Determinants of out-of-home food consumption: development and validation of a multidimensional scale in a Brazilian sample Authors: Luis Felipe Dias Lopes, Daniela Pegoraro, Rayssa Cleide de Oliveira, Eduarda Grando Lopes, Rosangela Arruda Saragozo, Deoclécio Junior Cardoso da Silva Published: 2026 ABSTRACT BackgroundOut-of-home food consumption has become an integral component of contemporary dietary patterns, particularly in urban contexts. Evidence suggests that eating outside the home is associated with diet quality, nutritional intake, and broader public health outcomes. However, Brazil lacks validated multidimensional instruments capable of comprehensively assessing the economic, behavioral, and experiential determinants underlying this behavior. This study aimed to develop and validate the Scale of Determinant Factors in Out-of-Home Food Consumption (SDF-OHFC) and to examine the structural relationships among its dimensions.MethodsA cross-sectional study was conducted with 426 Brazilian adults who reported consuming food outside the home at least once per week. Data were collected through an online survey and analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM). The measurement model was assessed for reliability, convergent and discriminant validity. Structural relationships were tested using bootstrapping procedures (5,000 resamples), and predictive performance was evaluated through R², Q², RMSE, and MAE. Measurement invariance was examined across sex, marital status, and geographic regions.ResultsThe SDF-OHFC comprises five dimensions: economic aspects, convenience, habits and personal preferences, social and experiential aspects, and perceived food quality. The scale demonstrated satisfactory internal consistency (Cronbach’s α = 0.705–0.762), composite reliability (> 0.70), and convergent validity (AVE > 0.50). All hypothesized structural relationships were statistically significant (p < 0.05). Economic aspects played a central antecedent role, positively influencing convenience and habits, and negatively influencing social and experiential aspects. Convenience and social and experiential dimensions were positively associated with perceived food quality. Structural invariance was confirmed across all tested demographic groups.ConclusionsThe SDF-OHFC is a valid and reliable multidimensional instrument for assessing determinants of out-of-home food consumption in Brazil. By integrating economic, behavioral, and experiential factors within a single framework, the scale supports public health research and policy development aimed at improving urban food environments, promoting healthier eating practices, and addressing structural determinants of dietary behavior. Introduction Over recent decades, the food sector has undergone substantial transformations driven by changes in production, processing, and distribution systems, alongside urbanization and shifting lifestyles [1–4]. These transformations have contributed to the expansion of out-of-home food consumption, which has progressively shifted from an occasional practice to a routine component of everyday life [5–7]. Evidence suggests that eating outside the home is particularly prevalent among young adults, men, and individuals with higher educational attainment, reflecting evolving work arrangements, family dynamics, urban lifestyles, and increasing time scarcity [8–12]. The growing reliance on ready-to-eat meals, restaurants, and food delivery services has intensified concerns regarding both public health and sustainable food systems. Out-of-home food consumption has frequently been associated with lower diet quality, including higher intake of energy, fats, and sodium, and lower consumption of fruits, vegetables, and fiber [13–16]. Simultaneously, these consumption patterns intersect with broader sustainability challenges related to food diversity, environmental impacts, and urban food environments [17–19]. In Brazil, these transformations have accompanied rapid urbanization and changes in work organization, reinforcing the externalization of food preparation through an expanding supply of ready meals and digitally mediated food services [20, 21]. Classic and contemporary literature suggests that out-of-home eating behavior is shaped by interrelated determinants, including convenience [5, 