The Impact of Consumer Participation Certification on the Trust of Eco-Agricultural Products Based on the Mediating Effects of Information and Identity

Abstract

With the increasing distrust of food safety, both third-party certification systems (TPC) and participatory guarantee systems (PGS) play a vital role in restoring consumer trust. Although the fact that previous research has focused on consumer trust and the factors that impact it in TPC products, little emphasis has been made on how consumer participation in certification affects trust. The goal of the study was to explore how consumer participation certification affects trust in eco-agricultural products under PGS. We constructed a theoretical framework of consumer trust in eco-agricultural products under PGS, based on consumer trust theory, and clarified the relationship between consumer participation certification, information quality, social identity, and consumer trust. After obtaining 238 valid questionnaires on consumers from 12 PGS organizations nationwide, a structural equation model (SEM) was conducted. The conclusions are as follows: (1) Consumer participation has a positive impact on consumer trust, and the direct effect is not significant, but the indirect effect is significant; (2) Information quality and social identity have been identified to play full intermediary roles in the relationship between consumer participation and trust. We suggest relevant research implications and recommendations for future research on consumer trust in PGS based on the findings.

1. Introduction

Consumer trust is a key factor affecting the purchase willingness of consumers and the size of the market in the sales of ecological agricultural products. At present, consumer trust in ecological agricultural products relies on officially authorized quality labels. These labels are authorized by government departments after an intermediary organization has passed the inspection of products. Quality labels can enhance the trust of ecological agricultural products [1]. In reality, there are some phenomena such as government departments regulating captives, false certification of intermediary agencies, and producer label fraud [2]. As a result, the authority of the third-party certification has been called into question, and most consumers have little trust in intermediary testing results or items labeled “green” or “organic” [3,4].

Consumer involvement certification was advocated by the Participatory Guarantee System (PGS) in response to the consumer trust issue of the 1970s. PGS fosters inter-personal connections in which all participants, including producers and consumers, interact, exchange information, and trust one another as a supplement to third-party certification. [5]. In recent years, PGS has developed well with the number of PGS organizations worldwide, having increased from 376 in 2014 to 235 in 2020. PGS has been established its legal status in some countries, such as India, Brazil, and the Philippines.

In previous studies, scholars have different views on how this certification system can address the problems of consumer trust. Some scholars believe that the reason why the third-party certification system was not worthy of trust was a lack of consumer groups. However, consumers are direct stakeholders. The effective participation of consumers can make up for the lack of supervision of relevant government departments, promote social supervision, restrict food operators, and so enhance consumer trust [6]. However, some scholars have pointed out that it is difficult to establish a stable trust, because consumers trust “human authentication” systems, based on the acquaintance network, not the understanding of the system [7]. Some scholars have said that short-chain agriculture will also reduce information asymmetry and dishonesty of producers [8]. Therefore, we were interested in seeing if consumer participation in certification may raise consumer trust, what factors are important to consider when improving consumer trust, and how to affect consumer trust.

When reviewing the research literature on consumer trust in PGS, it is clear that PGS plays a role in promoting social trust. First of all, PGS has an internal mechanism of mutual supervision, making certification results credible [9,10]. Second, various PGS activities are involved in the process of information, knowledge sharing, and social network formation of all participants, which promote social trust [11,12]. Third, PGS has achieved a close connection between producers and consumers, and consumers can obtain more open, transparent, and real information, thus improving the perceived value of products or services and enhancing trust [13,14,15,16]. In summary, from the research content, there is no direct research analyzing PGS consumer trust mechanisms. Consumer participation, information quality, and social identity are thought to be major sources of consumer trust, with consumer participation certification enhancing trust via knowledge exchange and social interaction. Although previous research has provided clues for the formation of PGS consumer trust, there is a lack of quantitative empirical research, and these hypotheses need to be verified by survey data.

Based on the above, this paper aimed to conduct an empirical study to testify the impact of consumer participation certification on consumer trust in the context of the participatory guarantee system, and the intermediary roles of information quality and social identity. To this end, we put forward research hypotheses that consumer participation positively affects consumer trust by affecting information quality and social identity at first. The questionnaires on consumer participation in PGS certification were designed and distributed to consumers in the Chinese PGS organization. Subsequently, a structural equation model (SEM) was used to verify the hypotheses proposed in this study. Finally, the total effect, direct effect, and mediation effects are discussed, and the implications and outlooks are pointed out. This study was conducted from June 2021 to January 2022.

The remaining parts of the study are structured as follows. Section 2 constructs the theoretical model of consumer trust and puts forward the main hypotheses. Section 3 designs the questionnaire, and selects the data analysis method. Section 4 provides the empirical results of our survey conducted in China. Section 5 discusses the enlightenment of the results in theory and practice. Section 6 presents the conclusion, limitations, and future research.

2. Conceptual Framework and Research Hypotheses Development

2.1. Conceptual Framework

Even though there are few studies on the link between consumer participation and consumer trust under PGS, several studies on consumer participation and consumer trust have been conducted on the subject of food safety. This is the theoretical basis of this study.

2.1.1. Theoretical Research on Consumer Participation

Consumer participation is the sum of consumer behaviors in the process of participating in the provision of products or services [17,18]. Ennew’s view is widely recognized by the academic community, which divides consumer participation behavior into three dimensions: information sharing, responsible behavior, and personal interaction [19]. Information sharing is when “needed information or knowledge”, obtained by consumers through various channels, can be transmitted to their partners, reflecting the important role of consumers as information contributors. Responsibility behavior is the duty that consumers have to perform during the interaction process. Personal interaction, defined as consumer and partner support, communication, and coordination, can strengthen relationships and enhance communication with other members [20,21].

From the content of consumer participation, with the development of modern information technology and the needs of the practice, the depth and breadth of consumer participation are constantly improving. It has been experienced from productive labor to marketing activities, as well as new product research and enterprise value creation [22]. Some scholars extended consumer participation to the field of food safety, pointing out that consumers can play a significant role in food safety governance, and then have studied consumer participation guarantee systems in food safety governance [6].

