According to Behavior Reasoning Theory (BRT), the rationality of behavior or the reasons behind actions is a more direct and critical factor influencing decision-making than traditional psychological variables such as values and attitudes (Westaby, 2005). Most scholars have conducted extensive and in-depth research on green low-carbon psychology and behavior using the Technology Acceptance Model (TAM), Theory of Reasoned Action (TRA), and Theory of Planned Behavior (TPB; Liu et al., 2017). These traditional models of green consumption have received substantial empirical support for the relationship between attitude and behavioral intention; however, they often overlook the rationality of behavior. Traditional theoretical models, predicated on the “rational consumer” assumption, posit that consumers engage in cost-benefit analyses to maximize utility, thereby illustrating the limitations of rational models in the study of user behavior (Jackson, 2005). Concurrently, consumer decision-making in related fields is often influenced by factors such as social norms, moral cognition, and status quo bias (Abhyankar, 2022). Furthermore, traditional models overlook the impact of contextual factors, such as habitual consumption and aesthetic preferences, thereby neglecting the rationalization of human behavior (Temizkan, 2022). Consequently, traditional theories struggle to explain the non-linear relationships between variables; for instance, the influence of attitudes on behavior may exhibit threshold effects due to the influence of other factors, which the traditional linear structures of models like TPB fail to capture (Velnampy and Achchuthan, 2016). The rationality of consumer behavior transcends the traditional rational assumptions, emphasizing the decision-making processes of consumers under conditions of bounded rationality, including the cognitive rationality of information acquisition and processing (Estrada, 2010), as well as contextual and habitual factors in decision-making (Kevin et al., 2024). Moreover, in the specific context of food packaging, numerous studies have explored the impact of factors such as the promotion of purchase intention through eco-friendly packaging and consumer preferences for recyclable and biodegradable materials (Kevin et al., 2024), as well as moderating factors like price and social influence (Chopde, 2024). Integrating these factors into the model is essential, highlighting the advantages of BRT in analyzing the interactions between variables. To investigate the impact of food packaging factors on consumer behavioral rationality, this study incorporates behavioral rationality into the traditional theoretical model. The BRT model offers unique advantages in analyzing behavioral rationality. Consequently, the BRT model is employed to analyze the underlying mechanisms of consumers' environmental behavioral intentions. Furthermore, values significantly influence consumers, exerting a long-term and fundamental effect. Traditional theoretical models, when discussing behavioral intentions, often focus on attitudes, subjective norms, and perceived behavioral control, yet overlook values as an antecedent variable. Therefore, this research integrates values into the discussion to examine how food packaging influences subsequent consumer attitudes, behavioral rationality, and behavioral intentions (Claudy and Peterson, 2013). Methodologically, this paper utilizes BRT to organically integrate consumer values, behavioral rationality, attitudes, and behavioral intentions, thereby exploring the intrinsic mechanisms of food packaging's impact on consumer behavior. Explaining consumer behavioral intentions through supporting/rejecting reasons is crucial, with values serving to predict behavioral rationality, attitudes, and behavioral intentions (Sahu et al., 2020).

Building upon the preceding discussion, this study centers on elucidating the mechanism by which food packaging influences consumers' behaviors in rejecting food waste. The research methodology is as follows: Grounded in the Behavior Reasoning Theory (BRT), with the incorporation of consumer context as a supplementary element, thereby addressing the limitations of traditional theories in overlooking external factors. This approach employs “reason for” and “reason against” variables to interpret behavioral motivations. Furthermore, it integrates specific aspects of food packaging, including functionality, visual design, and informational design, to construct a multi-dimensional influence pathway. The aim is to investigate their specific impacts on consumers' behavioral intentions and their actions in rejecting food waste.

