletschner D, Mesbah D (2011) Gender disparity in access to information: The datasets generated and analyzed in this paper are not publicly available do spouses share what they know? World Dev 39(8):1422–1433. https:// due to confidentiality reasons. However, the transcripts of the anonymous doi. org/ 10. 1016/j. world dev. 2010. 12. 014 interviews are available on reasonable request. 9. Kashif M, Awan MB, Nawaz S, Amjad M, Talib B, Farooq M, Nizami AS, Rehan M (2020) Untapped renewable energy potential of crop residues in Pakistan: challenges and future directions. J Environ Manag Declarations 256:109924. https:// doi. org/ 10. 1016/j. jenvm an. 2019. 109924 10. Munshi K (2004) Social learning in a heterogeneous population: technol- Ethics approval and consent to participate ogy diffusion in the Indian Green Revolution. J Dev Econ 73:185–215. Not applicable. https:// doi. org/ 10. 1016/j. jdeve co. 2003. 03. 003 11. Conley TG, Udry CR (2010) Learning about a new technology: pineapple Consent for publication in Ghana. Am Econ Rev 100(1):35–69. http://www.aeaweb.org/articles. Not applicable. php?doi=https:// doi. org/ 10. 1257/ aer. 100.1. 35 12. Krishnan P, Patnam M (2013) Neighbors and extension agents in Competing interests Ethiopia: who matters more for technology adoption? Am J Agric Econ The authors declare that they have no competing interests. 96(1):308–327. https:// doi. org/ 10. 1093/ ajae/ aat017 13. Maertens A (2017) Who cares what others think (or do)? Social learn- Author details 1 ing and social pressures in cotton farming in India. Am J Agric Econ School of Economics and Management, Hubei University of Technol- 2 99(4):988–1007. https:// doi. org/ 10. 1093/ ajae/ aaw098 ogy, 28 Nanlilu Street, Hongshan, Wuhan 430068, Hubei, China. F acult y 14. Cai H, Chen Y, Fang H (2009) Observational learning: evidence from a ran- of Agricultural, Life and Environmental Science, University of Alberta, General 3 domized natural field experiment. Am Econ Rev 99(3):864–882. https:// Services Building, 9007 116 St NW, Edmonton, AB T6G 2H1, Canada. College doi. org/ 10. 1257/ aer. 99.3. 864 of Economics and Management, Huazhong Agricultural University, 1 Shizishan Street, Hongshan, Wuhan 430070, Hubei, China. Z eng et al. Energy, Sustainability and Society (2022) 12:21 Page 15 of 16 15. Campbell A (2013) Word-of-mouth communication and percolation in Tigray, Ethiopia. Energy Econ 66:272–278. https:// doi. org/ 10. 1016/j. eneco. social networks. Am Econ Rev 103(6):2466–2498. https:// doi. org/ 10. 1257/ 2017. 06. 022 aer. 103.6. 2466 35. Reise C, Musshoff O, Granoszewsk K, Spiller A (2012) Which factors influ- 16. Ameri M, Honka E, Xie Y (2016) Word-of-mouth, observational learning, ence the expansion of bioenergy? An empirical study of the investment and product adoption: evidence from an anime platform. SSRN Electron behaviours of German farmers. Ecol Econ 73:133–141. https:// doi. org/ 10. J. https:// doi. org/ 10. 2139/ ssrn. 27989 621016/j. ecole con. 2011. 10. 008 17. Bandura A (1977) Social learning theory. Prentice Hall, Englewood Cliffs, 36. Muvhiiwa R, Hildebrandt D, Chimwani N, Ngubevana L, Matambo T NJ (2017) The impact and challenges of sustainable biogas implementation: 18. Berger J (2014) Word of mouth and interpersonal communication: a moving towards a bio-based economy. Energy Sustain Soc 7:20. https:// review and directions for future research. J Consum Psychol 24(4):586–doi. org/ 10. 1186/ s13705- 017- 0122-3 607. https:// doi. org/ 10. 1016/j. jcps. 2014. 05. 002 37. Ministry of Agriculture and Rural Areas of the People’s Republic of China. 19. Wang Y, Yu C (2017) Social interaction-based consumer decision-making 2016-08-11. Pilot Scheme of Promoting the Resource Utilization of Agri- model in social commerce: the role of word of mouth and observational cultural Wastes. http:// www. moa. gov. cn/ govpu blic/ FZJHS/ 201609/ t2016 learning. Int J Inf Manag 37:179–189. https:// doi. org/ 10. 1016/j. ijinf omgt. 0919_ 52778 46. htm 2015. 11. 005 38. National Development and Reform Commission. 2017a-01-25. National 20. Jones R, Audi PA, Tripp R (2001) The role of informal seed systems in dis- 13th five-year plan for rural biogas development. https:// www. ndrc. gov. seminating modern varieties: the example of pigeonpea from a semi-arid cn/ fzggw/ jgsj/ njs/ sjdt/ 201702/ t2017 0210_ 11948 08. html area of Kenya. Exp Agric 37:539–548. https:// doi. org/ 10. 