11, 22], perceived food quality [19, 23, 24], economic aspects [25–28], social and experiential meanings attached to meals [29–32], and socially constructed habits and preferences [20, 33]. However, these determinants have been progressively reinterpreted in light of contemporary food environments and urban lifestyles, reinforcing the need for integrative models capable of capturing their joint effects on food choices [34, 35]. From a structural perspective, economic aspects are expected to play an antecedent role, since price sensitivity, perceived value, and budget constraints shape consumers’ access to and selection of out-of-home food options [25, 26]. Evidence also indicates that individuals with higher income and educational levels tend to engage more frequently in out-of-home food consumption and food delivery services, whereas lower-income groups experience stronger trade-offs between affordability and convenience [14, 28]. Accordingly, economic conditions are expected to influence convenience-related decisions, leading to H1: Economic aspects positively influence convenience. Economic conditions may also affect eating routines and preference stability by constraining or enabling repeated consumption patterns, supporting H2: Economic aspects positively influence habits and personal preferences [20, 33]. Likewise, affordability and purchasing power may influence the extent to which eating out is experienced as a social and experiential practice, supporting H3: Economic aspects positively influence social and experiential aspects [29, 30]. Habits and personal preferences are also expected to influence both convenience-oriented behaviors and the social meanings attached to eating out. Food habits are socially constructed and reinforced through routine and repeated exposure, shaping where, how, and with whom individuals eat [20, 33]. Therefore, H4 proposes that habits and personal preferences positively influence social and experiential aspects, whereas H5 proposes that habits and personal preferences positively influence convenience [5, 11]. Perceived food quality in out-of-home settings is often constructed through contextual and experiential cues rather than exclusively through nutritional attributes [23, 24]. Dining experiences involving service quality, atmosphere, and social interaction may shape perceptions of quality, supporting H6: Social and experiential aspects positively influence perceived food quality [29, 30]. In addition, convenience associated with delivery services, time-saving practices, and ready-to-eat meals may also influence quality perceptions, supporting H7: Convenience positively influences perceived food quality [5, 11, 36]. Despite advances in the international literature, Brazil still lacks validated instruments capable of measuring these determinants in an integrated manner. Existing initiatives, such as the Frequency Consumption Scale (FCS) validated in the VAMOS program [37] and adapted instruments such as the Children’s Eating Behaviour Questionnaire (CEBQ) [38], demonstrate the importance of robust psychometric approaches but do not comprehensively capture the multidimensional sociocultural and economic determinants of out-of-home food consumption. Accordingly, this study aims to develop and validate the Scale of Determinant Factors in Out-of-Home Food Consumption (SDF-OHFC) and to test the structural relationships among economic aspects, habits and personal preferences, convenience, social and experiential aspects, and perceived food quality. Additionally, we examine structural invariance across gender, marital status, and Brazilian regions (H8–H10). Figure 1 presents the conceptual model and hypotheses tested. Fig. 1Habits and Personal Preferences in out-of-home food consumptions Habits and Personal Preferences in out-of-home food consumptions The scale was developed to assess the determinants of out-of-home food consumption among Brazilian adult consumers who regularly eat outside the home, encompassing diverse urban food service contexts such as restaurants, pizzerias, fast-food outlets, and similar establishments. This scope ensures that the instrument captures consumption behaviors across a wide range of everyday eating-out situations. Method Study design and participants This study focuses on the development and