Food packaging design factors represent a multidimensional and interdisciplinary field. It encompasses visual design, material science, environmental impacts, and consumer behavior. Multiple factors influence consumers' purchasing decisions. These include, but are not limited to, the packaging's visual attributes (Marques da Rosa et al., 2018; Ahmadi et al., 2013), functional properties (Yokokawa et al., 2019; Obersteiner et al., 2021), environmental attributes (Lindh and Olsson, 2016; d'Astous and Labrecque, 2021), and social connotations (Wang, 2013). The visual attributes of packaging, such as its color and shape, significantly impact consumers' product preferences and their impressions of a product's taste and health benefits (Marques da Rosa et al., 2018). Many elements of packaging design affect consumer behavior, including aspects such as its informational content, content protection, identification, smart features, geometric shape, eco-friendliness, durability, and color (Konstantoglou et al., 2020). The environmental attributes of food packaging include the use of recyclable materials, reduced resource consumption, and minimized environmental pollution (Wang and Sun, 2010). Meanwhile, the functional attributes of food packaging, including its durability, shelf-life extension capabilities, and controlled release technologies, are crucial in maintaining food quality and safety (Shima Jafarzadeh et al., 2023). Thus, food packaging designs contribute to consumers' confidence, ensuring that the materials do not contaminate or migrate to the food, as well as preventing physical damage (Helanto et al., 2019). Moreover, the eco-friendliness of food packaging influences not only purchasing intentions but also environmental intentions. In addition, consumers' awareness and understanding of eco-friendly packaging impact their purchasing decisions. Consumers' awareness and knowledge of eco-friendly packaging vary across different countries and regions. For instance, South African consumers possess limited knowledge about eco-friendly packaging and exhibit fewer environmentally friendly behaviors related to packaging (Scott and Vigar-Ellis, 2014). In Ghana, research also indicates that although consumers' awareness of environmental issues positively and significantly impacts their green purchasing decisions, the effect of green packaging itself on consumers' purchasing decisions is not significant (Rokka and Uusitalo, 2008). This finding suggests that consumers' sociodemographic characteristics can influence their awareness, behavior, and expectations regarding the environmental sustainability of food packaging. Factors such as their gender, age, and educational level affect consumers' perceptions and behaviors regarding the environmental sustainability of food packaging (Chirilli et al., 2022). To encourage consumers to adopt more sustainable food practices, packaging designers must consider visual attributes, functional properties, environmental characteristics, and social significance, while also ensuring that this information effectively reaches consumers. Additionally, they must account for cross-cultural differences and individual consumer traits to develop personalized and effective packaging design strategies (Lindh and Olsson, 2016; Borgman et al., 2018).

Meanwhile, poor packaging design negatively affects consumers' purchasing decisions, perceptions, and behaviors, which indirectly influences their food-saving habits. Excessive marketing claims on packaging, such as health and nutrition assertions, can mislead consumers, causing food to appear healthier than it actually is, leading to over-purchasing (Obersteiner et al., 2021). Furthermore, misleading information on packaging designs can cause consumers to develop misconceptions, resulting in food waste behaviors (Chandon, 2013; Cairo, 2015). Improper packaging sizes or difficult-to-open packages may prevent consumers from correctly storing or using their food, ultimately leading to premature spoilage and disposal (Boyce et al., 2008). Unreasonable packaging sizes can lead to consumer misunderstandings regarding the product capacity. If a container's shape is particularly large, it may be perceived as containing more of the product, even if the actual capacity is small (Valerie Folkes, 2004). The packaging dimensions can significantly impact consumers' perceptions of the product capacity; these may involve its volume, weight, and overall impressions, thus affecting consumers' perceptions of the amount of food purchased and contributing to food waste (Ordabayeva, 2013). Additionally, poorly functional packaging designs can increase food waste and exacerbate environmental issues (Konstantoglou et al., 2020). The functionality of packaging is often inadequate, failing to protect its contents and extend the shelf life of food, which consequently contributes to consumers' food waste behaviors (Ploom et al., 2020). Consumers' perceptions of the risks associated with packaging influence their decisions regarding food waste, where health consciousness and environmental awareness emerge as key factors. Poorly functional packaging tends to exacerbate food waste (Zeng, 2021). Research indicates that consumers' perceptions of the risks of packaging directly affect their food waste behaviors (Zeng, 2021). A lack of awareness among consumers regarding packaging design often results in misunderstandings about food's freshness and shelf life, leading to food waste (Obersteiner et al., 2021). Insufficient consumer awareness of food packaging safety can result in the neglect of both safety and environmental considerations while selecting packaging, further contributing to food waste (Kang and Wang, 2021). Overly complex or difficult-to-understand information on packaging diminishes consumers' engagement and attentiveness to this information, adversely affecting their purchasing decisions and food consumption and thus causing food waste (Mruk-Tomczak et al., 2019).

