Attitude is the likelihood that an individual will respond positively or negatively to behavior (Ajzen, 2011), and consumers’ attitudes are considered to be a key element in effectively predicting their behavioral intentions toward organic food (Sadiq et al., 2021). Drawing on Lee and Yun (2015), this paper classifies attitudes into two dimensions based on the function of use and purchase motivation: utilitarian attitudes and hedonistic attitudes. Utilitarian attitudes relate to the functional value of organic food, reflecting the consumer assessment of the benefits when purchasing organic food (Ditlevsen et al., 2019), while hedonistic attitudes relate to emotional satisfaction or sensory experience, reflecting the perceived value derived from the multisensory and emotional aspects of purchasing organic food, such as the taste, freshness, or pleasure derived from doing something good for one’s health and the environment (Wang et al., 2019).

Expectancy-value-attitude theory suggests that an individual’s attitude toward an object is determined by the subjective probability that the object has a particular attribute (Ajzen and Fishbein, 2000). Specifically, an individual will form an overall attitude toward an object by assessing the attributes associated with it and the extent to which its needs are met. Organic food is a credence product that is closely related to consumers’ health and safety and for which product information is difficult to obtain, and there is a certain degree of premium (Yu et al., 2022), so consumers tend to rely on credence attributes for psychological judgment (Janssen and Hamm, 2012; Massey et al., 2018). In organic food consumption, consumers usually pay attention to the food safety of organic food (Vapa-Tankosić et al., 2020; Joya et al., 2021). They believe that organic food is produced without pesticides, chemical fertilizers, artificial additives, and other harmful substances and is safer than ordinary food (Tan et al., 2021; Lang and Rodrigues, 2022), which makes them have a positive attitude toward organic food (Çabuk et al., 2014; Pham et al., 2019). Consumers make positive psychological judgments about food safety that meet their own needs. Such judgments are mainly motivated by the functional and emotional affirmation of organic food, which facilitates the formation of utilitarian attitudes and hedonistic attitudes among consumers. Therefore, the following hypotheses were proposed:

Eco-friendliness means that organic food follows the principles of sustainable development in the production process, using specific techniques to avoid genetic engineering and not causing harm to the environment and animals (Honkanen et al., 2006). Information about the sustainability benefits and environmental value of eco-friendliness can influence consumer attitudes toward organic food (Puteri et al., 2022). Consumers will psychologically discern whether the eco-friendliness of organic food meets their needs and further consider whether the pro-environmental characteristics of organic food enhances their personal emotional experience (Gkargkavouzi et al., 2019, Canio et al., 2021). The production process of organic food is concerned with animal welfare and production equity, which promotes ecological balance and makes consumers perceive that buying organic food not only satisfies their own needs but also benefits the ecological environment, which leads to more positive utilitarian attitudes and hedonistic attitudes. Therefore, the following hypotheses were proposed:

The WTPP is the willingness of consumers to pay a higher price for organic food compared to regular food (Konuk, 2018). Consumers’ positive attitudes toward organic food can lead to a greater WTPP (Nassivera et al., 2017). Consumers with a high WTPP are not paying a premium for the organic food itself but for the health and environmental benefits they feel when consuming organic food (Bryła, 2016; Molinillo et al., 2020). This perceived stimulus induces changes in consumers’ emotions and attitudes, which, in turn, affects their WTPP.

According to the theory of planned behavior, attitude is an important antecedent variable of willingness (Yazar and Burucuoğlu, 2019). Attitude reflects not only the individual’s emotional evaluation of a specific thing but also the individual’s behavioral intention. The more positive the individual’s attitude toward a particular thing, the more positive the individual’s intention to develop that behavior (Clark et al., 2019; Salvatore et al., 2021). A positive attitude is an important prerequisite for organic food consumption (Boobalan et al., 2021; Santos et al., 2021). Compared to consumers with a negative attitude, those with a positive utilitarian attitudes and hedonistic attitudes toward organic food believe that organic consumption is beneficial to health and the environment, and this belief not only results in psychological satisfaction but also stimulates their WTPP. Therefore, the following hypotheses were proposed:

S-O-R theory holds that an individual’s internal psychological state plays a mediating role between external stimuli and behavioral responses; that is, external characteristics stimulate consumers’ behavioral responses by influencing their mental states (Su and Swanson, 2017). Consumers also go through a series of psychological reactions before forming a WTPP for organic food. Specifically, consumers will carefully evaluate the credence attributes of organic food to satisfy their utilitarian and hedonistic goals when purchasing organic food (Sultan et al., 2021). With the external stimulus of the credence attributes of organic food, consumers will develop different attitudes (Brach et al., 2018). On the one hand, the functional value that consumers perceive from the credence attributes of organic food contributes to the development of utilitarian attitudes (Fuljahn and Moosmayer, 2011). On the other hand, consumers may develop hedonistic attitudes because they feel good about doing something that is good for their health and the environment (Maehle et al., 2015). In this case, credence attributes provide consumers a perceived stimulus that influences their attitudes toward organic food and further affects their WTPP. Therefore, the following hypotheses were proposed:

Uncertainty can be described as a state of incomplete information (Vieira, 2008). When consumers lack reliable information resources and professional knowledge to distinguish organic food, they will doubt the true properties of organic food and generate uncertainty (Yenipazarli, 2015). The current state of information credibility and standards of the organic food market is in a state of confusion, with a variety of certification systems and labels making it difficult for consumers to identify organic food. These potential information asymmetries and ambiguities can increase consumers’ uncertainty (Nuttavuthisit and Thøgersen, 2017) and prevent them from accurately assessing the functional and emotional value of organic food. In this case, uncertainty will reduce consumers’ WTPP even if they have a positive attitude toward organic food (Thøgersen, 2016). Conversely, if the information about the production and control processes of organic food is sufficient, credible, and easily accessible, consumers’ uncertainty about organic food will be reduced, which in turn will lead to a higher WTPP. Therefore, the following hypotheses were proposed:

Based on the above hypotheses, the theoretical model of this paper is shown in Figure 1.

We used the Questionnaire Star platform to design questionnaires. The first part of the questionnaire set up an introductory paragraph introducing organic food and the organic certification label so that participants could fully understand organic food. Two screening questions were also included in the design of the first section. The first was, “Would you like to take part in this survey?” The second was, “Have you ever purchased organic food?” This was done to verify that the participants agreed to participate in the survey and had purchased organic food in order to improve the accuracy of the questionnaire. The second part required participants to answer questions on a 5-point Likert scale regarding credence attributes, attitudes, uncertainty, and willingness to pay a premium, and each page was set to be nonreturnable to prevent participants from repeatedly checking or changing their answers which would invalidate the results of the study. In the third section, participants were asked to fill in some personal information.

Prior to the formal survey, an initial online questionnaire was sent to 10 PhDs in the field of marketing at a large university in Northeast China to ensure that the questions were designed to be easy to understand. Based on their suggestions, unclearly phrased questions were revised and repetitive questions were removed to create the final questionnaire. The formal survey was launched in mid-April. We joined several supermarket discount chat groups and community buying chat groups on the WeChat platform and asked participants to share the online survey among their network of acquaintances using a “snowball” approach to spread the questionnaire online. Given the diversity of organic food buyers, there was no geographical restriction on the distribution of the questionnaires, and a total of 1,000 questionnaires were collected. According to Huang et al. (2015), the presence of insufficient effort responding (IER) in survey data should be assessed prior to hypothesis testing. Therefore, we used response times (Huang et al., 2015; Cheung et al., 2017) to detect and screen out inattentive/careless responding to ensure the reliability of the questionnaire. We tested response times and derived a standard time based on the number of questions, and then excluded questionnaires that took less than the standard time to complete, as well as those with missing data and contradictory questions. A total of 647 valid questionnaires were finally collected.

The constructs in this study were measured by existing scales adopted from previous studies. Food safety was measured mainly from Steptoe et al. (1995), which contains three measurement items; Eco-friendliness was measured from the Lindeman and Väänänen (2000) four-item scale; Willingness to pay a premium was measured by three question items from Zhang et al. (2020), and the measurement of uncertainty was borrowed from Shiu et al. (2011). These four constructs were measured using a 5-point Likert scale, which classified subjects’ agreement with the question items on seven levels from 1 to 5 (strongly disagree to strongly agree). The utilitarian attitudes and hedonistic attitudes were measured using a scale developed by Voss et al. (2003), each containing four measurement items. The 5-point semantic differential scale was used to classify the subjects’ understanding of the question items on a 5-point scale from 1 to 5. To improve the validity of the questionnaire, opinions were sought from experts in the food and marketing fields, and appropriate modifications were made in the context of the study to ensure that subjects could better understand each question item. The specific question items are shown in Table 1.