validation of a multidimensional scale designed to capture the determinants of out-of-home food consumption. A quantitative, cross-sectional design was adopted. Data were collected from a non-probabilistic sample of Brazilian consumers who reported consuming food outside the home at least once per week. The instrument was specifically designed for adult consumers engaged in regular out-of-home food consumption across different types of food service establishments (e.g., self-service restaurants, à la carte restaurants, fast-food outlets, pizzerias, and ethnic restaurants). Participants were recruited through an online survey disseminated via social media platforms and messaging applications between March/2024 and January/2025. Eligibility criteria included being 18 years of age or older, residing in Brazil, and reporting regular out-of-home food consumption. A total of 426 valid responses were obtained and retained for analysis after data screening. The sample comprised respondents from all five Brazilian geographic regions, allowing for subsequent invariance testing. Sociodemographic information collected included sex, age, marital status, educational level, household income, and region of residence. Age was collected in predefined categorical ranges (≤ 30; 31–45; 46–55; > 55), designed to represent distinct life stages while maintaining analytical balance across groups. Educational level was categorized into five levels (high school, undergraduate degree, specialization/MBA, and master’s degree and doctorate) to capture increasing levels of formal education. Household income was collected in categories based on multiples of the Brazilian minimum wage (up to 2, 2–3, 3–4, and above 4 minimum wages) and subsequently converted into U.S. dollars for international comparability. Participation was voluntary and anonymous, and all respondents provided informed consent prior to completing the questionnaire. The study procedures complied with ethical standards for research involving human participants. The sample showed a strong concentration in the Southern region of Brazil, particularly in the state of Rio Grande do Sul (n = 305), followed by smaller proportions from Santa Catarina (n = 6) and Paraná (n = 19). Participants from other Brazilian regions were also included, although in smaller numbers, covering states from the Northeast, Southeast, Midwest, and North regions. Instrument development The Scale of Determinant Factors in Out-of-Home Food Consumption (SDF-OHFC) was developed based on an integrative review of the international literature addressing economic, behavioral, social, and experiential determinants of eating outside the home. Item generation was guided by prior empirical and theoretical contributions on food choice behavior, convenience, habits, social eating, and perceived food quality. Table 1 presents the theoretical mapping of the scale items and their respective supporting literature. Table 1Theoretical foundations and literature mapping of the SDF-OHFC itemsFactorItemTheoretical BasisMain ReferencesConvenienceCON1–CON4Time scarcity, accessibility, service speed, practicality in food choicesStewart et al. [5]; Jabs & Devine [11]; Lim et al. [22].Food QualityFQ1–FQ4Perceived quality, healthy choices, taste, menu varietyColozza [19]; Glanz et al. [23]; Tuncer et al. [24].Economic AspectsEA1–EA4Price sensitivity, perceived value, affordability, cost–benefit evaluationNayga Jr. [25]; Binkley [26]; Gupta et al. [28].Social and Experiential AspectsSEA1–SEA4Social interaction, experiential consumption, symbolic meanings of eatingSobal & Nelson [29]; Wansink [30]; Zarantonello et al. [36].Habits and Personal PreferencesHPP1–HPP4Food routines, preferences, advertising influence, socially constructed eating habitsRezende & Avelar [20]; Bourdieu [33]; Andretti et al. [39]. Table 1Theoretical foundations and literature mapping of the SDF-OHFC itemsFactorItemTheoretical BasisMain ReferencesConvenienceCON1–CON4Time scarcity, accessibility, service speed, practicality in food choicesStewart et al. [5]; Jabs & Devine [11]; Lim et al. [22].Food QualityFQ1–FQ4Perceived quality, healthy choices, taste, menu varietyColozza [19]; Glanz et al. [23]; Tuncer et al. [24].Economic AspectsEA1–EA4Price sensitivity, perceived value, affordability, cost–benefit evaluationNayga