From the influence of consumer participation, some scholars have pointed out that consumer participation behavior can meet their functional and social needs, according to the theories of consumer needs [23]. On the one hand, consumer participation may provide consumers with detailed and accurate product information. Consumers who are at a disadvantage in terms of information will find that the more information they obtain, the easier it will be to buy satisfactory products [24,25,26,27]. On the other hand, consumer participation behavior can produce positive recognition and a sense of belonging in an interpersonal environment with security and belonging, and so this helps to close the relationship with others [20,28]. Therefore, as a PGS stakeholder, the consumer can realize interpersonal interaction and information exchange while participating in certification [29].

2.1.2. Theoretical Research on Consumer Trust

From the perspective of modern sociology, scholars have studied interpersonal trust, which is based on interpersonal relationships. Interpersonal trust is defined as the degree of confidence in interpersonal relationships occurring in a specific social situation, and it is determined by rational calculation and emotional association in interpersonal interactions [30]. Thus, interpersonal trust is divided into two dimensions: cognitive trust and affective trust [31]. Cognitive trust is a cognitive judgment based on the other party’s ability and reliability, which is related to their ability, expertise, and reliability. Whether there is enough information matters with regards to consumer trust. Affective trust is a belief formed by both sides based on emotion, which involves the relationship itself. The reliance on identity and emotion makes both parties maintain positive interactions, limit opportunism, and build trust [31].

The theoretical basis of interpersonal trust shows that information and identity are two important perspectives affecting consumer trust. One of the perspectives is the information perspective. In real life, cognition often has a more significant impact on people’s behavior [32]. According to the “knowledge loss model”, information can improve trust, while distrust may be induced by a lack of information or misperceived information [33,34]. Consumers’ risk perception and trust are influenced by the amount of information they have. Scholars have pointed out that an important reason why consumers do not trust food safety is insufficient information disclosure and unreliable information authenticity [35,36]. Therefore, information quality is a critical aspect in establishing consumer trust. Consumer perceptions of product uncertainty can be eased to some extent by the richness, accuracy, authority, and readability of information, which can enhance consumer trust.

Another perspective is identity. People tend to establish a trusting relationship with people similar to themselves, and thus identity is an important step to building trust [37]. According to the theory of social identity, this means that individuals have a sense of belonging to a social group, feel similar to other members, and can perceive the feelings and value significance brought by the group [38]. Trust based on social identity is reflected in three aspects. First, when consumers focus on their social status, the perception of relevant groups or products will be strengthened [39]. Secondly, consumers can identify the social status contained in different objects when faced with a choice, so that social identity can affect consumers’ decisions [40]. Finally, it is a process in which people receive social influences that shape consuming behavior primarily through social identity, and in which group and environmental factors play a regulatory role [41]. Previous studies have also confirmed that social identity positively affects consumer trust [42,43]. The stronger a person’s sense of belonging to a social group, the more likely they are to accept or use services or products provided by the group.

2.1.3. Theoretical Research on the Relationship between Consumer Participation and Consumer Trust

Consumer participation is an important factor in studies of consumer food safety trust. Previous studies have explored the impact process of consumer participation on trust from different perspectives, and have confirmed that consumer participation has a direct and significant positive impact on consumer trust in food safety. From the perspective of information sources, such as consumer personal experience, trust comes from the information source led by consumers, and this is because consumers perceive the specific image of food through every contact with food, and gradually establish trust [1,44]. From the perspective of social embedding, the embedding relationship can generate trust, and the interaction can enhance trust based on certain social relations [45]. Embedded relationships can be a direct embedded relationship between producers and consumers, such as consumer social media involvement in community-supported agriculture having a significant, direct, and positive impact on consumer trust [46]. It can also be a third-party embedded relationship in which enterprise employees help to link producers and consumers in traditional food systems, so that the trust in food supply enterprises can also promote consumer trust in food safety [47,48].

2.1.4. Theoretical Model Construction

Consumer participation in this paper refers to the consumer who pays time, energy and money and who participates in the certification process of ecological, agricultural products to investigate and evaluate the products. Consumer trust specifically refers to the consumer trust in the safety of ecological agricultural products approved by PGS certification. It has been shown that not only are information and identity key sources of consumer trust in PGS, but they also play important intermediary roles between consumer participation and trust. Based on the above theoretical analysis, this paper constructed a theoretical framework for the impact of consumer participation certification on the trust of ecological agricultural products, as shown in Figure 1. Consumer participation, information quality, social identity, and consumer trust are the main variables of this study. H1, H2, H3, H4, H5, H6, and H7 are the research hypotheses of this paper. Specifically, through improving information quality and strengthening social identity, consumer participation affects consumer trust in ecological agricultural products under PGS.

2.2. Research Hypotheses

2.2.1. The Effect of Consumer Participation on Consumer Trust

Studies have shown that consumer participation has a direct and positive impact on consumer trust [46]. In the context of the participatory guarantee system, consumers will have an intuitive understanding of the safety of the production process of ecological agricultural products by personally participating in the process of product certification. When compared with the ecological agricultural products labeled as “green” and “organic” purchased from the supermarket, consumers rely more on the products they have inspected and recognized. Accordingly, we hypothesized the following:

Hypothesis 1:

Consumer participation has a direct positive impact on consumer trust.

2.2.2. The Effect of Consumer Participation on Information Quality

Information quality refers to the richness, integrity, accuracy, and authority of agricultural product safety information obtained by consumers. It is shown that the source of information led by consumers, such as consumer personal experience, has strong reliability [1]. Consumers actively seek information by engaging in product certification and forming complete and reliable product knowledge, rather than being fed single, one-sided information.

First, the consumer checks the production environment and technology system, and is aware of the original environment, product features, the standards of inputs, the status of the supply chain, etc. Consumers will also analyze the actual situation to evaluate objectively whether the producers have reliable ecological agricultural technology. Second, the consumer participation certification provides the opportunity to communicate face to face with producers. Consumers can obtain more reliable product information by interaction. Consumers can judge whether producers harm consumers with diseases and pests, and the possibility of its loss, thus considering the reliability of product information received from producers. Third, consumer participation ensures the authenticity and authority of product certification information. To avoid opportunistic behavior by producers, consumers supervise on-site sampling for inspection, especially the collection of soil samples, water samples, or product samples where the most likely pesticide drift. Accordingly, we hypothesized the following:

Hypothesis 2:

Consumer participation has a direct positive impact on information quality obtained by consumers.

