Cultural Worldview and Rural Consumer Preferences for Genetically Modified Foods
Department of Agricultural Economics and Agribusiness, University of Arkansas, Fayetteville, AR 72701, USA
Sustainability 2025, 17(15), 6843; https://doi.org/10.3390/su17156843
Submission received: 27 May 2025
/
Revised: 11 July 2025
/
Accepted: 17 July 2025
/
Published: 28 July 2025
Abstract
The consumer acceptance of genetically modified (GM) foods varies widely based on personal values and social context. This study investigates how cultural worldviews—measured through the Individualism–Communitarianism and Hierarchical–Egalitarianism dimensions—affect willingness to pay (WTP) for food labeling attributes related to GM content. We surveyed 932 U.S. primary grocery shoppers and conducted a discrete choice experiment (DCE) with poultry product options. Using a Mixed Logit model and supplemental cognitive data from Query Theory, we find that rural individualists are more accepting of GM-labeled products. At the same time, urban communitarians show a stronger preference for non-GM labels. These results offer insight into how values and thought patterns shape food technology perceptions, with implications for communication and policy.
Keywords:
cultural cognition theory; query theory; consumer preferences; genetically modified foods; rural consumersJEL Classification:
Q16; Q18; C91. Introduction
Consumer preferences for genetically modified (GM) and non-GM foods vary significantly across demographic and cultural groups [1,2]. While much prior work focuses on rurality as a geographic factor, this study investigates the deeper role of cultural worldviews in shaping consumer preferences. Specifically, we draw on Cultural Cognition Theory (CCT) to assess how Individualism–Communitarianism (IC) and Hierarchical–Egalitarianism (HE) orientations influence willingness-to-pay (WTP) for GM-related food labels. In addition, we incorporate Query Theory (QT) to explore how consumers mentally organize information about food risks and benefits.
This study contributes to ongoing discussions about the sustainability of food systems by examining how cultural values and decision-making processes influence consumer preferences for GM food products. While biotechnology is often promoted as a tool for increasing agricultural efficiency, reducing environmental impacts, and enhancing food security, its success depends in part on consumer acceptance. Cultural cognition and rural identities can shape responses to GM labeling, thereby influencing the broader adoption of sustainable innovations. By analyzing these behavioral dynamics, this study sheds light on key cultural barriers and opportunities related to sustainable food consumption.
Risk perception plays a crucial role in shaping consumer acceptance of GM foods, particularly in rural communities where attitudes toward technological adoption vary [3]. Some studies suggest that rural consumers who exhibit higher risk aversion are more likely to delay the adoption of GM-related agricultural technologies, mirroring trends observed among farmers hesitant to adopt GM crops due to perceived uncertainties [4]. Economic considerations further impact these preferences, with factors such as price, freshness, and accessibility shaping consumer choices [5]. Given the prominence of local food systems in rural economies, the way GM foods are perceived often aligns with community values and available information about food production methods [6].
CCT posits that individuals align their beliefs with cultural values, affecting their perception and acceptance of polarizing issues [7]. This theory, integrating the cultural theory of risk and the psychometric paradigm, suggests that people’s risk perceptions and preferences are influenced by their cultural worldview. This study extends prior research in cultural cognition [7] by exploring how rural cultural worldviews shape consumer preferences for GM and non-GM foods. Specifically, this study investigates how cultural worldviews influence WTP and discount preferences for food labeling across rural, suburban, and urban settings. By integrating CCT with QT, this study provides a cultural perspective and a cognitive account of the formation of preferences.
Consumer education has also been key in shaping rural GM food acceptance [8]. While risk perception and economic incentives play an important role in consumer decision-making, studies suggest that access to transparent information about GM foods, including proper labeling, can mitigate skepticism and foster more favorable attitudes toward biotechnology [6,8]. Thus, understanding the intersection of cultural values, risk perception, and economic priorities is critical to effectively addressing rural consumer concerns about GM foods.
This study builds upon existing research by integrating CCT with QT to examine both cultural and cognitive mechanisms influencing WTP for GM and non-GM foods. CCT helps explain how cultural values influence consumer preferences, while QT provides insight into the sequence and impact of thought processes shaping these preferences. Using a discrete choice experiment (DCE) and consumer survey, this study contributes to the growing body of literature on rural consumer behavior, risk perception, and food technology acceptance. Figure 1 illustrates the conceptual framework guiding this analysis, showing how locale and worldview shape cognitive processing and ultimately influence valuation outcomes.
2. Cultural Cognition Theory (CCT)
Cultural Cognition Theory (CCT) explores how individuals align their beliefs with their cultural values [9,10]. This theory integrates the cultural theory of risk and the psychometric paradigm [11]. The cultural theory of risk suggests that individuals perceive risks in ways that affirm their preferred way of life, organized along two key dimensions: group and grid. The group dimension spans from individualistic worldviews—favoring autonomy and competition—to communitarian worldviews—prioritizing collective welfare and social solidarity. The grid dimension ranges from hierarchical worldviews—supporting stratified social roles and authority—to egalitarian worldviews—favoring equality and inclusive decision-making.
CCT posits that people form opinions on contested issues like GM foods not purely through factual reasoning but via identity-protective cognition. This means individuals selectively accept or dismiss information in ways that reinforce and defend their cultural identity. For example, individualists may be more open to GM food technologies due to a preference for personal choice and market-driven solutions. In contrast, communitarians may view them with skepticism, emphasizing environmental or ethical concerns.
CCT has been used to explain differences in attitudes toward technologies, including GM foods, vaccination, nuclear power, and climate change. In food policy, worldview orientations have been shown to influence support for mandatory labeling and regulatory oversight [7]. This study extends prior work by testing whether cultural worldviews are more predictive than rurality alone and by integrating CCT with cognitive data from Query Theory.
3. Query Theory (QT)
Query Theory (QT) offers a cognitive lens through which to understand how preferences are formed during decision-making [12,13]. According to QT, individuals construct preferences by internally generating a sequence of thoughts or “queries” about the options available. These queries are shaped by the order in which they arise—initial thoughts tend to dominate due to a psychological phenomenon called output interference, which inhibits the recall of subsequent, contradictory ideas.
In this study, QT is used to explore how the sequence and balance of thoughts about GM foods vary across individuals with different cultural worldviews. We employ two cognitive measures from QT: (1) the Standardized Median Rank Difference (SMRD), which captures whether value-increasing or value-decreasing thoughts are listed earlier in a decision-making task, and (2) the Balance of Evidence (BOE), which reflects the net difference between value-increasing and value-decreasing thoughts. These measures help quantify cognitive bias and provide insight into the internal reasoning processes that accompany cultural predispositions. QT complements CCT by revealing not just what people think about GM foods but how they think about them—illustrating how cultural worldviews manifest in the sequence and framing of decision-making.