The descriptive statistical characteristics of the sample are specifically shown in Table 2. The results show that there are more women than men among the respondents, accounting for 62.3%, which reflects that the purchasers of organic food in households are mostly women. The age of the respondents is mainly distributed between 20 and 45 years old, and the consumers who buy organic food are mainly young and middle aged. An education level of college, bachelor’s degree, and above accounted for 80.4%, and the monthly income was mainly in the 3,000–6,000 RMB range, accounting for 39.6%, indicating that the average education level and income level of the respondents are relatively high. Overall, the sample has a good representation.

Cronbach’s α value was used to test the reliability. As seen in Table 3, the Cronbach’s α values of each construct were above 0.7 (Fornell and Larcker, 1981), and the reliability of the scale was good. Confirmatory factor analysis (CFA) was used to evaluate the validity of the scale, and the results showed that X²/df = 3.309, less than 5; RMSEA = 0.056, less than 0.08; GFI = 0.928, AGFI = 0.90, CFI = 0.964, NFI = 0.948, IFI = 0.964, all above the benchmark of 0.9, which indicated a good fit of the model and the data (Tabachnick et al., 2007). The standardized factor loadings (estimate) for each item were above the critical value of 0.5. The average variance extracted (AVE) for each construct was above the recommended value of 0.5, and the combined reliability (CR) was above the benchmark of 0.7, implying relatively good convergent validity (Schwab, 2006). Discriminant validity was tested by comparing the AVE with the correlation coefficients of the variables. The square roots of the AVEs on the diagonal in Table 4 were all greater than the correlation coefficients between them and the other constructs, indicating good discriminant validity between the variables (Schwab, 2006).

The hypothesis is tested by the SEM of maximum likelihood estimation. The model fit was X² = 418.478, df = 120, p < 0.01, X²/df = 3.487, GFI = 0.934, AGFI = 0.906, CFI = 0.966, NFI = 0.953, IFI = 0.966, RMSEA = 0.062. All fit indicators meet the requirements, indicating a good fit of the model (Browne and Cudeck, 1992). The results based on the structural equation model are shown in Table 5. Food safety had a significant positive influence on both utilitarian attitudes UA (β = 0.220, t = 2.528, p = 0.011) and hedonistic attitudes (β = 0.200, t = 2.674, p = 0.007). Thus, hypotheses H1a and H1b were verified. Eco-friendliness had a significant positive influence on both utilitarian attitudes (β = 0.170, t = 2.025, p = 0.043) and hedonistic attitudes (β = 0.220, t = 2.603, p = 0.009). Thus, hypotheses H2a and H2b were verified. Utilitarian attitudes had a significant positive influence on WTPP (β = 0.130, t = 3.471, p = 0.000). Hedonistic attitudes had a significant positive influence on WTPP (β = 0.490, t = 11.442, p = 0.000). Thus, hypotheses H3a and H3b were verified.

The mediating effects of utilitarian and hedonistic attitudes were tested using Model 4 (Model 4 is a simple mediation model) in the Process macro proposed by Hayes and Scharkow (2013). The Bootstrap test (a statistical method of multiple repeated sampling) was chosen and set with a number of repetitions of 5,000 and a 95% confidence interval. As shown in Table 6, the bootstrap 95% confidence intervals for both the indirect [0.071, 0.180] and direct [0.386, 0.548] effects of utilitarian attitudes in path 1 did not contain 0, which indicated that utilitarian attitudes partially mediate the relationship between food safety and WTPP. The bootstrap 95% confidence intervals for both the indirect [0.068, 0.169] and direct [0.405, 0.562] effects of utilitarian attitudes in path 3 did not contain 0, which indicated that utilitarian attitudes play a partially mediating role in the relationship between eco-friendliness and WTPP. Hypothesis H4a was verified. The bootstrap 95% confidence intervals for both the indirect [0.103, 0.218] and direct [0.353, 0.511] effects of hedonistic attitudes in path 2 did not contain 0, which indicated that hedonistic attitudes play a partially mediating role in the relationship between food safety and WTPP. In path 4, the bootstrap 95% confidence intervals for both the indirect [0.010, 0.207] and direct [0.372, 0.525] effects of hedonistic attitudes did not contain 0, which indicated that hedonistic attitudes play a partially mediating role in the relationship between eco-friendliness and WTPP. Thus, hypothesis H4b was verified.