2.2.3. The Effect of Consumer Participation on Social Identity

If the individual behavior is consistent with group cognition, the individual will produce satisfaction, maintain behavior, strengthen identity with higher consistency, and take the initiative to communicate with others [38,49]. Consumers find that the social labels they care about are somewhat similar to other members in the process of communication under PGS, such as parents, and then find it easy to identify with the group. The consistency of values among members is prominent, including the awareness of food safety, the consumption attitude of ecological agricultural products, and cognition of the economic, social, and cultural function of ecological agricultural products with continuous communication. Customer perceived value and social identity are improved as a result of the closer relationship between consumer and producer inside the group [34,50]. Accordingly, we hypothesized the following:

Hypothesis 3:

Consumer participation has a direct positive impact on social identity.

2.2.4. The Effect of Information Quality on Consumer Trust

Consumer access to rich and accurate information on food safety as far as possible is the key to establishing consumer trust [51]. In the PGS situation, consumers master more comprehensive and reliable information of ecological agricultural products by objective analysis and correct understanding during the certification process. Therefore, consumers’ risk perceptions are reduced, and consumer trust is enhanced. For example, consumers have a full understanding of the producer technology system, when the prevention and control technology of disease and insect pests is mature, and when management experience is rich. Consumers will believe producers to not use pesticides and insecticides at any time, and be optimistic about producers sticking to ecological production. Accordingly, we hypothesized the following:

Hypothesis 4:

Information quality obtained by consumers has a direct positive impact on consumer trust in ecological agricultural products.

2.2.5. The Effect of Social Identity on Consumer Trust

In addition to the information quality of products obtained by consumers, social identity can also affect consumption decisions [40]. The mechanism of action lies in the process from stimulation to confirmation in a specific situation. Environmental stimulation makes specific social identity and self-concept of the consumer aroused, and then produces consumer trust, followed by the repeated introspection process of social recognition and consumer self-consciousness [52]. In the PGS environment, the social identity generated by consumers can bring trust after communicating and the collision of thinking repeatedly. This is because it gives a person a sense of belonging to the social group, and adopts the group’s emotions or values from the heart. Therefore, consumers prefer to accept the authenticity of ecological agricultural products provided by producers, which leads to consumer trust. Previous studies have confirmed the positive impact of social identity on consumer trust [42,53]. Accordingly, we hypothesized the following:

Hypothesis 5:

Social identity has a direct positive impact on consumer trust in ecological agricultural products.

Based on Hypothesis 2 to Hypothesis 5, two hypotheses are set as follows:

Hypothesis 6:

Information quality mediates between consumer participation and consumer trust.

Hypothesis 7:

Social identity mediates between consumer participation and consumer trust.

3. Methods

3.1. Measurement of Variables

The main variables included in the assumptions are consumer participation, information quality, social identity, and consumer trust. To ensure reliability and validity, we mainly adopted scales, which have been proven to be good tools in past relevant studies. We also adjusted scales for practice in the PGS situation. All the items were evaluated on a 7-point Likert-scale ranging from strongly agree (7) to strongly disagree (1).

Consumer engagement was measured in three dimensions of information sharing, responsible behavior, and interpersonal interaction by Ennew and colleagues [19], and was assessed with a thirteen-item measurement by Kathryn and colleagues [54], Hoffman and Novak [55], and Bettencourt [56]. Information quality was assessed with five items to measure the richness, accuracy, completeness, and authority of product information by Kim and colleagues [57]. Social identity was assessed with a five-item measurement by Algesheimer and colleagues [58], as well as Bagozzi and Dholakia [59]. Consumer trust was measured in two dimensions of cognitive trust and affective trust, and was assessed with a ten-item measurement by McAllister [31], and Ng and Chua [60]. These items were designed to measure to what extent consumers think that ecological agricultural products are trustworthy under PGS. The measurement of the variables and their main sources is shown below

Table 1. Measurement scales and sources.

Variables	Dimension	Item	Source
Consumer participation (CP)	Information sharing (IS)	IS1. I pass along information that may be helpful for the certification work.	Kathryn et al. (2009); Hoffman and Novak (1996); Bettencourt (1997)
		IS2. I share the latest information with other PGS members that may increase the effectiveness of certification.
		IS3. I seek helpful information to share with other PGS members.
		IS4. I share information that I have when it can beneficial to other PGS members.
	Responsibility behavior (RB)	RB1. I am willing to participate in PGS certification activities, such as standard formulation, product testing, inspection, training, etc.
		RB2. I am willing to participate in the preparation and organization of PGS activities.
		RB3. I will respond and suggest as soon as possible when there are problems in the process of PGS certification.
		RB4. I will act with other PGS members to find solutions when there are problems in the process of PGS certification.
		RB5. I will help other PGS consumers to solve their problems and difficulties.
	Personal interaction (PI)	PI1. I am willing to talk with other PGS members, exchange emotions and establish relations with them when participating in certification.
		PI2. I am often involved in the discussion from others PGS members when participating in certification.
		PI3. I often comment on the topics from other PGS members positively.
		PI4. I often make friendly contacts with others PGS members.
Information quality (IQ)	——	IQ1. I can obtain a wealth of product information by participating in certification.	Kim et al. (2008)
		IQ2. I can obtain accurate product information by participating in certification.
		IQ3. I can obtain a piece of more complete product information by participating in certification.
		IQ4. I can obtain authoritative product information by participating in certification.
		IQ5. It is easier for me to understand product information by participating in certification.
Society identity (SI)	——	SI1. I see myself as a part of the PGS, and this identity is important for me.	Algesheimer et al. (2005); Bagozzi and Dholakia (2006)
		SI2. Other PGS members and I share the same objectives.
		SI3. The friendships I have with other PGS members mean a lot to me.
		SI4. I agree with the important role of consumer participation in ensuring the safety of ecological agricultural products.
		SI5. I fully recognize the certification process and results of ecological agricultural products.
		SI6. I would be glad to recommend PGS to those who need it.
Consumer trust (CT)	Cognitive trust (CCT)	CCT1. Through communication with PGS members, I see no reasons to doubt producers’ understanding of Organic.	McAllister (1995); Ng and Chua (2006)
		CCT2. Given product information in the process of certification, I have good reasons to believe in producers’ competence and technology of organic production.
		CCT3. I can rely on producers not to harm my interset by using chemical fertilizer and pesticide when they are faced with plant diseases and insect pests.
		CCT4. Producers discuss with me with professionalism and dedication.
		CCT5. Through communication, I have no reservations about acting on producers’ advice for the purchase of products.
	Affective trust (CAT)	CAT1. The producer’s attitude towards me is enthusiastic when I participate in certification.
		CAT2. If I shared my problems with producers, I know they would respond constructively and caringly.
		CAT3. I would feel lost if I could no longer buy PGS products.
		CAT4. We have a sharing relationship. I can freely share my ideas, feelings, and hopes with producers.
		CAT5. I can talk freely to producers about difficulties I am having in the consumption process, and know that they will want to listen.