4. Study Design and Experimental Methods
4.1. Survey Design and Discrete Choice Experiment
Data were collected through a national online survey conducted using Sawtooth Software, with respondents provided by Survey Sampling International [14], forming a nationally representative panel of 932 primary grocery shoppers. The sample was balanced across key sociodemographic characteristics and the four main U.S. Census regions. All respondents were 18 years or older and provided informed consent. Although the overall sample was designed to be broadly representative of U.S. food purchasers, no post-stratification weights were applied to adjust for differences in the subgroup size across geographic locales. As such, while comparisons between rural, suburban, and urban respondents are reported, subgroup-level results should be interpreted with appropriate caution due to the potential underrepresentation of rural participants.
The survey gathered data on respondents’ policy attitudes, food labeling preferences, and demographics. The discrete choice experiment (DCE) asked respondents to select between poultry products that varied by their GM labeling, production location, and carbon footprint across eight choice scenarios (Table 1). Each scenario included two product options and a “no-buy” option. Price levels (USD 2.99, USD 6.99, USD 10.99, USD 14.99) were based on retail market prices and USDA reports [15]. Carbon footprint levels (79, 90, 112 oz CO2e) reflected average emissions benchmarks in poultry production [16]. A sequential, D-efficient design minimized the respondent burden and optimized statistical efficiency [17]. A 250-respondent pilot informed the full survey design, which was split into four blocks of eight tasks to reduce choice fatigue. Respondents were randomly assigned to one block.
4.2. Cultural Worldview Dimensions Methods
Cultural worldviews were measured using four-item scales developed and validated by Kahan and colleagues as part of the Cultural Cognition Project [2,18]. These scales assess respondent alignment on two dimensions: Hierarchy–Egalitarianism and Individualism–Communitarianism. The reliability and validity of these short-form instruments have been established through extensive psychometric testing in prior research. We implemented these well-established measures without modification. The Hierarchy–Egalitarianism dimension measures preferences for a societal structure based on clearly defined social roles versus equality in power and resource distribution. The Individualism–Communitarianism dimension assesses whether individuals see prosperity as a personal responsibility or a collective priority [2].
Prior studies have validated the reliability of these four-item scales for identifying cultural worldviews [18], with factor analysis confirming their psychometric soundness. This analysis revealed that responses to these items align with two orthogonal factors, indicating distinct Hierarchy–Egalitarian and Individualism–Communitarian dimensions. Measurement precision for these scales is demonstrated by acceptable Cronbach’s alpha values: 0.779 for Individualism–Communitarianism and 0.710 for Hierarchy–Egalitarianism, as outlined in Table 2. This study reports standardized factor scores, with negative values indicating a tendency towards individualism or egalitarianism and positive values suggesting a hierarchical or communitarian disposition, respectively. This standardization facilitates analysis of cultural worldview impacts, particularly focusing on hierarchical individualists and egalitarian communitarians for clarity and relevance to this study’s hypotheses. While confirmatory factor analysis indices were not recalculated in this study, the short-form scales used here have been previously validated through confirmatory factor analysis and have been shown to align with two distinct worldview dimensions [2]. Sample characteristics, summarized by cultural worldview or “group–grid” dimensions, are shown in Table 3.
Table 1.
Attributes and levels in discrete choice experiment.
| Attributes | Levels |
|---|---|
| Price | USD 2.99 |
| USD 6.99 | |
| USD 10.99 | |
| USD 14.99 | |
| GM Content | No GM information |
| Non-GMO verified | |
| Contains GM | |
| Carbon Footprint | No carbon footprint information |
| 79 oz CO2e/lb (low) | |
| 90 oz CO2e/lb (medium) | |
| 112 oz CO2e/lb (high) | |
| Local | No information |
| Local production |
4.3. Query Theory Methods
To investigate the impact of cultural worldviews on decision-making valuations, this study applied a thought listing approach as modified for discrete choice experiments by Kemper et al. [19,20], originally outlined by Johnson et al. [12] and Weber et al. [13]. During each choice task, participants were prompted to list reasons guiding their decisions. These thoughts were coded as “value-increasing” or “value-decreasing” to capture cognitive biases in decision-making. For example, “health risks” or “I do not like GMOs” would be categorized as value-decreasing aspects, while “low cost” or “local is better” would be value-increasing. Each choice task was analyzed individually to maintain consistency with the measurement of other individual-level variables.
The Balance of Evidence (BOE) and Standardized Median Rank Difference (SMRD) metrics quantified the sequence and balance of thoughts, enabling assessment of the influence of cultural identity on cognitive evaluation patterns across rural, suburban, and urban contexts. This inclination is quantified through the Standardized Median Rank Difference (SMRD), calculated using the formula:
where MRd is the median rank of value-decreasing thoughts in a participant’s sequence, MRi is the median rank of value-increasing aspects in a participant’s sequence, and n is the total number of thoughts in a participant’s sequence. The SMRD score, therefore, ranges from −1 (all value-decreasing aspects are listed first) to 1 (all value-increasing aspects are listed first), providing a numerical representation of the decision-making bias towards positivity or negativity in aspect listing. BOE was calculated as the simple difference between value-increasing and value-decreasing aspects (MRi − MRd), at the choice task level.
2 (MRi − MRd)/n
4.4. Econometric Methods Used to Estimate WTP
Preferences were analyzed using a Mixed Logit model specified in WTP space [17], consistent with Random Utility Theory [21] and Lancaster’s Consumer Theory [22]. In line with Scarpa et al. [23], the Mixed Logit model was specified in WTP space rather than preference space. Estimating the utility function directly in WTP space offers greater control over the distribution of WTP, reduces the incidence of implausibly large WTP values, and allows for a clearer interpretation of price sensitivities. The utility function for individual i choosing option j in choice situation t is specified as (2)
Uijt = α [θ1NONE + PRICEijt + θ2NGEijt + θ3GMEijt + θ4LOEijt + θ5MDEijt + θ6HIEijt + θ7LCEijt] + ϵijt
The alternative-specific constant (NONE) captures the utility of the “no-buy” option and is dummy-coded as 1 for the no-buy option and 0 otherwise. Price represents the cost of each product, which is modeled with a fixed coefficient, and takes on four experimentally designed levels: USD 2.99, USD 6.99, USD 10.99, and USD 14.99. Each of the non-price attributes, including Non-GM (NGE), Contains Genetically Engineered Ingredients (GME), Low Carbon Footprint (LOE), Medium Carbon Footprint (MDE), High Carbon Footprint (HIE), and Local Production (LCE), are effects-coded. The effects coding for these attributes is specified as follows: 1 if the product displays the label, −1 if the label is absent, and 0 otherwise. This coding approach ensures that the attribute effects are interpreted relative to the overall sample mean effect of each attribute.