The factors influencing consumer behavior regarding packaged food waste are multifaceted, with packaging design playing a crucial role in encouraging proper food consumption, waste reduction, and effective sorting and recycling of both food and packaging waste. As a critical intermediary between production and consumption, packaging design characteristics significantly influence consumer purchasing decisions. Furthermore, they indirectly affect the extent of food waste through storage practices and usage frequency. Research indicates that 60% of consumers try new products due to their packaging (Uddin et al., 2022), highlighting the impact of packaging design on consumer purchasing behavior. Packaging features also exert influence; studies show that visual and sensory cues affect consumer decision-making. For instance, red-orange is associated with “sweetness,” while blue-green suggests “health,” and angular designs enhance a modern aesthetic (Marques da Rosa et al., 2018). Transparent packaging enhances product authenticity, thereby influencing consumer visual experiences (Simmonds et al., 2017). Packaging quality is another significant factor in consumer decision-making. While some packaging utilizes recyclable materials, the cost often includes a premium (Zeng and Durif, 2019). Storage behavior is also linked to packaging quality; large-capacity packaging can lead to food waste and over-purchasing, while smaller packaging, though convenient, increases packaging costs, also incurring a premium (Wikström et al., 2019). Resealable packaging extends shelf life post-opening, reducing waste from spoilage, but it also increases packaging costs, again with a premium (Nemat et al., 2020a). This underscores that both the visual aspects and quality of packaging are critical factors influencing consumer decision-making. Consequently, visual attributes and packaging quality serve as communication channels, thereby encouraging consumers to conserve food and properly categorize food packaging waste (Nemat et al., 2020b). The functional and educational aspects of food packaging design, which can extend food shelf life and diminish food waste, are also critical determinants of consumer behavior (Almli et al., 2018).

Consumers' perceptions of and attitudes toward packaging largely determine whether they will reduce their food waste. In this context, consumers are unlikely to prioritize food packaging or reductions in food waste as the primary motivation in their purchasing decisions (Kapoor and Kumar, 2019). At the same time, consumers' inadequate knowledge about optimized packaging affects their ability to use these packages correctly at home (Obersteiner et al., 2021). Additionally, consumers' perceptions of green packaging also influence their decisions regarding food waste (Zeng, 2021). However, an insufficient understanding of the environmental impacts of food packaging may lead to choices that inadvertently negate consumers' eco-friendly intentions (Lindh and Olsson, 2016). The functionality of packaging plays a significant role in reducing food waste, as the functional information provided on food packaging can greatly impact consumers' choices regarding optimized packaging (Almli et al., 2018). Although the information conveyed through packaging design holds tremendous potential to reduce food waste, consumers still seldom pay proper attention to such information (Rokka and Uusitalo, 2008). Perceptions regarding packaging also reflect consumers' age, gender, and educational levels, as older consumers may generate more food waste (Hazuchova et al., 2020). Furthermore, consumers' perceptions and strategies regarding imperfect or secondary foods are essential in reducing food waste (Varese et al., 2022). Consumers' perceptions and behaviors play a crucial role in minimizing food waste. Additionally, consumers' awareness and actions regarding food waste issues are vital in achieving societal sustainability (Norton et al., 2022).

Based on the analysis above, we construct the following theoretical model, as shown in Figure 1.

This study examines the influence of food packaging factors on consumer food waste reduction across eight categories, drawing on previous research and literature. Based on this foundation, a survey was designed to assess how various food packaging design elements and consumer values impact their support for or rejection of food waste. The aim is to validate the mechanisms and effects of food packaging design factors on consumers' cognitive attitudes and behavioral intentions.

From the perspective of cognitive attitudes, consumers' values—The acceptance of food packaging—affect their attitudes toward rejecting food waste, exhibiting both positive and negative influences. This is adapted from the viewpoints of Chan and Xiao regarding values (Chan, 2022; Xiao-b, 2014; V1–V3). Additionally, the eight types of food packaging design factors are analyzed for their impact on consumers' cognitive attitudes toward rejecting food waste, referencing the visual factors in food packaging design as studied by Wang (2013) and Marques da Rosa et al. (2018; VA1–VA3), functional attributes (Yokokawa et al., 2019; Obersteiner et al., 2021; FC1–FC3), environmental awareness factors (Lindh and Olsson, 2016; d'Astous and Labrecque, 2021; EA1–EA3), and social factors (Wang, 2013; SA1–SA3). These four dimensions are measured for their positive impact on consumers' rejection of wasteful behaviors.