Jr. [25]; Binkley [26]; Gupta et al. [28].Social and Experiential AspectsSEA1–SEA4Social interaction, experiential consumption, symbolic meanings of eatingSobal & Nelson [29]; Wansink [30]; Zarantonello et al. [36].Habits and Personal PreferencesHPP1–HPP4Food routines, preferences, advertising influence, socially constructed eating habitsRezende & Avelar [20]; Bourdieu [33]; Andretti et al. [39]. Theoretical foundations and literature mapping of the SDF-OHFC items Although several instruments addressing food consumption exist, no validated scale was identified that comprehensively captured the multidimensional determinants of out-of-home food consumption within the Brazilian context. Therefore, a new instrument was developed to address this gap. The initial pool of items was evaluated by a panel of six academic experts with expertise in consumer behavior, public health, and psychometrics to ensure content validity and contextual adequacy. Subsequently, the instrument was pilot-tested with a sample of 50 respondents, and an exploratory factor analysis (EFA) was conducted to assess the underlying factor structure and refine the scale. The results of this pre-test are presented in Table 2. Table 2Exploratory Factor Analysis Results and Reliability Indices (n = 50)FA / INCACOMFLKMOBST (p-value)Con0.8890.788< 0.001Con.010.4340.778Con.020.4370.693Con.030.5340.667Con.040.4170.663FQ0.7960.758< 0.001FQ.010.7660.875FQ.020.4810.626FQ.030.4340.684FQ.040.4250.658EA0.8620.872< 0.001EA.010.5310.729EA.020.4920.672EA.030.6360.957EA.040.4670.639SEA0.7860.792< 0.001SEA.010.4630.681SEA.020.5980.774SEA.030.4380.671SEA.040.4390.683HPP0.7700.726< 0.001HPP.010.4590.662HPP.020.4730.674HPP.030.5780.721HPP040.4390.653FA Factors, IN Indicators, CA Cronbach’s alpha, COM commonalities, FL factor loading, KMO Kaiser-Meyer-Olkin Index, BST Bartlett’s Sphericity Test, Con Convenience, FQ Food Quality, EA Economic Aspects, SEA Social and Experiential Aspects, HPP Habits and Personal Preferences Table 2Exploratory Factor Analysis Results and Reliability Indices (n = 50)FA / INCACOMFLKMOBST (p-value)Con0.8890.788< 0.001Con.010.4340.778Con.020.4370.693Con.030.5340.667Con.040.4170.663FQ0.7960.758< 0.001FQ.010.7660.875FQ.020.4810.626FQ.030.4340.684FQ.040.4250.658EA0.8620.872< 0.001EA.010.5310.729EA.020.4920.672EA.030.6360.957EA.040.4670.639SEA0.7860.792< 0.001SEA.010.4630.681SEA.020.5980.774SEA.030.4380.671SEA.040.4390.683HPP0.7700.726< 0.001HPP.010.4590.662HPP.020.4730.674HPP.030.5780.721HPP040.4390.653FA Factors, IN Indicators, CA Cronbach’s alpha, COM commonalities, FL factor loading, KMO Kaiser-Meyer-Olkin Index, BST Bartlett’s Sphericity Test, Con Convenience, FQ Food Quality, EA Economic Aspects, SEA Social and Experiential Aspects, HPP Habits and Personal Preferences Exploratory Factor Analysis Results and Reliability Indices (n = 50) FA Factors, IN Indicators, CA Cronbach’s alpha, COM commonalities, FL factor loading, KMO Kaiser-Meyer-Olkin Index, BST Bartlett’s Sphericity Test, Con Convenience, FQ Food Quality, EA Economic Aspects, SEA Social and Experiential Aspects, HPP Habits and Personal Preferences An initial pool of items was generated and reviewed by researchers with expertise in food behavior, consumer studies, and psychometrics to ensure content relevance and clarity. The items were phrased to reflect everyday out-of-home food consumption contexts and were designed to capture the frequency with which respondents experience each situation described. The preliminary version of the instrument was subjected to a pilot assessment to verify comprehension and response variability, resulting in minor wording adjustments. The final version of the SDF-OHFC consists of five latent dimensions (convenience, food quality, economic aspects, social and experiential aspects, and habits and personal preferences), comprising a total of 20 items (four items per dimension), all measured using a five-point Likert scale. The scale was theoretically grounded in the literature on food choice behavior, consumer decision-making, and food environments. The full list of items is presented in Appendix A. The questionnaire consisted of two sections: (i) sociodemographic and behavioral questions and (ii) the SDF-OHFC scale. All questions were closed-ended and predefined by the researchers, with no open-ended