The parameters for non-price attributes are modeled as random parameters with a normal distribution to account for heterogeneity in respondents’ preferences. However, to facilitate interpretation in WTP space, the price parameter is modeled using a fixed distribution. This approach follows Revelt and Train [24], which provides a direct way to estimate WTP for each attribute. Thus, θi represents the marginal WTP estimates for each non-price attribute. The error term ϵijt represents unobserved factors affecting utility and follows an extreme value type-I (Gumbel) distribution, which is independent and identically distributed (iid) over alternatives, consistent with the assumptions of the Mixed Logit model. To further capture preference heterogeneity across respondents, the model incorporates correlated errors and error components, which allows for greater flexibility in estimating individual-specific random parameters. Estimations were conducted using NLOGIT 6 with 1000 Halton draws to improve simulation accuracy for the random parameters [25]. This approach ensures a high level of precision in the WTP estimates and accounts for complex correlation structures in preferences across attributes.
4.5. Multiple Regression Analysis
Finally, WTP estimates were used as dependent variables in multiple regression analyses where CCT and QT variables were included among other independent variables. For the multiple regression models, the primary independent variables are the group (Individualism–Communitarianism) and grid (Hierarchical–Egalitarianism) measures from the factor analyses, as shown in Table 2. Utilizing these scales allows us to examine the relative influence of each cultural dimension on preferences, providing a deeper understanding of how distinct worldview dimensions shape consumer attitudes. To capture the potential joint influence of worldviews, a product interaction term is included (IC × HE).
In addition to cultural worldview dimensions, the final multiple regression models incorporate other explanatory factors. These include sociodemographic factors (age, number of children in the household, race, income, and locale) [7], risk preference as it relates to food consumption [26], and consumer behavior measures related to food labeling practices—such as frequency of reading food labels, knowledge of GM food consumption, and trust in sources of GM food information [27,28,29]. The final regression variables were selected using a backward stepwise approach. Variables were removed based on their significant contribution to model fit until removing further variables resulted in no improvements. A variance inflation factor (VIF) was calculated to detect multicollinearity in the models.
5. Results
This study’s analysis reveals notable variations in GM food preferences across rural, suburban, and urban communities, collectively referred to as the “locale.” Factor analysis results (Table 4) indicate significant differences in the Individualism–Communitarianism (IC) dimension by locale. Rural respondents display a more individualistic orientation than their suburban and urban counterparts. No significant differences are found for the Hierarchical–Egalitarian (HE) dimension across locales, suggesting that the primary worldview difference relates to individualism rather than hierarchy.
The full WTP space model results are presented in Appendix A Table A1. Individual-level WTP estimates from this model serve as the basis for the subgroup analysis. Table 5 offers a summary of mean WTP values by group-grid dimensions. Mean WTP values for non-GM and GM labels vary significantly across locale (Table 6), with rural participants showing the lowest WTP for non-GM labels (mean = USD 1.48/lb) and requiring the smallest discounts to accept GM labels (mean = −USD 1.53/lb). In contrast, urban participants exhibit higher premiums for non-GM labels (USD 3.02/lb) and require larger discounts for GM labels (−USD 2.42/lb). These differences highlight the role of locale in shaping labeling preferences.
Table 7 further disaggregates WTP values by both locale and cultural worldview dimensions. Rural individualists exhibit the lowest WTP for non-GM products (USD 3.80/lb) and the smallest discount requirement for GM labels (−USD 2.89/lb), whereas urban communitarians show the highest WTP (USD 5.45/lb) and the greatest aversion to GM products (−USD 3.79/lb). These subgroup differences underscore the importance of considering both cultural and geographic factors. Table 8 presents pairwise comparisons of WTP values for non-GM and GM labels across group–grid subgroups and locales. The pattern reveals a consistent divergence, particularly between rural individualists and urban communitarians. For instance, the difference in non-GM premiums between these two groups is USD 2.18/lb. While many of these differences are statistically significant, they are also meaningful from a consumer economics perspective given their relative size.
Table 9 summarizes the Standardized Median Rank Difference (SMRD) and Balance of Evidence (BOE) metrics across locales. SMRD scores do not differ significantly by locale, suggesting a similar order of thoughts during decision-making. In contrast, BOE scores indicate that rural participants tend to generate more value-decreasing thoughts (mean BOE = −0.172), while urban participants show slightly positive BOE scores (0.017), suggesting a greater emphasis on value-increasing considerations. Table 10 presents SMRD and BOE metrics by both locale and cultural worldview dimensions. Rural individualists again show the most negative BOE values (−0.14), while urban communitarians report a more positive cognitive balance (0.17), reinforcing the pattern seen in WTP results. These data suggest that worldview-linked cognitive patterns are associated with valuation differences. Table 11 reports pairwise differences in SMRD and BOE scores across subgroups. Urban communitarians generate value-increasing thoughts earlier in the decision process (SMRD = 0.45) and at a higher volume (BOE = 0.17) than rural individualists. These results help illustrate how cultural worldview influences not only valuation outcomes but also the cognitive process by which those values are formed.
Finally, Table 12 summarizes multiple regression models explaining the variation in non-GM premiums and GM discounts. The IC dimension has a significant negative effect on non-GM WTP (−2.020, p < 0.01) and a positive effect on GM discount values (1.139, p < 0.01), indicating that individualists are more accepting of GM food and less likely to pay a premium for non-GM labels. The interaction between individualism and rural residences is significant only in the GM model (0.551, p < 0.10), suggesting that cultural worldviews and geography jointly shape GM label responses. The QT variable’s BOE also significantly predicts both outcomes: positively associated with non-GM premiums (0.536) and negatively with GM discounts (−0.364), reinforcing the role of the cognitive balance in valuations. All VIF values are well below the common threshold of 10, confirming that multicollinearity is not a concern. These findings underscore the importance of cultural worldviews and thought processes in shaping consumer preferences. While some statistically significant differences are modest in absolute terms, they reflect consistent and interpretable patterns relevant to communication and labeling strategies.
6. Discussion
This research provides a comprehensive examination of preferences for GM and non-GM foods, emphasizing the cultural distinctions within rural, suburban, and urban communities. By integrating Cultural Cognition Theory (CCT) and Query Theory (QT), this study reveals how rural cultural worldviews—particularly individualism—shape consumer attitudes and willingness to pay (WTP) for labeled food attributes. Consistent with the prior literature [7,22], we find that rural respondents exhibit stronger individualistic worldviews, which are associated with a lower WTP for non-GM attributes and smaller discounts required to accept GM foods. These patterns suggest that individualism, rather than a rural residence alone, is a key explanatory factor in shaping responses to GM labeling. The significant effect of the IC dimension on both non-GM premiums and GM discounts reinforces the idea that personal responsibility and market-oriented beliefs reduce the aversion to biotechnology.