Furthermore, regarding the negative impact of food packaging design factors on consumers' attitudes toward rejecting food waste, it includes four dimensions. Zeng believes that excessive marketing information, health information, and nutritional information in food packaging design can lead to consumer food waste behavior and affect food conservation (Zeng, 2021; OI1–OI3). Ploom and Pentus believe that Unreasonable size in food packaging design can lead to food waste behavior among consumers due to unreasonable packaging size, difficulty in opening, and size that is not easy to carry and store, thereby affecting food conservation (Ploom et al., 2020; US1–US3). Chandon believes that low functionality in food packaging design can lead to food waste behavior among consumers due to the inability to protect food, short shelf life, and unfavorable transportation, thereby affecting food conservation (Chandon, 2013; LF1–LF3). Obersteiner and Cociancig et al. believe that the differences in other social and cultural levels lead to High awareness of environmental protection in food packaging design. The phenomenon of protection leads to food freshness cognitive bias, shelf life cognitive bias and other information bias, which lead to food waste behavior and affect the item design of food conservation (Obersteiner et al., 2021; Ploom et al., 2020; HA1–HA3).

On the perspective of rejecting food waste, according to Kunda and Zhuo, Ren et al., it is believed that behavioral rationality can enhance individual awareness and self-worth, promote the generation of positive attitudes toward food conservation and environmental protection, but attitudes may not necessarily translate into action intentions (Zhuo et al., 2022; Kunda, 1990). On this perspective, item design is carried out to reduce food waste, protect the environment, reflect personal morality, and demonstrate the willingness to take practical actions for Attitude toward reusing to waste food (Neff et al., 2015; Fornell and Larcker, 1981; RA1–RA3).

From the perspective of behavioral intention, Wang, Li, Guchi and Anon's research on behavioral intention, including the influence of self-belonging and self-efficacy and attitude on behavior (Wang et al., 2022; Guchi and Syafrizal, 2022). This dimension examines the personal intention to act against food waste, the sense of self belonging and self-efficacy of behavior, and the implementation of specific behaviors (Neff et al., 2015; Gelman, 2006; BI1–BI3).

Based on the above analysis, the items for questionnaire design are shown in Table 1.

To facilitate data collection, an online questionnaire survey method was selected. Given the geographically dispersed consumer distribution discussed earlier, an online survey approach was deemed more appropriate than offline methods. Consequently, all respondents were recruited through the Wenjuanxing online platform. Furthermore, this study received ethical approval from the School of Fine Arts, Nanjing Normal University. An online informed consent form was presented prior to the commencement of the survey. Due to the online nature of participation, all participants provided their consent verbally. Respondents were directed to the survey questionnaire only after confirming their informed consent by selecting the “agree” option. Participants who selected the “disagree” option were excluded from the survey. This approach ensured the collection of authentic and comprehensive consumer survey data. All respondents were consumers who had purchased or used food packaging. Moreover, this study adhered to the Declaration of Helsinki and was approved by the Institutional Review Board of the School of Fine Arts, Nanjing Normal University (ID: NO. NNU SFA-E-2024-002, June 19, 2024). All participants provided informed consent, and their privacy was strictly protected. Personal information was kept confidential throughout the study. Participants were informed of their voluntary participation and their right to withdraw at any time. All respondents were adults, with no minors included.

Data collection for this study utilized an online questionnaire format, distributed via online platforms to a broad demographic to ensure the sample's representativeness and mitigate regional biases, thereby enhancing the validity of the collected data. The survey was conducted between July and September 2024, utilizing the “Wenjuanxing” data survey platform to collect a total of 562 samples. After eliminating invalid samples, 513 valid samples were ultimately obtained, resulting in an effective rate of 91.3%. The reasons for sample exclusion were three-fold: (1) Duplicate submissions from the same IP address were consolidated, with only one retained as a valid sample; (2) Surveys completed in an excessively short timeframe were deemed invalid; (3) Samples containing sequences of repeated numerical entries were excluded. All items were assessed using a Likert scale, with scores ranging from 1 (strongly disagree) to 7 (strongly agree). Prior to participation, an online informed consent form was presented. Given the online nature of the survey, verbal consent was obtained from all respondents. Furthermore, participants were required to confirm their comprehension of the informed consent and indicate their agreement by selecting an affirmative option to access the questionnaire. At the same time, the respondents could only participate in the questionnaire survey once and could not participate in the questionnaire repeatedly, which ensured the accuracy of the questionnaire survey. Non-agreement resulted in the termination of the survey. This study did not involve minors, as they are considered to lack full behavioral capacity due to their developmental stage.