By comparing the relative magnitude of the mediating effect, in the indirect relationship between food safety and WTPP, the proportion of the mediating effect of utilitarian attitudes is 0.156/0.588 = 26.53% and the proportion of the mediating effect of hedonistic attitudes is 0.121/0.588 = 20.58%, indicating that the mediating effect of utilitarian attitudes is greater than that of hedonistic attitudes. However, in the indirect relationship between eco-friendliness and WTPP, the proportion of the mediating effect of utilitarian attitudes is 0.114/0.598 = 19.06%, and the proportion of the mediating effect of hedonistic attitudes is 0.149/0.598 = 24.91%, indicating that the mediating effect of hedonistic attitudes is greater than that of utilitarian attitudes.

The moderated mediation analysis model (Process Model 14) proposed by Hayes (2018) was used to test for moderating effects and mediating effects with moderation. Food safety and eco-friendliness were placed into the model as independent variables, utilitarian and hedonistic attitudes as mediating variables, uncertainty as a moderating variable, and WTPP as a dependent variable. The results of the test for the moderating effect of uncertainty (Table 7) show that uncertainty plays a significant negative interaction effect on the relationship between utilitarian attitudes and WTPP (β = −0.185, p = 0.012; β = −0.186, p = 0.011). Thus, hypothesis H5a was verified. Uncertainty plays a significant positive interaction effect on the relationship between hedonistic attitudes and WTPP (β = 0.132, p = 0.040; β = 0.138, p = 0.030). Thus, hypothesis H5b was not verified.

The test for mediating effects with moderation is shown in Table 8. When the independent variable was food safety, the moderating effect of low-uncertainty consumers on the mediating effect of utilitarian attitudes was significant (indirect effect = 0.007, boot CI = [0.029, 0.140]); the moderating effect of high-uncertainty consumers on the mediating effect of utilitarian attitudes was not significant (indirect effect = −0.015, boot CI = [−0.080, 0.067]). The results indicate that there is a significant difference between the low-uncertainty and high-uncertainty groups in terms of whether they influence the WTPP through utilitarian attitudes. The mediating effect of utilitarian attitudes is moderated by uncertainty. In addition, according to the determination index, it can be further shown that uncertainty strengthens the mediating role of utilitarian attitudes in food safety and WTPP (index = −0.069, boot CI = [−0.140, −0.003]).

The high-uncertainty group (indirect effect = 0.174, boot CI = [0.106, 0.251]) moderated the mediating effect of hedonistic attitudes more than the low-uncertainty group (indirect effect = 0.100, boot CI = [0.042, 0.160]), and there was a significant difference in the indirect effect between the two groups, indicating that the mediating effect of hedonistic attitudes was moderated by uncertainty. In addition, uncertainty strengthened the mediating role of hedonistic attitudes between food safety and WTPP (index = 0.056, boot CI = [0.001, 0.121]). Similarly, when the independent variable is eco-friendliness, uncertainty strengthens the mediating role of utilitarian attitudes and hedonistic attitudes between eco-friendliness and WTPP.

The foremost theoretical contribution of this research is the extension of the S-O-R model in the context of organic food consumption. Although S-O-R models have been used for predicting purchase intentions of organic food (Kim et al., 2021; Sultan et al., 2021), the S-O-R framework in the organic food context is currently still underrepresented in the literature, especially in the Chinese context. Moreover, there is still a lack of studies to measure other causal and moderating factors in the framework (Lee and Yun, 2015), so a more robust exploration of the framework is necessary. This study is among the first to develop a model that explains how consumers’ willingness to pay a premium for organic food is driven and constrained by integrating food safety, eco-friendliness, utilitarian attitudes, hedonistic attitudes, and uncertainty. It extends the S-O-R theory and its application context. At the same time, the results of the study provide a new theoretical basis for companies to develop organic food marketing strategies.