The QT results provide additional insight by illustrating the cognitive mechanisms behind these preferences. BOE scores reveal that rural individualists are more likely to generate value-decreasing thoughts during decision-making, such as concerns about health risks or skepticism toward labeling claims. In contrast, urban communitarians tend to prioritize value-increasing thoughts, such as support for local production or environmental benefits. These patterns suggest that identity-protective cognition influences not only what consumers value but also how they evaluate competing information during the choice process. This finding has important implications for public communication and food labeling strategies. Rather than adopting a one-size-fits-all approach, policymakers and producers should consider tailoring messages to resonate with different cultural groups. For rural individualists, emphasizing personal economic benefits—such as cost savings, support for local farmers, or voluntary market choices—may be more persuasive than messages focused on collective or regulatory outcomes. Conversely, for urban communitarians, communication strategies could highlight shared environmental goals, sustainability outcomes, or ethical food sourcing. The interaction between cultural worldviews and geography also reveals that certain messaging strategies may need to address overlapping values. For instance, a rural communitarian may still be sensitive to collective outcomes but approach biotechnology from a risk-averse position. These nuances underscore the need for segmentation in consumer outreach and food policy design.
Finally, while this study focuses on the U.S. context, the framework may be applicable in other countries where cultural worldviews vary across regions or social groups. Further research could explore whether similar patterns hold in Europe or Asia, particularly given differences in regulatory environments and consumer trust in science [3,4].
7. Conclusions
This study demonstrates that cultural worldviews—particularly individualism and communitarianism—play a significant role in shaping consumer preferences for GM and non-GM food labels, especially within rural populations. By integrating Cultural Cognition Theory with Query Theory, we show that not only do individualists tend to value GM products more favorably, but they also approach food-related decisions through a different cognitive lens, often emphasizing value-decreasing aspects.
These findings suggest that the rural consumer skepticism toward GM labeling is better explained by cultural identity than geographic location alone. The observed variations in the willingness to pay and cognitive evaluation patterns reinforce the importance of tailoring food policy and communication strategies to resonate with distinct cultural orientations. For rural individualists, framing biotechnology in terms of personal economic benefit and market autonomy may be more effective, while messaging for communitarian audiences may benefit from emphasizing collective welfare and sustainability.
One limitation of the current analysis is that no weighting adjustments were made to account for the sample size imbalance between rural and non-rural groups. Although our findings highlight significant differences in preferences and cognitive patterns by locale, these results should be interpreted with modest caution given the relatively smaller sample of rural respondents. Future research could extend this framework to other technologies or international contexts and explore longitudinal shifts in worldviews and trust as biotechnology becomes more prevalent in global food systems. Understanding the intersection of culture, cognition, and valuation remains critical for navigating the social acceptance of food innovations and advancing sustainability goals.
Funding
This research was funded by the Arkansas Soybean Promotion Board. No specific grant number is associated with this funding.
Institutional Review Board Statement
This study was conducted in accordance with and approved by the Institutional Review Board of the University of Arkansas (Protocol Code IRB15-10-192. and date of approval 19 October 2015).
Informed Consent Statement
Informed consent was obtained from all subjects involved in this study.
Data Availability Statement
The data presented in this study are available on request from the corresponding author. The data are not publicly available due to participant confidentiality, particularly related to the open-ended responses to the Query Theory portion of this study.
Acknowledgments
The authors thank the Arkansas Soybean Promotion Board for their support and the anonymous reviewers for their valuable comments.
Conflicts of Interest
The authors declare no conflicts of interest.
Appendix A
Table A1.
Willingness to pay space model results.
| Variables | Coefficient | Estimate | S.E. | 95% Confidence Interval | ||
|---|---|---|---|---|---|---|
| Random Parameters | ||||||
| NON-GM (NGE) | µ | 1.98 | *** | 0.30 | 1.40 | 2.56 |
| σ | 7.07 | *** | 0.29 | |||
| GM (GME) | µ | −1.87 | *** | 0.19 | −2.25 | −1.50 |
| σ | 1.21 | *** | 0.12 | |||
| LOWCO2 (LOE) | µ | 0.30 | ** | 0.14 | 0.03 | 0.58 |
| σ | 0.78 | *** | 0.28 | |||
| MEDIUMCO2 (MDE) | µ | −0.07 | 0.13 | −0.33 | 0.18 | |
| σ | 0.25 | 0.28 | ||||
| HIGHCO2 (HIE) | µ | −0.06 | 0.13 | −0.31 | 0.20 | |
| σ | 0.04 | 0.51 | ||||
| LOCAL (LCE) | µ | 0.45 | *** | 0.10 | 0.27 | 0.64 |
| σ | 0.68 | *** | 0.26 | |||
| Nonrandom Parameters | ||||||
| No-Buy (NONE) | µ | −3.54 | *** | 0.08 | ||
| Price (in Preference Space) | µ | 0.73 | *** | 0.0352 | ||
| Model Fit | ||||||
| Respondents | 932 | |||||
| Log Likelihood | −5586.3 | |||||
| AIC | 11,230.6 | |||||
| AIC/N | 1.506 | |||||
*,**, and *** indicate significance at the 10%, 5%, and 1% levels, respectively.
References
- Lusk, J.L.; Roosen, J.; Bieberstein, A. Consumer Acceptance of New Food Technologies: Causes and Roots of Controversies. Annu. Rev. Resour. Econ. 2014, 6, 381–405. [Google Scholar] [CrossRef]
- Kahan, D.M. Cultural Cognition as a Conception of the Cultural Theory of Risk. In Handbook of Risk Theory: Epistemology, Decision Theory, Ethics and Social Implications of Risk; Hillerbrand, R., Ed.; Springer: Dordrecht, The Netherlands, 2012; pp. 725–759. [Google Scholar]
- Meerza, S.; Dsouza, A.; Ahamed, A.; Mottaleb, K. Risk Propensity and Acceptance of Gene-Edited and Genetically Modified Food among US Consumers: A Comparison Between Plants and Animal Products. J. Agric. Appl. Econ. 2024, 56, 575–596. [Google Scholar] [CrossRef]
- Wang, Y.; Yue, C.; Watkins, E.; Straw, C. How Do Risk Preferences Affect Golf Course Superintendents’ Adoption of Precision Irrigation Technologies? Implications from Prospect Theory. J. Agric. Appl. Econ. 2023, 55, 516–535. [Google Scholar] [CrossRef]
- Seo, F.; Hudson, D. Attributes That Influence Consumers’ Preferences for Choosing Locally Grown Food Sources During and After the COVID-19 Pandemic. J. Agric. Appl. Econ. 2023, 55, 626–650. [Google Scholar] [CrossRef]
- Hovhannisyan, V.; Kondaridze, M.; Bastian, C.; Shanoyan, A. Empirical Evidence of Changing Food Demand and Consumer Preferences in Russia. J. Agric. Appl. Econ. 2020, 52, 480–501. [Google Scholar] [CrossRef]
- Kemper, N.P.; Popp, J.S.; Nayga, R.M.; Kerr, J.B. Cultural Worldview and Genetically Modified Food Policy Preferences. Food Policy 2018, 80, 68–83. [Google Scholar] [CrossRef]
- Twine, E.; Ndindeng, S.; Mujawamariya, G.; Futakuchi, K. Pricing Rice Quality Attributes and Returns to Quality Upgrading in Sub-Saharan Africa. J. Agric. Appl. Econ. 2022, 54, 175–196. [Google Scholar] [CrossRef]
- Kahan, D.M.; Braman, D. Cultural Cognition and Public Policy. Yale Law Policy Rev. 2006, 24, 147–170. [Google Scholar]
- Douglas, M.; Wildavsky, A.B. Risk and Culture: An Essay on the Selection of Technical and Environmental Dangers; University of California Press: Berkeley, CA, USA, 1982. [Google Scholar]
- Slovic, P. The Perception of Risk; Earthscan: London, UK; Sterling, VA, USA, 2000. [Google Scholar]
- Johnson, E.J.; Häubl, G.; Keinan, A. Aspects of Endowment: A Query Theory of Value Construction. J. Exp. Psychol. Learn. Mem. Cogn. 2007, 33, 461–474. [Google Scholar] [CrossRef]
- Weber, E.U.; Johnson, E.J.; Milch, K.F.; Chang, H.; Brodscholl, J.C.; Goldstein, D.G. Asymmetric Discounting in Intertemporal Choice: A Query-Theory Account. Psychol. Sci. 2007, 18, 516–523. [Google Scholar] [CrossRef]
- Survey Sampling International (SSI). Available online: https://www.surveysampling.com (accessed on 7 November 2016).