The questionnaire comprised 34 items, yielding 513 valid responses. Statistical power analysis, conducted using G^*Power, was performed for a linear multiple regression model, specifically focusing on the R² deviation in a fixed model. As illustrated in Figure 2, the input parameters included an effect size f² of 0.15, a significance level (α error probability) of 0.05, and a desired statistical power (1-β error probability) of 0.90, representing a 90% probability of correctly rejecting the null hypothesis. Consequently, under these parameters, G^*Power calculated a required total sample size of 108. This sample size ensures sufficient statistical power to detect effects within the model, given the specified effect size and significance level. The actual statistical power was 0.9013736, slightly exceeding the target of 0.90, indicating that the sample size meets the study design requirements and supports the reliability of the findings. Given the effective sample size of 513, which substantially surpasses the minimum requirement of 108, the sample size is deemed adequate for the study.

Therefore, the data analysis proceeded based on this standard, primarily utilizing SPSS 22.0 and Smart-PLS 4.1 software. Descriptive statistical analysis of the sample demographic variables is presented in Table 2.

The survey encompassed 513 questionnaires, with sample regions strategically selected across 19 provinces, including Jiangsu, Fujian, Sichuan, Shandong, Anhui, Zhejiang, Hunan, Jiangxi, Hubei, Heilongjiang, and Guangdong, thereby demonstrating the comprehensiveness of the sample collection. Nevertheless, regional disparities in economic development influenced the distribution of consumers who had utilized or experienced food packaging. Notably, a significant proportion of respondents hailed from economically advanced regions such as Jiangsu, Fujian, Guangdong, and Zhejiang, indicating a heightened level of attention toward food packaging in more developed areas, thus validating the authenticity of the sample collection. The online survey participation comprised a slightly higher number of male respondents compared to female respondents, with the age distribution predominantly skewed toward young and middle-aged individuals. In terms of the occupational distribution, government workers and public institution staff accounted for the largest proportion, followed by freelancers and students, and other occupations accounted for approximately 10%. At the same time, according to the questionnaire, the proportion of respondents who overbought packaged food was as high as 86.8%, of which 31.8% of these respondents often overbought packaged food. When faced with a large amount of excess packaged food, 57.3% of respondents stated that they would directly discard it, and only 28.1% were willing to continue to eat it. Meanwhile, 71.9% of the 513 respondents would throw food away or use it for other purposes, and 49.7% did not feel sad or upset about food waste. This means that nearly half of the respondents were indifferent to food waste. In summary, the proportion of the sample that responded to each part of the questionnaire was reasonable, meeting the requirements of the research. According to the data on the respondents' activities related to buying packaged food, dealing with excessive packaged food, and discarding packaged food, it is necessary to study the influence of food packaging design factors on consumers' behavior in terms of rejecting food waste.

Reliability refers to the consistency or stability of measurement results. Research indicates that a Cronbach's α value between 0.7 and 0.9 is generally considered to reflect good internal consistency (Oviedo and Campo-Arias, 2005). The Cronbach's α values for each measurement variable, calculated using SPSS 22.0, were all greater than 0.7, while the overall Cronbach's α for the questionnaire exceeded 0.6 (Yin and Wei, 2012; Chan and Idris, 2017). The total correlations after item deletion were all above 0.5 (Yin and Wei, 2012), indicating that the constructs in this study possess good reliability. The variance inflation factor (VIF) serves as a metric for assessing multicollinearity among independent variables, quantifying the extent to which each predictor variable is linearly explained by others. A higher VIF value indicates a more severe multicollinearity issue (Willis and Perlack, 1978). Research suggests that a VIF exceeding 5 suggests the presence of some degree of multicollinearity within the measurement scale (Alhamami et al., 2024); ideally, the VIF should be below 3.3 (Bria et al., 2021). At this level, the correlation between predictor variables is low, resulting in more stable regression estimates, which is conducive to constructing robust analytical models. As shown in Table 3, the VIF values for all variables are below 3.3, indicating the absence of multicollinearity issues. Overall, the Cronbach's alpha coefficient (a reliability indicator) and the VIF values (a multicollinearity indicator) in Table 3 collectively demonstrate that the measurement scale used in this study has good reliability and does not exhibit multicollinearity issues, thus supporting the construction of a stable and reliable analytical model.