1017/ s0014 47970 39. National Development and Reform Commission. 2017b-12–28. Guiding 10004 61 Opinions on the Construction of Straw Gasification Clean Energy Utiliza- 21. Mauceri M, Alwang J, Norton G, Barrera VH (2007) Eec ff tiveness of tion Project. https:// www. ndrc. gov. cn/ xxgk/ zcfb/ tz/ 201801/ t2018 0102_ integrated pest management dissemination techniques: a case study 962631. html of potato farmers in Carchi, Ecuador. J Agric Appl Econ 39(3):765–780. 40. General Office of Hubei Provincial People’s Government. 2019-06-25. https:// doi. org/ 10. 1017/ s1074 07080 00234 03 Evaluation Report on the Performance of “Special Project for Modern 22. Zilberman D, Zhao J, Heiman A (2012) Adoption versus adaptation, with Agricultural and Rural Energy Development” in 2018. http:// nyt. hubei. emphasis on climate change. Annu Rev Econ 4:27–53. https://www.gov. cn/ fbjd/ xxgkml/ czgk/ ysjx/ 201906/ t2019 0625_ 123499. shtml annualreviews.org/doi/pdf/https:// doi. org/ 10. 1146/ annur ev- resou 41. General Office of Hubei Provincial People’s Government. 2007–4–17. rce- 083110- 115954 Opinions on vigorously promoting biogas construction in rural areas. 23. Mobius M, Rosenblat T (2014) Social learning in economics. Annu http:// www. hubei. gov. cn/ zfwj/ ezbf/ 201112/ t2011 1210_ 17125 55. shtml Rev Econ 6:827–847. https:// doi. org/ 10. 1146/ annur ev- econo 42. Han H, Wu S (2018) Rural residential energy transition and energy con- mics- 120213- 012609 sumption intensity in China. Energy Econ 74:523–534. https:// doi. org/ 10. 24. Oreszczyn S, Lane A, Carr S (2010) The role of networks of practice and 1016/j. eneco. 2018. 04. 033 webs of influencers on farmers’ engagement with and learning about 43. Christiaensen L, Heltberg R (2014) Greening China’s rural energy: new agricultural innovations. J Rural Stud 26(4):404–417. https:// doi. org/ 10. insights on the potential of smallholder biogas. Environ Dev Econ 1016/j. jrurs tud. 2010. 03. 003 19(1):8–29. https:// doi. org/ 10. 1017/ s1355 770x1 30003 75 25. Sol J, Beers PJ, Wals AEJ (2013) Social learning in regional innovation 44. Sun D, Bai J, Qiu H, Cai Y (2014) Impact of government subsidies on networks: trust, commitment and reframing as emergent properties of household biogas use in rural China. Energy Policy 73:748–756. https:// interaction. J Clean Prod 49:35–43. https:// doi. org/ 10. 1016/j. jclep ro. 2012. doi. org/ 10. 1016/j. enpol. 2014. 06. 009 07. 041 45. He P, Veronesi M (2017) Personality traits and renewable energy technol- 26. Joffre OM, De Vries JR, Klerkx L, Poortvliet PM (2020) Why are cluster ogy adoption: a policy case study from China. Energy Policy 107:472–479. farmers adopting more aquaculture technologies and practices? The role https:// doi. org/ 10. 1016/j. enpol. 2017. 05. 017 of trust and interaction within shrimp farmers’ networks in the Mekong 46. Si S, Lyu M, Lawell CYC, Chen S (2018) The effects of energy-related poli- Delta, Vietnam. Aquaculture 523:735181. https:// doi. org/ 10. 1016/j. aquac cies on energy consumption in China. Energy Econ 76:202–227. https:// ulture. 2020. 735181doi. org/ 10. 1016/j. eneco. 2018. 10. 013 27. Kropf B, Schmid E, Schönhart M, Mitter H (2020) Exploring farmers’ behav- 47. Senaviratna NAMR, Cooray TMJA (2019) Diagnosing multicollinearity of ior toward individual and collective measures of Western Corn Rootworm logistic regression model. Asian J Prob Stat 5(2):1–9. https:// doi. org/ 10. control—a case study in south-east Austria. J Environ Manag 264:110431. 9734/ ajpas/ 2019/ v5i23 0132 https:// doi. org/ 10. 1016/j. jenvm an. 2020. 110431 48. Mengistu MG, Simane B, Eshete G, Workneh TS (2016) Factors affect - 28. Mahajan A, Bishop JW, Scott D (2012) Does trust in top management ing households’ decisions in biogas technology adoption, the case of mediate top management communication, employee involvement and Ofla and Mecha Districts, northern Ethiopia. Renew Energy 93:215–227. organizational commitment relationships. J Manag Issues 24(2):173–190. https:// doi. org/ 10. 1016/j. renene. 2016. 02. 066 https:// www. jstor. org/ stable/ 43488 132 49. Campbell RJ, Jeong SH, Graffin SD (2019) Born to take risk? The effect of 29. Jin S, Bluemling B, Mol APJ (2015) Information, trust and pesticide ceo birth order on strategic risk taking. Acad Manag J 62(4):1278–1306. overuse: interactions between retailers and cotton farmers in China. https:// doi. org/ 10. 