- United States Department of Agriculture (USDA) Economic Research Service (ERS). Retail Prices for Beef, Pork, and Poultry Cuts, Eggs, and Dairy Products. Available online: https://www.ers.usda.gov/data-products/meat-price-spreads (accessed on 11 October 2016).
- Van Loo, E.J.; Caputo, V.; Nayga, R.M., Jr.; Verbeke, W. Consumers’ valuation of sustainability labels on meat. Food Policy 2014, 49, 137–150. [Google Scholar] [CrossRef]
- Bliemer, M.C.; Rose, J.M. Construction of Experimental Designs for Mixed Logit Models Allowing for Correlation Across Choice Observations. Transp. Res. Part B Methodol. 2010, 44, 720–734. [Google Scholar] [CrossRef]
- Kahan, D.M.; Jenkins-Smith, H.; Braman, D. Cultural Cognition of Scientific Consensus. J. Risk Res. 2011, 14, 147–174. [Google Scholar] [CrossRef]
- Kemper, N.P.; Popp, J.S.; Nayga, R.M. A Query Theory Account of a Discrete Choice Experiment under Oath. Eur. Rev. Agric. Econ. 2020, 47, 1133–1172. [Google Scholar] [CrossRef]
- Kemper, N.P.; Popp, J.S.; Nayga, R.M.; Kerr, J.B. A Query Approach to Modeling Attendance to Attributes in Discrete Choice Experiments. J. Agric. Resour. Econ. 2024, 49, 392–409. [Google Scholar]
- McFadden, D. Conditional Logit Analysis of Qualitative Choice Behavior. In Frontiers in Econometrics; Zarembka, P., Ed.; Academic Press: New York, NY, USA, 1974; pp. 105–142. [Google Scholar]
- Lancaster, K.J. A New Approach to Consumer Theory. J. Political Econ. 1966, 74, 132–157. [Google Scholar] [CrossRef]
- Scarpa, R.; Thiene, M.; Train, K. Utility in Willingness to Pay Space: A Tool to Address Confounding Random Scale Effects in Destination Choice to the Alps. Am. J. Agric. Econ. 2008, 90, 994–1010. [Google Scholar] [CrossRef]
- Revelt, D.; Train, K. Mixed Logit with Repeated Choices: Households’ Choices of Appliance Efficiency Level. Rev. Econ. Stat. 1998, 80, 647–657. [Google Scholar] [CrossRef]
- Train, K. Discrete Choice Methods with Simulation; Cambridge University Press: Cambridge, UK, 2009. [Google Scholar]
- Lusk, J.L.; Coble, K.H. Risk Perceptions, Risk Preference, and Acceptance of Risky Food. Am. J. Agric. Econ. 2005, 87, 393–405. [Google Scholar] [CrossRef]
- McFadden, B.R.; Lusk, J.L. Cognitive Biases in the Assimilation of Scientific Information on Global Warming and Genetically Modified Food. Food Policy 2015, 54, 35–43. [Google Scholar] [CrossRef]
- Dannenberg, A.; Scatasta, S.; Sturm, B. Does Mandatory Labeling of Genetically Modified Food Grant Consumers the Right to Know? Evidence from an Economic Experiment; ZEF–Discussion Papers on Development Policy, No. 123; Zentrum für Europäische Wirtschaftsforschung (ZEW): Mannheim, Germany, 2008. [Google Scholar]
- Dannenberg, A.; Scatasta, S.; Sturm, B. Mandatory versus Voluntary Labelling of Genetically Modified Food: Evidence from an Economic Experiment. Agric. Econ. 2011, 42, 373–386. [Google Scholar] [CrossRef]
Figure 1.
Conceptual Framework: Linking locale, cultural worldview, cognition, and consumer preferences. Note: Solid arrows indicate hypothesized causal pathways. The dotted arrow represents a potential direct effect of the geographic locale on the willingness to pay, independent of worldviews or cognitive processing.
Figure 1.
Conceptual Framework: Linking locale, cultural worldview, cognition, and consumer preferences. Note: Solid arrows indicate hypothesized causal pathways. The dotted arrow represents a potential direct effect of the geographic locale on the willingness to pay, independent of worldviews or cognitive processing.
Table 2.
Cultural cognition short-form scales and group–grid dimensions.
| Variable Name | Variable Type | Description | |
|---|---|---|---|
| Group or Individualism–Communitarianism 1 | |||
| CHARM (reverse code) | 1 = strongly agree to 5 = strongly disagree | Sometimes the government needs to make laws that keep people from hurting themselves. | |
| IPRIVACY | 1 = strongly disagree to 5 = strongly agree | The government should stop telling people how to live their lives. | |
| Grid or Hierarchy–Egalitarianism 2 | |||
| HEQUAL | 1 = strongly disagree to 5 = strongly agree | We have gone too far in pushing equal rights in this country. | |
| EWEALTH (reverse code) | 1 = strongly agree to 5 = strongly disagree | Our society would be better off if the distribution of wealth was more equal. | |
| Cultural Worldview Group–Grid Dimensions | Scale Interpretation | Frequencies | |
| n | percent | ||
| Individualism | Negative Group Factor Score | 409 | 43.9% |
| Communitarianism | Positive Group Factor Score | 523 | 56.1% |
| Egalitarian | Negative Grid Factor Score | 509 | 54.6% |
| Hierarchical | Positive Grid Factor Score | 423 | 45.4% |
| Hierarchical–Individualism (HI) | Positive Grid × Positive Group | 297 | 31.9% |
| Egalitarian–Individualism (EI) | Negative Grid × Positive Group | 226 | 24.2% |
| Hierarchical–Communitarianism (HC) | Positive Grid × Negative Group | 212 | 22.7% |
| Egalitarian–Communitarianism (EC) | Negative Grid × Negative Group | 197 | 21.1% |
Note: Respondents were not forced to agree or disagree with the above statements. A factor analysis was used to examine responses and place all subjects into corresponding dimensions and worldviews. 1 The Individualism–Communitarianism scale reflected acceptable levels of measurement precision, with a Cronbach’s α = 0.7792; 2 The Hierarchy–Egalitarianism scale reflected acceptable levels of measurement precision with a Cronbach’s α = 0.710.