Secondly, this study reveals the internal psychological mechanisms involved in the decision-making process of consumers in purchasing organic food. While a large number of studies have highlighted the role of attitudes in intention to purchase organic food and actual purchase behavior (Boobalan et al., 2021; Santos et al., 2021), they have mainly focused on unidimensional attitudes (Lee and Goudeau, 2014). This research investigated utilitarian and hedonic dimensions of attitudes toward purchasing organic foods. This conceptualization is a more integrative approach than the previous studies in which one-dimensional concept of attitudes is considered. Further, the findings suggest that hedonistic attitudes have a greater impact on consumers’ WTPP, implying that consumers may be more inclined to seek a superior emotional experience when purchasing organic food, which will provide new dimensions of exploration for future research in the area of consumer attitudes and behavior.

Thirdly, this study strengthens the explanatory power of the S-O-R model by adding uncertainty as a moderating variable to the model. Our study identifies the moderating role of uncertainty in the relationship between different dimensions of attitudes and WTPP, contributing to the explanation of the attitude-behavior gap in organic food research. Furthermore, our study found that uncertainty strengthens the mediating role of utilitarian and hedonistic attitudes between credence attributes and WTPP. This moderated mediating role could provide new and substantial insights into marketing theory and future research directions and could expand the boundaries of existing organic food research. Finally, the study spotlights the behavior of Chinese consumers in regard to organic food. Findings bridge the gap of deficient theoretical knowledge on the unique characteristics of Chinese consumers and their responses to organically grown food items.

Firstly, we suggest that manufacturers could consider carefully designing safety and environmental messages to showcase how organic food contributes to the health and well-being of consumers. Specifically, in terms of eco-friendliness, manufacturers may consider reporting the carbon footprint of their products and creating a consumer sustainability index (Nikolaou and Kazantzidis, 2016) to improve the perceived environmental value of a product among consumers. In terms of food safety, retailers could cite the latest and most authoritative scientific findings or endorsements by national agencies to support their claims on the safety of organic food, such as less additives and pesticide residues, more beneficial nutrients, etc. In addition, retailers could communicate the safety and environmental benefits of organic foods through social networks, mass media, and interactive screens in-store, and include integrated promotional mix elements in their communication campaign (Chekima et al., 2019).

Secondly, as established consumer attitudes are difficult to change (Aaker, 1996), we suggest that marketers could stimulate positive attitudes toward organic food by providing consumers with cues that organic food contains hedonic and utilitarian attributes that will satisfy their needs. Specifically, for utilitarian consumers, marketers may participate or sponsor research related to the organic food safety testing and reflect the findings in advertising, packaging, or leaflets, which will help stimulate utilitarian attitudes among consumers. For hedonistic consumers, marketers could focus on hedonic gratification derived from purchasing organic foods. For example, by emphasizing that an organic food purchase is a way to ultimately save animals and the environment, or by using experiential marketing methods, organizing food tastings, food festival competitions, etc. to stimulate hedonic feelings such as satisfaction. According to our findings, hedonistic attitudes have a stronger impact on consumers’ WTPP compared to utilitarian attitudes. Therefore, marketers could leverage the hedonic benefits of the consumption of organic food by engaging in emotional appeal advertising and highlight the hedonic attributes of their products through a multi-sensory marketing approach (attractive packaging and good taste, etc.), thus enhancing consumer pleasure and enjoyment.

Finally, we recommend that manufacturers may consider introducing the latest blockchain technology. Blockchain food traceability system has the characteristics of traceability and tamper-proof, which can ensure the authenticity of information (Wei et al., 2022), thus reducing the perceived information asymmetry between consumers and organic food suppliers. Manufacturers could also introduce QR codes with traceability information on product packaging. By scanning the QR code, consumers could have information on the entire process of organic products from production to distribution, which will help increase transparency and address consumers’ uncertainty. In addition, organic producers could show more production details to consumers through media and social platforms to reduce consumers’ uncertainty and increase their organic purchase intentions.