Utilizing SPSS 22.0, the Kaiser-Meyer-Olkin (KMO) measure and Bartlett's test of sphericity were conducted. Relevant studies indicate that the KMO value serves as an indicator of the suitability of data for factor analysis, with a KMO value greater than 0.7 generally considered acceptable for this purpose (Campos et al., 2013; Yordanova and Krastev, 2018; Sara et al., 2020). Furthermore, a significance level of less than 0.05 in Bartlett's test suggests that the covariance matrix of the variables is not an identity matrix, indicating significant correlations among the variables, thus making them suitable for factor analysis (Wu and Wong, 2003; Dyer and Keating, 1980). As shown in Table 4, the KMO values for the variables ranged from 0.704 to 0.829, all exceeding the threshold of 0.7, and the significance of Bartlett's test was less than 0.05, confirming that all variables passed the Bartlett's test of sphericity. This establishes a solid foundation for the factor analysis of the data. Consequently, principal component analysis was employed for the factor analysis of the variables, revealing that each variable could extract a factor with an eigenvalue greater than 1, and the cumulative variance contribution of all variables exceeded 50%. This indicates that the factors identified in this study provide a robust explanation for the variables analyzed (Feng, 2012). Additionally, the commonalities for all items were greater than 0.5 (Liang and Khan, 2024), and the factor loadings were all above 0.6 (Elias et al., 2020), which are within reasonable limits. In summary, this study concludes that the survey results exhibit strong unidimensionality, indicating that there is no issue of common method bias in the data, as illustrated in Table 4.

When Cronbach's α and composite reliability exceed the 0.7 threshold, the scale is deemed acceptable, as indicated in Table 3, suggesting that the scale demonstrates satisfactory reliability. Validity analysis was conducted using Smart-PLS4.1, with validity assessed through Average Variance Extracted (AVE) and Composite Reliability (CR) values. As presented in Table 5, the CR values range from 0.768 to 0.862, and the AVE values range from 0.683 to 0.722. According to Fornell et al., both CR and AVE values are critical for assessing convergent validity, with CR values needing to be greater than 0.7, and 0.5 being the minimum acceptable AVE value (Fornell and Larcker, 1981; Churchill, 1979). The results in Table 5 indicate that all AVE values surpass the 0.5 threshold, and all CR values exceed the 0.7 threshold, confirming that all variables exhibit robust convergent validity.

As shown in Table 6, the outer loadings derived from the indicators of each construct in this study exceed the results of each additional construct's cross-loadings. It is noteworthy that cross-loadings are commonly employed as a preliminary step in assessing the discriminant validity of indicators (Leguina, 2015). Therefore, as indicated in Table 6, all variables demonstrate good discriminant validity.

The Heterotrait-Monotrait Ratio (HTMT) is an index used to assess the discriminant validity of variables within a structural equation model. HTMT values below 0.85 are generally considered to indicate that the variables exhibit satisfactory discriminant validity (Henseler et al., 2015). It is essential to evaluate these metrics to mitigate potential multicollinearity issues arising from highly correlated constructs. The results indicate that all HTMT scores are within acceptable limits ( ≤ 0.85), with values ranging from 0.528 to 0.825 (Table 7). Therefore, it can be concluded that there is sufficient evidence to support the model's satisfactory discriminant validity in this study.

Discriminant validity refers to the differences among various latent variables. According to Fornell's recommendations, this can be verified by comparing the correlation coefficients between latent variables with the square root of the Average Variance Extracted (AVE). If the square root of the AVE is greater than the correlation coefficients of the variables, it indicates that the scale possesses good discriminant validity (Fornell and Larcker, 1981). As shown in Table 7, the square roots of the AVE for the latent variables are all greater than the correlation coefficients among the latent variables. Therefore, significant correlations exist among the main variables of this study, namely V, BI, EA, FC, HA, LF, OI, RA, SA, US, and VA, and the square roots of the AVE exceed the correlation coefficients between the variables. Moreover, the correlations of the exogenous structures are all less than 0.85 (Gelman, 2006), indicating good discriminant validity. According to Table 8, the square roots of the AVE for the latent variables are greater than the correlation coefficients among the latent variables, and the correlations of the exogenous structures are all less than 0.85. Thus, it can be concluded that all variables exhibit good discriminant validity.