responses. Data collection procedure Data were collected using an online questionnaire hosted on a secure survey platform. Respondents first completed the informed consent form, followed by sociodemographic questions and the SDF-OHFC items. The questionnaire was self-administered and required approximately 10 min to complete. To minimize missing data, all items were set as mandatory fields. Statistical analysis Prior to the structural analysis, an exploratory factor analysis (EFA) was performed with a pilot sample of 50 consumers to examine the underlying factor structure of the initial item pool and to assess item adequacy, following classical recommendations for multivariate data analysis and scale development [39]. Although relatively small, this sample size is considered acceptable for exploratory purposes when the factor structure is theoretically supported and the indicators demonstrate adequate communalities and strong factor loadings [40]. The EFA supported the dimensional structure of the scale and informed item refinement for subsequent analyses. The analysis was conducted using IBM SPSS Statistics (version 31). The criteria adopted for the exploratory factor analysis included factor loadings greater than 0.60, Cronbach’s alpha values above 0.70, communalities exceeding 0.40, a Kaiser–Meyer–Olkin (KMO) index above 0.70, and a significant Bartlett’s test of sphericity (p < 0.05), indicating the suitability of the data for factor analysis. Data analysis was conducted using Partial Least Squares Structural Equation Modeling (PLS-SEM), implemented with SmartPLS (version 4.1.1.6) [41]. PLS-SEM was selected due to its suitability for predictive modeling, complex structural relationships, and scale development purposes, as well as its robustness to non-normal data distributions [42, 43]. The analysis followed a three-step procedure. First, the measurement model was evaluated by examining indicator reliability, internal consistency reliability (Cronbach’s alpha and composite reliability), convergent validity (average variance extracted, AVE), and discriminant validity using the Fornell–Larcker criterion and the heterotrait–monotrait ratio (HTMT) [44, 45]. Second, the structural model was assessed by analyzing path coefficients, coefficients of determination (R²), and predictive relevance (Q²). Collinearity among predictor factors was evaluated by examining variance inflation factors (VIF’s), with values below recommended thresholds indicating the absence of multicollinearity [43]. The statistical significance of the structural paths was tested using a bootstrapping procedure with 5,000 resamples [42]. In addition, out-of-sample predictive performance was assessed through robustness tests based on root mean square error (RMSE) and mean absolute error (MAE), allowing the evaluation of the model’s predictive accuracy and stability [46]. As a third step, measurement invariance of composite models (MICOM) was assessed to examine the stability of the measurement and structural models across subgroups [47]. Invariance was tested across sex, marital status, and Brazilian geographic regions, following the three-step MICOM procedure: configural invariance, compositional invariance, and equality of composite mean values and variances. Establishing invariance allowed for meaningful comparisons of the structural relationships across groups. Results Content validity assessment The content validity of the initial item pool was assessed by a panel of six independent academic experts. The experts evaluated each indicator individually, as well as the corresponding constructs, in terms of relevance, clarity, and representativeness of the theoretical domains. A Content Validity Index (CVI) was calculated for each item (I-CVI) based on the proportion of experts rating the item as relevant. The results indicated high levels of agreement across items, with I-CVI values ranging from 0.83 to 1.00. At the dimension level, the average CVI values were as follows: Convenience (0.915), Food Quality (0.915), Economic Aspects (1.000), Social and Experiential Aspects (0.958), and Habits and Personal Preferences (0.958). The overall scale-level CVI (S-CVI) was 0.949, indicating excellent content validity. Pre-test of the developed scale As shown in Table 2, the results confirmed the