.

3.2. Sample and Procedure

The target population analyzed in this study was composed of PGS consumers in China. Although PGS is at the initial development stage, the prototype of its organizational structure has emerged in China. In 2017, Beijing Shunyi Ecological Agriculture Development Association established the PGS National Union, committing to publicize and promote PGS. In recent years, PGS regional organizations have emerged with the support of the National Union. Currently, 20 PGS organizations have been developed in Shaanxi, Anhui, Fujian, Beijing, Hebei, Shandong, Sichuan, Ningxia, Guangdong, Yunnan, Yunnan, Henan, Jiangsu, Hunan, Guizhou, Jilin, Chongqing, Shanghai, and Gansu. Affected by the development time, internal governance, and external environment, the degree of consumer participation among PGS organizations is different. To ensure the quality of the data, this paper selected the PGS organizations with more PGS investigation or certification activities, and then conducted questionnaires on the participating consumers.

To select the appropriate PGS organization for the investigation, the research team first solicited advice from a project officer who worked in the PGS National Union. The project official, as the main force, has promoted PGS nationwide since 2019, has rich experience in PGS mobilization and organization, and has a good understanding of the development of the PGS national organizations. He recommended 12 organizations with more frequent PGS certification activities in the Northeast (Jilin), North China (Beijing and Hebei), East China (Shanghai, Anhui, and Jiangsu), Central China (Henan), South China (Guangdong), Southwest (Sichuan and Chongqing), and Northwest (Shaanxi and Gansu). Second, the research team evaluated the 12 organizations through their past long-term participatory research experience. Finally, we communicated with the heads of 12 PGS organizations to understand the basic situation of consumers’ participation in investigation and certification, which confirmed the scientific and feasibility of the research on the 12 organizations.

The survey questionnaires were designed by using an online survey. We entrusted each organization’s director with sending PGS questionnaires to participants through WeChat. Data were collected from October 2021 to December 2021. As a result, 238 complete responses were gathered and included in the final data set.

3.3. Data Analysis Method

This paper employed a structural equation model (SEM) and Amos 24.0.0 software. When compared with traditional statistical analysis, SEM is a multivariable statistical technology that solves the problem of multicollinearity. It has become the main analysis method in many study disciplines, because it can distinguish the indirect effects of variables in the process of path analysis, present the complex relationship visually and clearly in the model, and provide more rigorous and accurate information [61].

4. Results

4.1. Descriptive Statistic

Females and males represented 51.3 and 48.7% of the sample size, respectively, in this study. The respondents in the middle group of 26–55 years old accounted for 76.9%. Moreover, 72.3% of the respondents were at a high education level (college). The respondents mainly had more than three family members, 35.3% have three family members, 28.1% have four family members, and 26.1% have five family members. Additionally, 26.9% of respondents’ families earned between CNY 5000 and CNY 10,000, half of the respondents’ families earned more than CNY 10,000.

Consumers participating in the inspection and certification of ecological agricultural products pay close attention to food safety and 94.5% of respondents know organic food and green food. Only 10.9% of respondents trust agricultural products labeled “organic” and “green”, while 50.4% of respondents’ attitudes were not clear. The reason why respondents do not trust comes from the following aspects. First, one of the reasons is negative news from the certification market. Second, the main reason is the lack of key information about the producer, origin, and production process of food. In addition, some consumers worry about the interest collusion between producers and testing institutions under the third-party certification system. Consumers’ interests in safe and reliable products are believed be infringed in this profit-seeking access and regulatory system.

For the experience of consumer participation, 47.1% of respondents had a year or less than a year of participation, 52.9% had more than 3 years, and nearly half of respondents often participated in PGS certification more than four times, and some consumers participated only one to three times. The original intention of consumers to participate in certification activities follows four aspects. First, 65.1% of respondents love ecological agriculture, recognize the problems brought by unsustainable food systems, and are interested in seeking social innovative solutions to solve problems (such as PGS). Second, 62.2% of respondents obtained enough product information. Third, 53.8% of respondents are satisfied with PGS for establishing a friendship with producers. Fourth, 45% of respondents believe that PGS can bring them a safe and reliable channel to buy ecological agricultural products. Moreover, we also investigated consumers’ satisfaction with PGS organizations using a seven-point Likert scale. Consumers scored each item on a scale of one (representing strongly disagree) to seven (representing strongly agree), and the average score of each item is near six. It was said that consumers are satisfied with PGS organizations. They believe that the organizations cover the participation of multiple subjects, have strong cohesion and consistent values, and have good friendships between members.