SRMR is one of the commonly used fit indices in Structural Equation Modeling (SEM) to evaluate the degree of fit between the model and the data. A smaller SRMR value indicates a better fit of the model to the data. Related research shows that SRMR performs well in handling ordered factor analysis models, particularly when data distribution is abnormal; in such cases, the testing results of SRMR are more reliable. An SRMR value less than 0.08 indicates a very high fit of the model to the data (Ximénez et al., 2022). Furthermore, an NFI greater than 0.8 reflects a high fit of the model; the closer the value is to 1, the stronger the model fit, signifying that the model can accurately capture the structures and relationships within the data (Gowen et al., 2011). As shown in Table 9, the SRMR value of 0.058 is below the threshold of 0.08, indicating a good model fit. Additionally, the NFI value of 0.964 is greater than 0.8, indicating a high model fit.

Finally, the hypotheses were tested using PLS-SEM, and path coefficients were calculated by employing the Bootstrap method with 5,000 resamples, as shown in Table 10. The P-values for the second-order model hypotheses were all found to be less than 0.05, indicating statistical significance. Therefore, the second-order model hypotheses were validated. As presented in Table 8, the statistical results demonstrate that all eight second-order model path relationships are significant, indicating that visual attributes, social attributes, environmental attributes, and functional attributes serve as positive reasons for consumers to reject food waste, while excessive information, low functionality, unreasonable sizes, and overly high cognitive load are identified as negative reasons for consumers' rejection of food waste.

The analysis of the path coefficients from Table 11 indicates that values (The acceptance of food packaging) significantly influence the reasons for support (T = 84.088, P < 0.01) and the reasons for rejection (T = 83.473, P < 0.01), thereby providing support for hypotheses H1 and H2. However, the results regarding attitudes toward food conservation (T = 1.215, P > 0.01) are not significant, leading to the rejection of hypothesis H3. Furthermore, the reasons for support significantly affect attitudes toward rejecting food waste (T = 8.302, P < 0.01) and the behavioral intention to reject food waste (T = 4.372, P < 0.01), confirming hypotheses H4 and H6. Similarly, the reasons for rejection significantly influence consumers' attitudes toward rejecting food waste (T = 7.333, P < 0.01) and their behavioral intention to reject food waste (T = 6.131, P < 0.01), thus supporting hypotheses H5 and H7. Lastly, consumers' attitudes toward rejecting food waste significantly impact their behavioral intention to reject food waste (T = 1.758, P < 0.05), confirming hypothesis H8.

The results of hypothesis testing are shown in the above table, as shown in Figure 3.

Through empirical analysis, this study reconfirms the influence of consumer values and food packaging design elements on consumer emotions, clarifying their crucial roles in shaping consumers' conservation attitudes and behavioral intentions, thereby further enriching and refining the theoretical framework of the relationship between consumer behavior and food packaging design. Simultaneously, a novel theoretical model is innovatively constructed to deeply explore the relationship paths between food packaging design elements and consumers' attitudes and behavioral intentions toward rejecting food waste, providing a new perspective and research direction for the theoretical development in the field of food packaging design, which has significant reference value for the theoretical research of food packaging design. Furthermore, the research not only analyzes the impact of food packaging design elements and consumer values on consumers' supporting/rejecting reasons, attitudes, and behavioral intentions but, more importantly, builds a theoretical framework for future analysis of food packaging design guiding consumer behavioral intentions, which is conducive to the in-depth development of subsequent related research in this field.

The research findings offer valuable insights for the advancement of design elements in food packaging, encompassing visual appeal, functionality, environmental sustainability, and social impact. These insights assist designers in understanding the roles of various elements in influencing consumer attitudes and behavioral intentions, thereby facilitating the optimization of food packaging design. Furthermore, the research outcomes delineate key priorities for the future development of food packaging, emphasizing the importance of enhancing consumers' attitudes and behavioral intentions toward food conservation. This directs designers to prioritize the strategic application and integration of packaging design elements, tailoring designs to the specific attributes of different food products. This approach aims to capture consumer attention and encourage food-saving behaviors. The study also identifies unfavorable aspects of food packaging design from the consumer's perspective, offering recommendations to mitigate these issues. This contributes to a more positive consumer experience, strengthens the perceived value of food conservation, and ultimately fosters the formation of intentions to conserve food. In addition, the research underscores the significance of innovative food packaging design, encouraging designers to leverage design effects and objectives to reshape consumer perceptions of food waste. This, in turn, promotes the adoption and practice of food-saving behaviors among consumers, providing direct guidance for food packaging design practices.