statistical and theoretical adequacy of the proposed factor structure for assessing determinants of out-of-home food consumption. All five theoretical factors demonstrated satisfactory internal consistency, with Cronbach’s alpha coefficients ranging from 0.770 (Habits and Preferences) to 0.889 (Convenience), indicating good reliability of the item sets. The KMO values ranged from 0.726 to 0.872, supporting sample adequacy for exploratory factor analysis. In addition, Bartlett’s test of sphericity was significant for all factors (p < 0.001), indicating sufficient inter-item correlations to justify factor extraction. Communality values ranged from 0.417 to 0.766, suggesting that the indicators were well represented by their respective latent dimensions. Factor loadings ranged from 0.626 to 0.957, indicating strong associations between items and their underlying factors. Overall, these findings provide robust initial evidence supporting the factorial structure of the scale and indicate that the indicators exhibit adequate psychometric properties to proceed with confirmatory validation using PLS-SEM. Sample profile Taken together, the sociodemographic characteristics (Table 3) and consumption-related variables (Table 4) provide a comprehensive overview of the profile and out-of-home eating practices of the respondents. The sample is predominantly composed of young and middle-aged adults, with a high level of educational attainment and a strong presence of individuals who are formally employed or working in the public sector, suggesting a population with relatively stable occupational conditions and structured daily routines. Table 3Sociodemographic profile of respondents (n = 426)VariablesCategoriesFrequency%SexFemale29168.3Male13531.7Marital statusMarried20447.9Single19044.6Others327.5Age range (years old)≤ 3015135.431–4514133.146–557517.7> 555913.8Educational AttainmentHigh school7216.9Undergraduate Degree15436.2Specialization or MBA9121.3Master’s degree and Doctorate10925.6Region of Brazil whereSouth33077.5you live?Other regions9622.5 Table 3Sociodemographic profile of respondents (n = 426)VariablesCategoriesFrequency%SexFemale29168.3Male13531.7Marital statusMarried20447.9Single19044.6Others327.5Age range (years old)≤ 3015135.431–4514133.146–557517.7> 555913.8Educational AttainmentHigh school7216.9Undergraduate Degree15436.2Specialization or MBA9121.3Master’s degree and Doctorate10925.6Region of Brazil whereSouth33077.5you live?Other regions9622.5 Sociodemographic profile of respondents (n = 426) Table 4Consumer Profile and Out-of-Home Eating Habits (n = 426)VariablesCategoriesFrequency%What is your occupation?Formal Employee11025.8Government Employee10725.1Student9722.8Self-Employed7517.6Others378.7What is your income? (USD)Up to $39011326.5$391 to $6959622.5$696 to $1,39011929.0Above $1,3909823.0How many times per week doOnce19245.1you eat out?2 to 3 times12930.24 or more times10524.7What is your preferred placeSelf-Service Restaurant29168.3to eat lunch?À la carte Restaurant337.7Steakhouse225.2Others368.5I don’t eat lunch outside the home4410.3What is your preferred placeFast Food Burger Restaurant13229.9to eat dinner?Pizzerias9121.4À la carte Restaurant7016.4Ethnic Restaurant368.5All-You-Can-Eat Restaurant307.0Others337.7I usually don’t have dinner348.0 Table 4Consumer Profile and Out-of-Home Eating Habits (n = 426)VariablesCategoriesFrequency%What is your occupation?Formal Employee11025.8Government Employee10725.1Student9722.8Self-Employed7517.6Others378.7What is your income? (USD)Up to $39011326.5$391 to $6959622.5$696 to $1,39011929.0Above $1,3909823.0How many times per week doOnce19245.1you eat out?2 to 3 times12930.24 or more times10524.7What is your preferred placeSelf-Service Restaurant29168.3to eat lunch?