4.2. The Measurement Model

4.2.1. Confirmatory Factor Analysis

The data used in SEM are derived from the investigation, so it is very important to test whether these data have reliability and validity. Therefore, we firstly performed the measurement model before the structural model. Confirmatory factor analysis (CFA) is an important tool that can judge whether a latent variable is truly and accurately explained by its measurement items. Specifically, we did single-factor CFA for each latent variable (consumer participation, information quality, social identity, and consumer trust) in Amos 24.0.0, and removed the non-significant items (IQ3, SI1, SI3, C CT2, CAT1, CAT4, and CAT5). Moreover, a measurement model of consumer participation with item parceling to simplify the model as shown in Figure 2. As shown in Table 3, the values of standardized factor loadings were 0.7–0.95 (>0.7), and the values of squared multiple correlations (SMC), representing item reliability, were 0.4–0.9 (>0.36). The results of the overall fit are displayed in

Table 2. The overall fit of the confirmatory factor analysis (CFA) models of latent variables.

Indicators	Ideal Standards	Model
		Consumer Participation CFA Model	Information Quality CFA Model	Social Identity CFA Model	Consumers Trust CFA Model
χ2	The smaller the better	——	5.810	4.298	16.061
Df	——		2	2	9
χ2/df	1–3		2.905	2.149	1.785
GFI	>0.80 Acceptable >0.90 Good fitting		0.987	0.991	0.977
AGFI	>0.80 Acceptable >0.90 Good fitting		0.937	0.957	0.947
IFI	>0.90		0.996	0.998	0.995
TLI	>0.90		0.989	0.993	0.992
CFI	>0.90		0.996	0.998	0.995
RMSEA	<0.08		0.090	0.070	0.058

. The χ² and Df values of CFA indicate that the models were consistent with their covariance data. Other overall fit indices such as RMSEA (<0.08), GFI (>0.9), AGFI (>0.9), and CFI (>0.9) met the cutoff criteria. Therefore, the measurement models of latent variables were found to be statistically acceptable.

4.2.2. Reliability and Validity Analysis

We then needed to examine the reliability and validity of the measurement model. We adopted composite reliability (CR, >0.7) to confirm the reliability of the measurement, reflecting the good internal consistency between variables [62]. We used the convergent and discriminant validity to confirm the construct validity of the measurement. The convergent validity shows the explanatory ability of potential variables to measure variables on average. It is judged by the value of the average variance extracted (AVE). The AVE measures the amount of variance captured by the construct through its items, relative to the amount of variance due to the measurement error. The discriminant validity is to ensure variables within different constructs are unrelated. A guideline for evaluating discriminant validity is to check that the square root of AVE should be larger than the correlations between the construct and any other factors in the model [62].

As results are shown in

Table 3. The measurement model.

Constructs	Indicators	Parametric Significance Estimates				Factor Loading	Items Reliability	CR	AVE
		Unstd.	S.E.	t-Value	p	Std.	SMC
Consumer participation	IS	1.000				0.886	0.785	0.945	0.851
	RB	0.991	0.044	22.438	***	0.933	0.870
	PI	1.033	0.045	23.052	***	0.947	0.897
Information quality	IQ1	1.000				0.921	0.848	0.953	0.835
	IQ2	1.052	0.038	27.586	***	0.956	0.914
	IQ4	1.068	0.050	21.219	***	0.874	0.764
	IQ5	0.931	0.040	23.504	***	0.903	0.815
Social identity	SI2	1.000				0.732	0.536	0.938	0.792
	SI4	1.212	0.080	15.233	***	0.954	0.910
	SI5	1.256	0.084	14.877	***	0.937	0.878
	SI6	1.160	0.079	14.667	***	0.919	0.845
Consumer trust	CCT1	1.000				0.894	0.799	0.951	0.766
	CCT3	1.061	0.046	23.285	***	0.928	0.861
	CCT4	1.041	0.042	24.621	***	0.948	0.899
	CCT5	1.007	0.044	22.941	***	0.921	0.848
	CAT2	0.814	0.045	18.224	***	0.834	0.696
	CAT3	0.919	0.069	13.371	***	0.703	0.494
Ideal value						>0.7	>0.36	>0.7	>0.5

Note: *** stands for p < 0.001.

, all CR values exceed the criteria value of 0.7, ranging from 0.93 to 0.95, indicating a high degree of internal consistency. Likewise, all AVE values for all constructs were more than 0.5, confirming that all constructs are established with satisfactory convergent validity. As shown in

Table 4. Square root of AVE.

Construct	CP	CT	IQ	SI
CP	0.922
CT	0.807	0.875
IQ	0.886	0.858	0.914
SI	0.819	0.876	0.871	0.890

, the values of the square root of AVE for the constructs in the diagonal (0.922, 0.875, 0.914, 0.890) were all greater than the correlations among the constructs that are below the diagonal. However, one index did not meet the evaluation criteria, but the tiny difference can be ignored. Hence, the results passed the discriminant validity test, and questionnaire items that measured different constructs should not be correlated.

4.3. The Structural Model

Structural models were used to verify the path of consumer participation on consumer trust, and to assess the degree of impact. As shown in Figure 3, a two-factor mediation analysis model of “the trust mechanism of consumer participation in ecological agricultural products’ certification” was constructed in Amos 24.0.0 software. IS, RB, and PI are three observation indicators of consumer participation, IQ1, IQ2, IQ4, and IQ5 are four observation indicators of information quality, SI2, SI4, SI5, and SI6 are four observation indicators of social identity, and CCT1, CCT3, CCT4, CCT5, CAT2, and CAT3 are six observation indicators of social identity. The circles, boxes, and ellipses in the figure represent the measurement error, observation variable, and potential variable respectively. The measurement error is named e1 to e20 in turn.

4.3.1. Assessment of Model Fit

The overall model fitting was not good, and χ²/df, RMSEA, and AGFI did not pass. By judging the variable normal distribution, we found |Skew| < 2 and |Kurtosis| < 7 indicating the data fit with univariate normal distribution [63]; CR value was 81 (>5), which showed that the data did not conform to multivariate normal distribution. It caused the expansion of the Chi-square value and the poor model fitting [64]. Then, this paper corrected the model by the Bollen-Stine Bootstrap method. Specifically, we used another Chi-square value estimated by the Bollen-Stine p-value to re-estimate model fitting [65]. The revised model fitting indicators are shown in

Table 5. The structural model fitting results.