Consequently, this study has yielded significant findings at both theoretical and practical levels. Theoretically, it addresses research gaps, refines the existing theoretical framework, and establishes a robust foundation for future academic investigations. Practically, it offers clear guidance and actionable strategies for the food packaging design industry, assisting the sector in promoting consumer behaviors conducive to food conservation through optimized packaging design. The research outcomes hold immediate practical significance and provide substantial support and positive direction for the future advancement of food packaging design, both in terms of theoretical research and practical innovation, thereby contributing to the evolution of the food packaging field toward greater scientific rigor and enhanced efficiency.

This study delves into the correlation between consumer values and food packaging design elements, with a specific focus on their impact on consumers' value attitudes and behavioral intentions regarding food waste reduction. Positive food packaging factors encompass visual appeal, functional utility, environmental sustainability, and social significance. Conversely, negative factors include information overload, diminished functionality, inappropriate sizing, and elevated cognitive costs. By integrating these factors, a comprehensive research model was constructed, effectively capturing the behavioral shifts experienced by consumers after interacting with food packaging. This model offers significant guidance for the advancement of food packaging design.

Empirical data validation robustly demonstrates a significant association between food packaging design elements and consumers' attitudes and behavioral intentions. Notably, consumers exhibit a high degree of endorsement for the conservation and environmental principles embodied in food packaging design, coupled with a sustained intention to act accordingly. This finding not only elucidates consumers' value preferences in food packaging selection but also provides direction for the future development of food packaging design.

The findings of this study offer significant insights and practical value for packaging designers. They can leverage these results to gain a deeper understanding of consumer needs and expectations, enabling more precise articulation of design concepts. This approach ensures that food packaging effectively addresses consumer requirements on a functional level while also resonating with their values. By optimizing food packaging design, we anticipate fostering a positive consumer attitude and behavior toward reducing food waste, thereby contributing to resource conservation and environmental protection. Future research should build upon these findings, exploring how food packaging design can further influence consumer behavior and how innovation in this area can be achieved across diverse cultural contexts.

The present study acknowledges certain limitations inherent in its methodology. The primary reliance on online questionnaires for data collection, while facilitating rapid and large-scale information gathering, introduces potential biases in sample representativeness. Specifically, the demographic profile of online survey respondents may skew toward younger age groups, leading to an underrepresentation of both adolescent and elderly populations. Adolescents, as the future drivers of societal change, are a critical demographic for cultivating food conservation awareness, given their formative consumption patterns. Simultaneously, the burgeoning elderly population in China, influenced by established habits and potentially slower adoption of new concepts, warrants significant attention in consumer behavior research.

To address these limitations, future research should adopt a mixed-methods approach. Quantitatively, this could involve optimizing online questionnaire distribution channels, collaborating with schools and community centers to target specific demographics, and expanding the sample size of adolescent and elderly participants. Furthermore, incorporating offline surveys, such as paper-based questionnaires administered in senior activity centers and schools, would enhance sample diversity. Qualitatively, in-depth exploration of the underlying motivations and behavioral drivers related to food conservation among adolescents and the elderly is essential. For instance, in-depth interviews and discussions can be employed to ascertain the conservation education they received within their familial and scholastic environments, as well as the influence of peer groups on their food conservation attitudes. For the elderly demographic, an exploration of traditional life experiences and memories of material scarcity can illuminate how these factors have shaped their current consumption habits and the underlying reasons for their resistance to emerging conservation concepts. Furthermore, organizing discussions on food conservation topics among adolescents and senior citizens from diverse age groups and backgrounds can facilitate the observation of idea exchange and consensus-building within these groups, thereby yielding richer and more profound insights.

Integrating large-scale data statistical analysis from quantitative research with the in-depth insights from qualitative research allows for a comprehensive understanding of the overall trends and disparities in food conservation awareness and behavior across different groups at a macro level. Simultaneously, it enables an in-depth analysis of the complex psychological and social factors underlying individual behaviors at a micro level. This comprehensive approach provides a more complete and thorough understanding of the research subjects, offering a more robust theoretical and data foundation for refining related research, thereby enhancing the practical guidance value and social significance of the research findings.