Table 3.
Sample characteristics, by group–grid dimension.
| Sample | Communitarian | Individualism | Chi-Square Test | Hierarchical | Egalitarianism | Chi-Square Test | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| rural | Count (%) | 221 | (24%) | 112 | (21%) | 109 | (27%) | χ2: 5.195 | 98 | (23%) | 123 | (24%) | χ2: 0.2771 |
| suburban | Count (%) | 476 | (51%) | 267 | (51%) | 209 | (51%) | p-value: 0.074 | 220 | (52%) | 256 | (50%) | p-value: 0.871 |
| urban | Count (%) | 235 | (25%) | 144 | (28%) | 91 | (22%) | 105 | (25%) | 130 | (26%) | ||
| female | Count (%) | 488 | (52%) | 348 | (67%) | 140 | (34%) | χ2: 0.061 | 147 | (35%) | 168 | (33%) | χ2: 0.315 |
| male | Count (%) | 444 | (48%) | 175 | (33%) | 269 | (66%) | p-value: 0.805 | 276 | (65%) | 341 | (67%) | p-value: 0.575 |
| 18 to 24 years | Count (%) | 70 | (8%) | 47 | (9%) | 23 | (6%) | 26 | (6%) | 44 | (9%) | ||
| 25 to 34 years | Count (%) | 194 | (21%) | 122 | (23%) | 72 | (18%) | 83 | (20%) | 111 | (22%) | ||
| 35 to 44 years | Count (%) | 155 | (17%) | 90 | (17%) | 65 | (16%) | χ2: 15.929 | 64 | (15%) | 91 | (18%) | χ2: 7.743 |
| 45 to 54 years | Count (%) | 170 | (18%) | 79 | (15%) | 91 | (22%) | p-value: 0.007 | 76 | (18%) | 94 | (18%) | p-value: 0.171 |
| 55 to 64 years | Count (%) | 186 | (20%) | 106 | (20%) | 80 | (20%) | 92 | (22%) | 94 | (18%) | ||
| 65 years and over | Count (%) | 157 | (17%) | 79 | (15%) | 78 | (19%) | 82 | (19%) | 75 | (15%) | ||
| children | Count (%) | 196 | (21%) | 100 | (19%) | 96 | (23%) | χ2: 2.617 | 82 | (19%) | 395 | (78%) | χ2: 1.262 |
| no children | Count (%) | 736 | (79%) | 423 | (81%) | 313 | (77%) | p-value: 0.106 | 341 | (81%) | 114 | (22%) | p-value: 0.261 |
| some high school | Count (%) | 19 | (2%) | 11 | (2%) | 8 | (2%) | 10 | (2%) | 9 | (2%) | ||
| high school diploma | Count (%) | 302 | (32%) | 160 | (31%) | 142 | (35%) | 135 | (32%) | 167 | (33%) | ||
| associate′s degree | Count (%) | 199 | (21%) | 108 | (21%) | 91 | (22%) | χ2: 5.430 | 96 | (23%) | 103 | (20%) | χ2: 2.149 |
| bachelor′s degree | Count (%) | 270 | (29%) | 153 | (29%) | 117 | (29%) | p-value: 0.366 | 115 | (27%) | 155 | (30%) | p-value: 0.828 |
| master′s degree | Count (%) | 108 | (12%) | 68 | (13%) | 40 | (10%) | 51 | (12%) | 57 | (11%) | ||
| doctoral degree | Count (%) | 34 | (4%) | 23 | (4%) | 11 | (3%) | 16 | (4%) | 18 | (4%) | ||
| not white | Count (%) | 142 | (15%) | 99 | (19%) | 43 | (11%) | χ2: 12.587 | 55 | (13%) | 87 | (17%) | χ2: 2.992 |
| white | Count (%) | 790 | (85%) | 424 | (81%) | 366 | (89%) | p-value: 0.000 | 368 | (87%) | 422 | (83%) | p-value: 0.084 |
| under USD 20,000 | Count (%) | 110 | (12%) | 58 | (11%) | 52 | (13%) | 53 | (13%) | 57 | (11%) | ||
| 20,000–39,999 | Count (%) | 220 | (24%) | 112 | (21%) | 108 | (26%) | 85 | (20%) | 135 | (27%) | ||
| 40,000–59,999 | Count (%) | 184 | (20%) | 101 | (19%) | 83 | (20%) | 79 | (19%) | 105 | (21%) | ||
| 60,000–79,999 | Count (%) | 160 | (17%) | 94 | (18%) | 66 | (16%) | χ2: 13.212 | 69 | (16%) | 91 | (18%) | χ2: 28.599 |
| 80,000–99,999 | Count (%) | 105 | (11%) | 54 | (10%) | 51 | (12%) | p-value: 0.105 | 55 | (13%) | 50 | (10%) | p-value: 0.000 |
| 100,000–119,999 | Count (%) | 59 | (6%) | 40 | (8%) | 19 | (5%) | 26 | (6%) | 33 | (6%) | ||
| 120,000–139,999 | Count (%) | 30 | (3%) | 20 | (4%) | 10 | (2%) | 19 | (4%) | 11 | (2%) | ||
| 140,000–159,999 | Count (%) | 26 | (3%) | 18 | (3%) | 8 | (2%) | 14 | (3%) | 12 | (2%) | ||
| USD 160,000 and over | Count (%) | 38 | (4%) | 26 | (5%) | 12 | (3%) | 23 | (5%) | 15 | (3%) | ||
| social strong liberal | Count (%) | 86 | (9%) | 57 | (11%) | 29 | (7%) | 15 | (4%) | 71 | (14%) | ||
| social liberal | Count (%) | 196 | (21%) | 130 | (25%) | 66 | (16%) | 63 | (15%) | 133 | (26%) | ||
| social weak liberal | Count (%) | 329 | (35%) | 189 | (36%) | 140 | (34%) | 148 | (35%) | 181 | (36%) | ||
| social weak conservative | Count (%) | 189 | (20%) | 86 | (16%) | 103 | (25%) | χ2: 36.713 | 114 | (27%) | 75 | (15%) | χ2: 83.666 |
| social conservative | Count (%) | 72 | (8%) | 24 | (5%) | 48 | (12%) | p-value: 0.000 | 54 | (13%) | 18 | (4%) | p-value: 0.000 |
| social strong conservative | Count (%) | 60 | (6%) | 37 | (7%) | 23 | (6%) | 29 | (7%) | 31 | (6%) | ||
| fiscal strong liberal | Count (%) | 61 | (7%) | 42 | (8%) | 19 | (5%) | 12 | (3%) | 49 | (10%) | ||
| fiscal liberal | Count (%) | 133 | (14%) | 88 | (17%) | 45 | (11%) | 45 | (11%) | 88 | (17%) | ||
| fiscal weak liberal | Count (%) | 357 | (38%) | 220 | (42%) | 137 | (33%) | 141 | (33%) | 216 | (42%) | ||
| fiscal weak conservative | Count (%) | 224 | (24%) | 106 | (20%) | 118 | (29%) | χ2: 50.941 | 127 | (30%) | 97 | (19%) | χ2: 67.669 |
| fiscal conservative | Count (%) | 94 | (10%) | 27 | (5%) | 67 | (16%) | p-value: 0.000 | 68 | (16%) | 26 | (5%) | p-value: 0.000 |
| fiscal strong conservative | Count (%) | 63 | (7%) | 40 | (8%) | 23 | (6%) | 30 | (7%) | 33 | (6%) | ||
Table 4.