		Initial Model		The Model Was Corrected by BS
Indicator	Ideal Standard	Estimated Value	Result	Estimated Value	Result
χ2	Smaller is better	383.454	——	221.605	Significantly reduce
D f		114	——	114	——
χ2/df	1< χ2/df < 3	3.364	Not pass	1.944	Pass
RMSEA	<0.08	0.100	Not pass	0.063	Pass
GFI	>0.9 Good fitting >0.8 Acceptable	0.841	Acceptable	0.959	Good fitting
AGFI	>0.9 Good fitting >0.8 Acceptable	0.787	Not pass	0.937	Good fitting
TLI(NNFI)	>0.9	0.939	Pass	0.975	Pass
CFI	>0.9	0.949	Pass	0.979	Pass
IFI	>0.9	0.949	Pass	0.980	Pass

, referring to Zhang Weihao [61]. χ²/df was less than 2; GFI, AGFI, NNFI, IFI, and CFI were higher than 0.9; RMSEA was less than 0.08. It can be seen that all fitting indices were in line with the standard, and the model fitting was satisfactory.

4.3.2. Hypotheses Testing

Statistically, the results of unstandardized estimates were used to test the research hypotheses. Generally speaking, if the C.R. value above the recommended threshold (1.96), * p < 0.05, the parameter reaches 0.05 significance level; with C.R. values above 3.29, *** p < 0.001, the parameter reaches a significant level of 0.001 [66]. Meanwhile, the results of unstandardized estimates reflect the importance of variables. We operated the model in Amos 24.0.0 software with the data collected by our study, and the standardized path coefficient diagram among four of the proposed latent research variables as output, as shown in Figure 4. Then we sorted out the results of standardized estimates and unstandardized estimates as shown in

Table 6. Results of path coefficient estimation for the structural model.

Hypotheses	Path	Standardized Path Coefficients	Unstandardized Path Coefficients	S.E.	C.R.	p	Result of Hypothesis Test
Hypothesis 1	CT <- - - CP	0.063	0.063	0.104	0.611	0.541	Reject
Hypothesis 2	IQ <- - - CP	0.907	0.936	0.049	19.163	***	Accept
Hypothesis 3	SI <- - - CP	0.849	0.74	0.06	12.368	***	Accept
Hypothesis 4	CT <- - - IQ	0.36	0.351	0.098	3.574	***	Accept
Hypothesis 5	CT <- - - SI	0.528	0.61	0.093	6.577	***	Accept

Note: *** stands for p < 0.001.

.

For Hypothesis 1, the results indicated that the path of consumer participation to consumer trust (C.R. = 0.611 < 1.96, p = 0.541, γ11 = 0.063 (γ = path coefficient between an exogenous variable and an endogenous variable)) was not statistically significant. Thus, Hypothesis 1 was rejected. With regard to Hypotheses 2 and 3, the results of the structural model showed that the positive and direct effects of consumer participation on information quality (C.R. = 19.163 > 1.96, γ21 = 0.907) and social identity (C.R. = 12.368 > 1.96, γ31 = 0.849) were statistically significant. Thus, Hypotheses 2 and 3 were accepted. Regarding Hypotheses 4 and 5, the positive and direct effect of information quality on consumer trust (C.R. = 3.574 > 1.96, β21 = 0.36 (β = path coefficient between endogenous variables)) and the positive and direct effect of social identity on consumer trust (C.R. = 6.577 > 1.96, β31 = 0.528) were both statistically significant. Thus, Hypotheses 4 and 5 were accepted. Concerning Hypotheses 6 and 7, a bootstrap estimate approach was performed to test the mediating effects of work information quality and social identity.

4.3.3. Mediating Effects Test

For tests of mediation effects, the causal method was first proposed by Baron and Kenny [67]. Subsequently, Sobel proposed the coefficient product method [68]. However, both methods have obvious limitations, the former failing to test the significance of a multiplied by b, and the latter failing to address the effect of the non-normal distribution of a multiplied by b on significance results [69,70]. According to MacKinnon, the confidence interval method (Bootstrap method) is the most powerful test method for specific indirect effects [71]. This paper tested the mediating effects through the BC bootstrapping procedure with 1000 bootstrap samples. We set the confidence level to 95%. If Z values are greater than 1.96 and confidence intervals of the paths do not contain zero, it is said that the mediating effects are significant. The results of the bootstrap estimates are presented in

Table 7. Bootstrap estimates of the mediating effects of variables.

Path	Point Estimation	Product of Coefficients		Bootstrapping
				Bias-Corrected 95%CI		Percentile 95%CI
		SE	Z	Lower	Upper	Lower	Upper
Consumer participation ↓ Information Quality social identity ↓ Consumers trust	Total Effect
	0.844	0.084	10.048	0.672	0.997	0.672	1.000
	Indirect Effect
	0.780	0.162	4.815	0.571	1.259	0.542	1.171
	Direct Effect
	0.063	0.147	0.429	−0.240	0.352	−0.267	0.348

Note: CI stands for confidence interval; 1000 bootstrap samples.

and

Table 8. Specific analysis of the mediating effects.

SIE	Point Estimation	Product of Coefficients		Bias-Corrected		Percentile		Result of Hypothesis Test
		SE	Z	Lower	Upper	Lower	Upper
CP → IQ → CT	0.329	0.162	2.031	0.061	0.673	0.062	0.674	Accept H6
CP → SI → CT	0.452	0.117	3.863	0.281	0.755	0.255	0.711	Accept H7
IE difference	0.123	0.231	0.532	−0.194	0.529	−0.222	0.481	——

Note: 1000 bootstrap samples.

.

As shown in Table 7, Z values for both total and indirect effects were greater than 1.96, and the confidence intervals for the Bias-Corrected and Percentile tests in the Bootstrap method did not contain zero; thus, the total indirect effect was significant in this model. It indicated that the intermediary effects of information quality and social identity were both significant. The Z value of the direct effect was 0.429, less than 1.96, and the confidence intervals of Bias-Corrected and Percentile were (−0.240, 0.352) and (−0.267, 0.348), respectively, so that the direct effect does not exist. The direct influence of consumer participation on trust was only 6%, and about 94% of the impact was transmitted by two mediating variables. It meant the information quality together with social identity played a complete mediating role between consumer participation and consumer trust. Therefore, this model was a two-factor complete mediation model.