À la carte Restaurant337.7Steakhouse225.2Others368.5I don’t eat lunch outside the home4410.3What is your preferred placeFast Food Burger Restaurant13229.9to eat dinner?Pizzerias9121.4À la carte Restaurant7016.4Ethnic Restaurant368.5All-You-Can-Eat Restaurant307.0Others337.7I usually don’t have dinner348.0 Consumer Profile and Out-of-Home Eating Habits (n = 426) This profile is consistent with the observed patterns of out-of-home food consumption. Respondents reported moderate to high frequency of eating outside the home, with nearly half eating out at least once per week and approximately one quarter doing so four or more times weekly. Such frequency reflects the integration of out-of-home eating into everyday life, particularly among economically active individuals facing time constraints associated with work and urban living. Income distribution further contextualizes these behaviors. Although a substantial proportion of respondents reported lower to middle income levels, out-of-home food consumption remains frequent, indicating that eating outside the home is not restricted to high-income groups but has become a routine practice across income strata. This finding reinforces the relevance of economic considerations in shaping food choices, especially regarding affordability and perceived value. Clear temporal and contextual differences emerge when lunch and dinner consumption are examined jointly. While evening consumption has been associated with more experiential and indulgent motivations, it may also reflect practical considerations, such as time constraints, fatigue after daily activities, and the search for convenience. In this sense, the preference for fast food during the evening period can be interpreted not only as a hedonic choice but also as a functional response to situational demands. In contrast, evening consumption was more diversified and concentrated in hamburger restaurants, fast-food outlets, and pizzerias, indicating a shift toward variety, indulgence, and experiential consumption. This distinction suggests that lunch is more strongly governed by functional and economic constraints, whereas dinner choices are more closely associated with leisure, social interaction, and hedonic motivations. Overall, the combined evidence from Tables 3 and 4 highlights that out-of-home food consumption among this sample constitutes a multidimensional behavior, shaped by the interaction of sociodemographic characteristics, occupational structure, income conditions, and situational contexts. These patterns support the relevance of adopting an integrative analytical framework that simultaneously considers economic aspects, convenience, habits and personal preferences, and social and experiential factors, as proposed in the conceptual model of this study. Measurement and structural model assessment and predictive validity With regard to the structural model, convergence was achieved after ten iterations, indicating stability of the estimated parameters. Overall model fit indices showed satisfactory results. The standardized root mean square residual (SRMR) was 0.071, below the recommended threshold of 0.08, indicating good global model fit to the observed data [47]. The normed fit index (NFI) reached 0.830, exceeding the minimum recommended value of 0.80 and further supporting the adequacy of the structural model [48]. Distance-based discrepancy measures also indicated low divergence between the observed and estimated covariance matrices. The squared Euclidean distance (dULS = 0.528) and the geodesic distance (dG = 0.252) suggest satisfactory model fit. Although no universally accepted cutoff values exist for these indices, the obtained values are considered acceptable and consistent with the proposed theoretical structure [49]. Table 5 presents the results of the measurement model validation for the five factors of the scale. All factors exhibited acceptable levels of internal consistency reliability, with Cronbach’s alpha (CA) values ranging from 0.705 to 0.762, and composite reliability (CR) values between 0.734 and 0.816, exceeding the recommended minimum threshold of 0.70 [42]. Table 5Measurement model validationFactorCACRAVEConvenience (Con)0.7100.7340.519Economic Aspects (EA)0.7620.8160.528Food Quality (FQ)0.7480.7670.528Habits and Personal Preferences (HPP)0.7050.7460.517Social and Experiential Aspects (SEA)0.7260.7820.557CA Cronbach’s alpha, CR composite reliability, AVE average variance extracted Table 5Measurement model validationFactorCACRAVEConvenience (Con)0.7100.7340.519Economic Aspects (EA)0.7620.8160.528Food Quality (FQ)0.7480.7670.528Habits and Personal Preferences (HPP)0.7050.7460.517Social and Experiential Aspects (SEA)0.7260.7820.557CA Cronbach’s alpha, CR composite reliability, AVE average variance extracted Measurement model validation CA Cronbach’s alpha, CR composite