5465/ ambpp. 2018. 186 NJAS-Wagen J Life Sc 72:23–32. https:// doi. org/ 10. 1016/j. njas. 2014. 10. 003 50. Hoetker G (2007) The use of logit and probit models in strategic manage- 30. Buck S, Alwang J (2011) Agricultural extension, trust, and learning: results ment research: critical issues. Strat Manag J 28:331–343. https:// doi. org/ from economic experiments in Ecuador. Agric Econ 42(6):685–699. 10. 1002/ smj. 582 https:// doi. org/ 10. 1111/j. 1574- 0862. 2011. 00547.x 51. Brambor T, Clark WR, Golder M (2006) Understanding interaction models: 31. Hunecke C, Engler A, Jara-Rojas R, Poortvliet PM (2017) Understanding improving empirical analyses. Polit Anal 14:63–82. https:// doi. org/ 10. the role of social capital in adoption decisions: an application to irrigation 1093/ pan/ mpi014 technology. Agric Syst 153:221–231. https:// doi. org/ 10. 1016/j. agsy. 2017. 52. Zeng Y, Zhang J, He K (2019) Eec ff ts of conformity tendencies on house - 02. 002 holds’ willingness to adopt energy utilization of crop straw: evidence 32. Wang H, Yu F, Reardon T, Huang J, Roelle S (2013) Social learning and from biogas in rural China. Renew Energy 138:573–584. https:// doi. org/ parameter uncertainty in irreversible investments: evidence from green-10. 1016/j. renene. 2019. 02. 003 house adoption in northern China. China Econ Rev 27:104–120. https:// 53. Mittal S, Ahlgren EO, Shukla PR (2018) Barriers to biogas dissemination doi. org/ 10. 1016/j. chieco. 2013. 09. 003 in India: a review. Energy Policy 112:361–370. https:// doi. org/ 10. 1016/j. 33. Chen L, Zhao L, Ren C, Wang F (2012) The progress and prospects of rural enpol. 2017. 10. 027 biogas production in China. Energy Policy 51:58–63. https:// doi. org/ 10. 54. Ziegler A (2017) Political orientation, environmental values, and climate 1016/j. enpol. 2012. 05. 052 change beliefs and attitudes: an empirical cross country analysis. Energy 34. Kelebe HE, Ayimut KM, Berhe GH, Hintsa K (2017) Determinants for adop- Econ 63:144–153. https:// doi. org/ 10. 1016/j. eneco. 2017. 01. 022 tion decision of small scale biogas technology by rural households in Zeng et al. Energy, Sustainability and Society (2022) 12:21 Page 16 of 16 55. Franken JRV, Pennings JME, Garcia P (2018) Graphical illustration of inter- action effects in binary choice models: a note. J Agric Econ 69(3):852– 858. https:// doi. org/ 10. 1111/ 1477- 9552. 12257 56. Williams M (2001) In whom we trust: Group membership as an affective context for trust development. Acad Manage Rev 26(3):377–396. https:// doi. org/ 10. 2307/ 259183 57. Abrams LC, Cross R, Lesser E, Levin DZ (2003) Nurturing interpersonal trust in knowledge-sharing networks. Acad Manag Perspect 17:64–77. https:// www. jstor. org/ stable/ 41660 07 58. Ma W, Renwick A, Yuan P, Ratna N (2018) Agricultural cooperative mem- bership and technical efficiency of apple farmers in China: an analysis accounting for selectivity bias. Food Policy 81:122–132. https:// doi. org/ 10. 1016/j. foodp ol. 2018. 10. 009 59. Liang Q, Huang Z, Lu H, Wang X, Ifamr I (2015) Social capital, member par- ticipation, and cooperative performance: Evidence from China’s Zhejiang. Int Food Agribus Man Assoc 18(1):49–77. https:// www. resea rchga te. net/ publi cation/ 28176 2869 60. Oi JC, Rozelle S (2000) Elections and power: the locus of decision-making in Chinese villages. China Q 162:513–539. https:// doi. org/ 10. 1017/ S0305 74100 00082 37 61. Yip W, Hsiao W (2009) China’s health care reform: a tentative assessment. China Econ Rev 20(4):613–619. https:// doi. org/ 10. 1016/j. chieco. 2009. 08. 62. Wang HH, Zhang YP, Wu L (2011) Is contract farming a risk management instrument for Chinese farmers? China Agric Econ Rev 3(4):489–504. https:// doi. org/ 10. 1108/ 17561 37111 11923 47 Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in pub- lished maps and institutional affiliations. Re Read ady y to to submit y submit your our re researc search h ? Choose BMC and benefit fr ? Choose BMC and benefit from om: : fast, convenient online submission thorough peer review by experienced researchers in your field rapid publication on acceptance support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year At BMC, research is always in progress. Learn more biomedcentral.com/submissions

Journal

"Energy, Sustainability and Society" – Springer Journals

Published: May 16, 2022

There are no references for this article.