Then, we identified two specific mediating effects. As shown in Table 8, the results demonstrated the statistically significant mediating effects of information quality in the relationship between consumer participation and consumer trust of ecological agricultural products (Z = 2.031 > 1.96, Bias-Corrected CI (0.115, 0.278), Percentile CI (0.062, 0.674)), as well as the statistically significant mediating effects of social identity in the relationship between consumer participation and consumer trust of ecological agricultural products (Z = 3.963 > 1.96, Bias-Corrected CI (0.281, 0.755), Percentile CI (0.255, 0.711)). This assumed H6 and H7 were correct. The intermediary effect value of information quality was 0.329, and it accounted for about 42% of the total mediating effects (0.329/0.780). The intermediary effect value of social identity was 0.452, and it accounted for about 58% of the total mediating effects (0.452/0.780). The difference between the two mediating effects was not significant, indicating that both information quality and social identity were equally important.

5. Discussion

This paper studied the factors affecting consumer trust in ecological agricultural products and the relationship between these factors under the participatory guarantee system. For this purpose, a two-factor mediation analysis model was constructed. We took consumers who have participated in the quality and safety inspection of agricultural products as respondents for the research. It empirically examined consumer participation during the process of product certification, which will affect consumer trust through two mediation variables: information quality and social identity. The results of this paper showed that the model matched research data very well, and fitting indices all reached the ideal state. Apart from the fact that customer participation has no direct impact on trust, the rest of the research hypotheses have been supported. Consumer participation influences consumer trust through information quality and social identity significantly and indirectly. Therefore, the following results can be discussed:

(1) The direct impact of consumer participation on trust is not significant, assuming H1 is incorrect. It shows that consumer participation in certification has little impact on trust directly. Consumers need clear clues to build trust, such as obtaining product information and strengthening their sense of identity. Although existing studies have confirmed that consumer participation has a direct positive impact on trust, the effect is very small in the PGS of this study. Consumer personal qualities may have an impact on the rationale. Consumers in the process of certification must have a high level of cognitive ability and professionalism in ecological agriculture. If consumers participate in the certification without having any knowledge about ecological agriculture, they may misunderstand farmers or not trust products of safety. For example, some consumers think ecological agriculture cannot apply any additions, however, some biological preparations are permitted in moderate amounts.

(2) Consumer participation in certification indirectly affects consumer trust through two mediator variables: information quality and social identity. The findings reveal that the direct effect of consumer participation on trust is not significant, but there is a total effect and a total indirect effect, those being information quality and social identity, which provided 42% and 58% of the mediation effects, respectively, in the two-factor complete mediation model developed in this work. Therefore, this study shows that in the path of consumer participation in ecological product certification on consumer trust, obtaining high-quality information and realizing social identity are two more important processes of the intermediary.

On the one hand, information quality plays a complete intermediary role between consumer participation and trust. This conclusion confirms that information led by consumers can better promote the establishment of trust [1,44]. However, information led by consumers comes from the experience of themselves and the recommendation of relatives and friends, mainly mentioned in previous studies. It focuses on consumers’ intuitive feelings about products in the sale of ecological agricultural products. Consumers master specific information about product images, such as appearance, taste, and service. Therefore, the establishment of consumer trust is based on the improvement of consumers’ perceived value [44]. This paper studied consumer participation in the certification of ecological agricultural products, so that consumers could obtain more abundant and reliable information. When consumers participate in PGS, they have the ability to see producers and other PGS stakeholders face to face in addition to seeing the product image. Consumers can get familiar with the production process, listen to the diagnosis of technicians and experts, and get the real sampling results. This information can help consumers objectively evaluate whether the producer of the ecological agricultural product has reliable ecological agricultural technology and whether they are willing to pay for it, resulting in a significant increase in consumers’ perceived value and the establishment of a stable trust relationship.

On the other hand, social identity plays a complete intermediary role between consumer participation and trust. Based on social identity theory, the researchers showed that, in a specific situation, individuals believe that they are similar to members of the group and can perceive the feelings and values brought by the group, and then have a sense of belonging to the group [38]. The stronger the sense of belonging, the more inclined individuals are to accept or use the services or products provided by the group [42,43]. According to our findings, PGS participants have a high level of education, are mostly young adults, and have family members. They support the development of ecological agriculture to improve environmental pollution and food insecurity. However, they do not trust the “green” and “organic” products sold on the market and hope to find real ecological products with childhood flavor. For the individual characteristics of consumers, it can be seen that consumers and PGS have the same goals. Consumers feel the efforts made by these ecological farmers for the ecological environment and consumer health in the loose, open, and integrated environment of PGS, and are willing to pay a fair price. As a consumer, they internalize this specific social identity and, finally realize the impact of social identity on trust. Therefore, consumer participation in the product quality certification process and results over time through interactive communication, thereby increasing trust.

(3) Consumer participation positively affects information quality (standardized path coefficient 0.907) and social identity (standardized path coefficient 0.849), indicating that consumer participation in certification has a great impact on information quality and social identity. This is consistent with the conclusions of previous studies [20,24,29]. On the one hand, it can deal with the problems of asymmetrical information regarding the ecological agricultural products market, with consumers participating personally. The modern agricultural supply chain is complicated, with numerous participants. In many cases, consumers passively obtain limited product information about origin and identification from various channels in the market. Consumers who directly participate in certification can gain accurate information. On the other hand, consumer participation in PGS not only acquires test reports that meet strict third-party certification standards, but also realizes the value of co-creation with consumer participation. Based on communication in the certification process, consumers gradually agree with the value of ecological agriculture. Although ecological agricultural products are the necessities of life, they also symbolize human health, sustainable agriculture, a sustainable lifestyle, and a win–win connection between producers and consumers.