Summary of factor analysis results for group–grid dimensions, by locale.
| n | Sample | Rural | Suburban | Urban | Statistics | ||
|---|---|---|---|---|---|---|---|
| 932 | 221 | 476 | 235 | ||||
| IC Dimension | mean | −0.11 | 0.01 | −0.12 | −0.21 | F: 2.790 | p-value: 0.062 |
| HE Dimension | mean | 0.05 | 0.05 | 0.07 | 0.01 | F: 0.260 | p-value: 0.774 |
| Individualism | count | 409 | 109 | 209 | 91 | χ2: 5.1947 | p-value: 0.074 |
| Communitarianism | count | 523 | 112 | 267 | 144 | ||
| Hierarchical | count | 423 | 98 | 220 | 105 | χ2: 0.2271 | p-value: 0.871 |
| Egalitarianism | count | 509 | 123 | 256 | 130 | ||
Note: Standardized factor values range from −1 to 1. Statistics based on one-way ANOVA and χ2.
Table 5.
Mean willingness to pay values, by group–grid dimensions.
| Non-GMOWTP | GMWTP | |||||
|---|---|---|---|---|---|---|
| n | Mean | Std. Dev. | Mean | Std. Dev. | ||
| Sample | 932 | 2.05 | 6.24 | −1.91 | 3.65 | |
| Group Dimensions | Individualism | 409 | −0.94 | 5.20 | −0.21 | 3.13 |
| Communitarian | 523 | 4.38 | 5.99 | −3.24 | 3.47 | |
| F statistic: | 203.35 | 190.79 | ||||
| p-value: | 0.000 | 0.000 | ||||
| Grid Dimensions | Hierarchical | 423 | 1.66 | 5.97 | −1.70 | 3.47 |
| Egalitarian | 509 | 2.37 | 6.45 | −2.08 | 3.78 | |
| F statistic: | 2.97 | 2.42 | ||||
| p-value: | 0.085 | 0.120 | ||||
Note: Statistics based on one-way ANOVA. The values listed for WTP for the GM label are in absolute dollars. Consistent with findings in the literature, the GM label is associated with negative preferences, resulting in negative WTP values. The WTP values for the GM label should be interpreted as a measure of the discount required for consumers to purchase the product containing GM ingredients. WTP space model details are provided in Appendix A Table A1.
Table 6.
Mean willingness to pay values, by locale.
| Non-GMOWTP | GMWTP | ||||
|---|---|---|---|---|---|
| Locale | n | Mean | Std. Dev. | Mean | Std. Dev. |
| Sample | 932 | 2.05 | 6.24 | −1.91 | 3.65 |
| Rural | 221 | 1.48 | 5.54 | −1.53 | 3.28 |
| Suburban | 476 | 1.83 | 6.34 | −1.83 | 3.68 |
| Urban | 235 | 3.02 | 6.57 | −2.42 | 3.86 |
| F statistic: | 4.11 | 3.68 | |||
| p-value: | 0.017 | 0.026 | |||
Note: Statistics based on one-way ANOVA. The values listed for WTP for the GM label are in absolute dollars. Consistent with findings in the literature, the GM label is associated with negative preferences, resulting in negative WTP values. The WTP values for the GM label should be interpreted as a measure of the discount required for consumers to purchase the product containing GM ingredients. WTP space model details are provided in Appendix A Table A1.
Table 7.
Mean willingness to pay values, by locale and group–grid dimensions.
| Locale | Group/Grid Dimension | n | Non-GMOWTP | GMWTP | ||
|---|---|---|---|---|---|---|
| Mean | Std. Dev. | Mean | Std. Dev. | |||
| Rural | Individualism | 66 | 3.80 | 5.46 | −2.89 | 3.23 |
| Communitarian | 46 | 3.69 | 4.63 | −2.79 | 2.73 | |
| Egalitarian | 57 | −0.50 | 4.63 | −0.29 | 2.71 | |
| Hierarchical | 52 | −1.26 | 5.33 | −0.04 | 3.25 | |
| Suburban | Individualism | 151 | 4.49 | 6.54 | −3.31 | 3.77 |
| Communitarian | 116 | 4.06 | 5.71 | −3.10 | 3.25 | |
| Egalitarian | 105 | −1.10 | 5.18 | −0.18 | 3.13 | |
| Hierarchical | 104 | −1.59 | 4.82 | 0.08 | 2.92 | |
| Urban | Individualism | 80 | 4.66 | 6.59 | −3.40 | 3.85 |
| Communitarian | 64 | 5.45 | 5.78 | −3.79 | 3.33 | |
| Egalitarian | 50 | 0.91 | 6.83 | −1.17 | 4.10 | |
| Hierarchical | 41 | −1.39 | 4.13 | 0.09 | 2.57 | |
| F statistic: | 35.33 | 29.65 | ||||
| p-value: | 0.000 | 0.002 | ||||
Note: Statistics based on one-way ANOVA. The values listed for WTP for the GM label are in absolute dollars. Consistent with findings in the literature, the GM label is associated with negative preferences, resulting in negative WTP values. The WTP values for the GM label should be interpreted as a measure of the discount required for consumers to purchase the product containing GM ingredients. WTP space model details are provided in Appendix A Table A1.
Table 8.
Pairwise mean comparisons, non-GMO premium and GM discount, by locale and group–grid dimensions.
Table 8.
Pairwise mean comparisons, non-GMO premium and GM discount, by locale and group–grid dimensions.
 |
Note: Values in the table represent pairwise comparisons of differences between each subgroup. The lower left portion of the table is the comparison of the non-GMO WTP premiums, while the upper right portion is a comparison of GM WTP discounts. A single solid outline and regular text represents a statistical significance at the 0.10 level, a double solid outline and italicized text at the 0.05 level, and a bolded dashed outline and bolded text at the 0.01 level. Statistics are based on a t-test of independent samples.
Table 9.
Mean SMRD and BOE, by locale.
| SMRD | BOE | ||||
|---|---|---|---|---|---|
| n | Mean | S.D. | Mean | S.D. | |
| Sample | 932 | −0.005 | 0.66 | −0.071 | 1.00 |
| Rural | 221 | −0.032 | 0.65 | −0.172 | 0.99 |
| Suburban | 476 | 0.004 | 0.66 | −0.067 | 0.99 |
| Urban | 235 | 0.001 | 0.68 | 0.017 | 1.03 |
| F statistic: | 0.23 | 2.04 | |||
| p-value: | 0.796 | 0.131 | |||
Note: Statistics based on one-way ANOVA. SMRD is valued on a scale from −1 (all decreasing aspects) to +1 (all increasing aspects). Balance of Evidence (BOE) is the difference between value-increasing and -decreasing aspects, on a per choice task basis.
Table 10.
Mean SMRD and BOE, by locale and group–grid dimensions.
| SMRD | BOE | |||||
| n | Mean | Std. Dev. | Mean | Std. Dev. | ||
| Rural | Individualism | 66 | −0.01 | 0.67 | −0.14 | 1.01 |
| Communitarian | 43 | 0.06 | 0.70 | −0.02 | 0.98 | |
| Egalitarian | 56 | −0.11 | 0.63 | −0.30 | 0.98 | |
| Hierarchical | 50 | −0.06 | 0.59 | −0.21 | 1.00 | |
| Suburban | Individualism | 147 | 0.01 | 0.63 | 0.08 | 1.00 |
| Communitarian | 110 | 0.01 | 0.63 | −0.16 | 0.94 | |
| Egalitarian | 103 | 0.04 | 0.69 | −0.10 | 1.01 | |
| Hierarchical | 104 | −0.05 | 0.71 | −0.13 | 1.01 | |
| Urban | Individualism | 76 | 0.09 | 0.61 | 0.15 | 1.02 |
| Communitarian | 63 | 0.10 | 0.71 | 0.17 | 0.97 | |
| Egalitarian | 49 | −0.09 | 0.74 | −0.12 | 1.00 | |
| Hierarchical | 41 | −0.20 | 0.64 | −0.32 | 1.08 | |
| F statistic: | 0.23 | 1.79 | ||||
| p-value: | 0.796 | 0.051 | ||||
Note: Statistics based on one-way ANOVA. SMRD is valued on a scale from −1 (all decreasing aspects) to +1 (all increasing aspects). Balance of Evidence (BOE) is the difference between value-increasing and -decreasing aspects, on a per choice task basis.
Table 11.
Pairwise mean comparisons, SMRD and BOE, by locale and group–grid dimensions.
 |
Note: Values in the table represent pairwise comparisons of differences between each subgroup. The lower left portion of the table is the comparison of SMRD values, while the upper right portion is the comparison of the BOE. A single solid outline and regular text represents a statistical significance at the 0.10 level, a double solid outline and italicized text at the 0.05 level, and a bolded dashed outline and bolded text at the 0.01 level. Statistics are based on a t-test of independent samples.
Table 12.
Multiple regression results for non-GMO premiums and GM discounts.
| Non-GMO Premium | GM Discount | VIF | ||||||
|---|---|---|---|---|---|---|---|---|
| Category | Level | Parameter | Robust SE | Parameter | Robust SE | |||
| Cultural Dimensions and Interaction | Individualism–Communitarian | −2.020 | *** | 0.184 | 1.139 | *** | 0.107 | 1.27 |
| Hierarchical–Egalitarian | −0.492 | *** | 0.174 | 0.248 | ** | 0.102 | 1.13 | |
| IC × HE Interaction | 0.161 | 0.153 | −0.097 | 0.089 | 1.08 | |||
| Demographics | Rural | −0.922 | * | 0.481 | 0.551 | * | 0.284 | 1.59 |
| Suburban | −0.974 | ** | 0.437 | 0.478 | * | 0.257 | 1.54 | |
| Age | −0.340 | *** | 0.125 | 0.184 | ** | 0.073 | 1.33 | |
| Child | 0.194 | 0.411 | −0.151 | 0.243 | 1.12 | |||
| Income | 0.370 | *** | 0.085 | −0.227 | *** | 0.050 | 1.12 | |
| Non-white | 1.311 | ** | 0.528 | −0.736 | ** | 0.310 | 1.08 | |
| Food Labeling and Risk Preferences | mand | 1.060 | *** | 0.190 | −0.638 | *** | 0.111 | 1.48 |
| vol | 0.069 | 0.176 | −0.027 | 0.102 | 1.34 | |||
| frisk | −0.163 | ** | 0.074 | 0.087 | ** | 0.043 | 1.30 | |
| read_labels | 1.464 | *** | 0.188 | −0.837 | *** | 0.110 | 1.21 | |
| eatgm | −0.399 | ** | 0.186 | 0.201 | * | 0.109 | 1.18 | |
| uinfo | −0.770 | *** | 0.217 | 0.380 | *** | 0.125 | 1.35 | |
| pinfo | 0.474 | ** | 0.208 | −0.243 | ** | 0.121 | 1.16 | |
| ginfo | −0.586 | *** | 0.194 | 0.384 | *** | 0.113 | 1.38 | |
| loc_frt | 0.870 | ** | 0.392 | −0.579 | ** | 0.229 | 1.34 | |
| loc_wrn | 1.930 | *** | 0.527 | −1.164 | *** | 0.306 | 1.45 | |
| Query Theory Measures | smrdi | 0.534 | * | 0.289 | −0.328 | * | 0.172 | 1.36 |
| boe | 0.536 | *** | 0.201 | −0.364 | *** | 0.119 | 1.43 | |
| constant | −3.155 | * | 1.683 | 1.421 | 0.975 | n/a | ||
| Model Fit Statistics | R2 | 0.376 | 0.369 | Mean VIF: 1.30 | ||||
| F (21, 886) | 33.77 | 32.15 | ||||||
| No. respondents | 908 | 908 | ||||||
*,**,and *** indicate significance at the 10%, 5%, and 1% levels, respectively. Note: The variance inflation factor (VIF) measures how much the variance of a regression coefficient is inflated due to multicollinearity among the predictors. The low VIF values indicate that our predictor variables are not highly correlated, and multicollinearity is not a concern in our models.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2025 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
MDPI and ACS Style
Kemper, N.P. Cultural Worldview and Rural Consumer Preferences for Genetically Modified Foods. Sustainability 2025, 17, 6843. https://doi.org/10.3390/su17156843
AMA Style
Kemper NP. Cultural Worldview and Rural Consumer Preferences for Genetically Modified Foods. Sustainability. 2025; 17(15):6843. https://doi.org/10.3390/su17156843
Chicago/Turabian StyleKemper, Nathan P. 2025. "Cultural Worldview and Rural Consumer Preferences for Genetically Modified Foods" Sustainability 17, no. 15: 6843. https://doi.org/10.3390/su17156843
APA StyleKemper, N. P. (2025). Cultural Worldview and Rural Consumer Preferences for Genetically Modified Foods. Sustainability, 17(15), 6843. https://doi.org/10.3390/su17156843
Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.