(4) The information quality and social identity positively influence consumer trust, and social identity has a more important impact on trust under PGS. From the perspective of information asymmetry, scholars put information into the theoretical framework of consumer trust mechanisms in food safety, demonstrating that information disclosure and intervention have become the primary ways to solve the problem of mistrust [44,51,72]. This viewpoint has been confirmed in this paper’s discussion of PGS. During consumer participation, the richness, accuracy, authority, and readability of information can be gathered to ease the ambiguity of consumers’ perception of products to some level and then enhance consumer trust.

The results of this study also showed that, in addition to information quality, the impact of social identity on trust cannot be ignored. Although there are few studies on the influence of identity factors on consumer trust in ecological agricultural products, through combing the theory of social identity, this paper found that identity and consumer have a close relationship, which provides a theoretical basis for the research on the influence of identity factors on consumer trust. People tend to build trust relationships with people who are similar to themselves, and identity is an important step in building trust [37]. This paper confirms this view. Communication between PGS members when participating in certification can constantly improve consumers’ sense of identity to the certification results, and then enhance trust. When consumers and other PGS members have common ideas, consumers tend to trust that ecological products certified by PGS are reliable, and the higher the consistency of ideas is, the greater the consumer tendency to trust.

The results show that the standardized path coefficient of SI to CT (0.528) is slightly higher than the standardized path coefficient of IQ to CT (0.36). It indicates that consumer trust depends more on social identity under PGS. Now, consumers distrust certified products for the reason that consumers see producers and testing institutions blindly pursuing interests through some events that promote bad food safety practices. In this case, the effect was limited, in which trust was maintained only by information disclosure. The information provided by profit-seekers is likely to be false. Consumers are more willing to believe in people who are consistent with their values, trust producers to maintain original intention, unswervingly develop ecological agriculture, stabilize the mutual assistance relationship between consumers, and ensure dietary health of consumers under this circumstance.

6. Conclusions, Implications, and Limitations

6.1. Contributions and Implications

This study answers the question about whether consumer participation can affect consumer trust in ecological agricultural products under PGS, and the action path. The results of this study show that consumer participation affects trust through information quality and social identity. There are two main contributions to this study. On the one hand, this paper reveals the mechanism of consumer trust under PGS from the perspective of consumer participation. It enriches the related theories of PGS. On the other hand, the theoretical model of this paper analyzes the impact of consumer participation factors and identity factors on consumers’ trust in ecological agricultural products. It expands the research framework of consumer trust in ecological agricultural products. The following two implications can be proposed for the practice of PGS and the development of TPC based on this paper.

(1) First, we discuss how to win the trust and continuous attention of consumers to promote the steady development of PGS according to the current situation of PGS in China. Nowadays, PGS organizers in China enable consumers to participate in certification activities by mobilizing various resources. However, the results of this paper show that the direct impact of consumer participation on trust is small, and that information and identity play important intermediary roles in promoting consumer trust. Therefore, PGS organizers should not only invite consumers to participate in certification activities but should also pay close attention to the personal situation of consumers, such as whether consumers themselves have obtained a lot of valuable information and whether consumers agree with this kind of quality assurance system, to establish a solid consumer trust relationship.

To this end, there are three suggestions. Firstly, PGS organizers should do a good job in the detailed design of certification activity and increase the inspection items. For example, organizing team trains for inspection, having discussion meetings during the inspection, and holding summary meetings after inspection. For the above measures, this can create an atmosphere of full interaction of important information, mutual understanding, and mutual trust for involved consumers, rather than only bringing them into the process symbolically, and not caring about their actual participation. Second, relying on modern information technology, specifically, PGS organizations can realize the application of blockchain to PGS, which can store various information of certificate process and results on the Internet. At present, PGS organizations in China are working together to build this data monitoring system, which provides a convenient channel for consumers to obtain information. Consumers can not only participate in the field investigation but also participate online to monitor the production of producers. Third, organizing consumer activities regularly is important. PGS is in the initial stage of development in China. It is necessary to invite more consumers to participate in co-construction and sharing, and enhance the cohesion of the PGS team in a relaxed and pleasant atmosphere.

(2) Second, this paper provides a reference to the third-party certification systems for enhancing consumer trust. In the context of the continuous expansion of the sales market of ecological agricultural products, trust is an important factor affecting consumers’ purchase behavior. In some cases, the products with nationally authorized logos have low consumer trust, such as those labeled “organic food” and “green food”. Therefore, this will restrict the development of the sales market of ecological agricultural products. The reason for this may be the lack of checks and balances from consumers in third-party certification system. The findings confirm that the private certification of consumer participation such as PGS can enhance consumer trust, such that consumer autonomy in food safety regulation is improved. In China, a variety of informal certification groups, such as farmer markets, consumer (life) cooperatives, ecological agriculture cooperatives, and so on, have actively emerged in the past decade. These organizations follow the concept of PGS and invite consumers to inspect products, so that they have maintained a loyal consumer base. Therefore, the current product certification system needs to learn from PGS and consider the interest of consumers and improve the discourse right of consumers, to build a food safety supervision system that is self-recognized by consumers and widely trusted by the public.

6.2. Limitations and Future Research

The following several limitations should be considered when conducting future research. Firstly, future research needs to enlarge the sample size. The survey in this study is not enough. This is because of the low degree of Chinese consumers’ participation in the early development of PGS in China. However, the concept and practices of PGS have been spread in China in recent years. Consumers, scholars, officials, and other stakeholders are paying more attention to the development of PGS than before. That is to say, more consumers will participate. The samples of future research must be enlarged.

Secondly, moderator variables (i.e., organizational environment, and organizational characteristics) can be put into the structural model of consumer trust created in this paper in future research. Consumers participating in different organizational environments will have distinct behaviors and make different decisions. It can also conduct group analysis by classifying consumers according to different characteristics of PGS organizations. Through comparison between groups, it can be analyzed which type of PGS organizations tend to win consumer trust.

Last, future research can explore the impact of three behaviors (information sharing, responsible behavior, and personal interaction) of consumer participation on trust respectively. It can indicate which kind of consumer participation behavior is more important to trust, and then it can provide PGS organizers with more specific measures to promote consumer trust.