reliability, AVE average variance extracted Regarding convergent validity, all factors achieved average variance extracted (AVE) values above the minimum criterion of 0.50, indicating that a substantial proportion of the variance of the indicators is explained by their respective latent factors. Specifically, AVE values ranged from 0.517 to 0.557, supporting adequate convergent validity across all factors. Overall, these findings demonstrate that the measurement model exhibits satisfactory reliability and convergent validity, providing robust support for the internal consistency and adequacy of the proposed scale and enabling subsequent interpretation of the structural relationships among the factors. As shown in Table 6, the discriminant validity of the measurement model was confirmed based on the Fornell–Larcker criterion and the heterotrait–monotrait ratio (HTMT). The square roots of the average variance extracted (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\sqrt{AVE}$$\end{document}) for each factor were greater than the corresponding inter-factor correlations, satisfying the Fornell–Larcker criterion [44]. Table 6Fornell-Larcker and HTMT criterionFactor \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\sqrt{AVE}$$\end{document} Pearson Correlation MatrixConEAFQHPPSEACon0.7201.000EA0.7400.2981.000FQ0.7270.1160.0171.000HPP0.7190.2170.2690.0731.000SEA0.746-0.018-0.0980.2290.2401.000 Upper Limit (HTMT) 97.5% EA0.213FQ0.2110.630HPP0.1330.7020.784SEA0.1280.7570.7580.908Con Convenience, EA Economic Aspects, FQ Food Quality, HPP Habits and Personal Preferences, SEA Social and Experiential Aspects, AVE average variance extracted Table 6Fornell-Larcker and HTMT criterionFactor \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\sqrt{AVE}$$\end{document} Pearson Correlation MatrixConEAFQHPPSEACon0.7201.000EA0.7400.2981.000FQ0.7270.1160.0171.000HPP0.7190.2170.2690.0731.000SEA0.746-0.018-0.0980.2290.2401.000 Upper Limit (HTMT) 97.5% EA0.213FQ0.2110.630HPP0.1330.7020.784SEA0.1280.7570.7580.908Con Convenience, EA Economic Aspects, FQ Food Quality, HPP Habits and Personal Preferences, SEA Social and Experiential Aspects, AVE average variance extracted Fornell-Larcker and HTMT criterion Con Convenience, EA Economic Aspects, FQ Food Quality, HPP Habits and Personal Preferences, SEA Social and Experiential Aspects, AVE average variance extracted In addition, all HTMT values, including the upper limits of the 97.5% confidence intervals, were below the recommended threshold of 1.0, indicating adequate discriminant validity among the constructs [44]. Overall, these results provide robust evidence that the factors are empirically distinct, supporting the adequacy of the measurement model and confirming satisfactory discriminant validity. As reported in Table 7, the structural model assessment indicated the absence of multicollinearity among the predictor factors. All variance inflation factor (VIF) values were below 1.08, well under the conservative threshold of 5.0, indicating optimal collinearity conditions and ensuring the stability of the estimated path coefficients. Table 7Multicollinearity analysis, coefficient of determination, and predictive relevanceExogenousFactorsEndogenous Factors(VIF value)ConFQHPPSEACon1.000EA1.0781.0001.078HPP1.0781.078SEA1.000 Predictive Validity Coefficients R2 (p-value)0.209 (< 0.001)0.167 (< 0.001)0.172 (< 0.001)0.186 (< 0.001)Q20.1730.1310.1640.149RMSE0.8660.8080.9710.804MAE0.7610.7880.7560.708VIF variance inflation factor, Con Convenience, EA Economic Aspects, FQ Food Quality, HPP Habits and Personal Preferences, SEA Social and Experiential Aspects Table 7Multicollinearity analysis, coefficient of determination, and predictive relevanceExogenousFactorsEndogenous Factors(VIF value)ConFQHPPSEACon1.000EA1.0781.0001.078HPP1.0781.078SEA1.000 Predictive Validity Coefficients R2 (p-value)0.209 (< 0.001)0.167 (< 0.001)0.172 (< 0.001)0.186 (< 0.001)Q20.1730.1310.1640.149RMSE0.8660.8080.9710.804MAE0.7610.7880.7560.708VIF variance inflation factor, Con Convenience, EA Economic Aspects, FQ Food Quality, HPP Habits and Personal Preferences, SEA Social and Experiential Aspects Multicollinearity analysis, coefficient of determination, and predictive relevance VIF variance inflation factor, Con Convenience, EA Economic Aspects, FQ Food Quality, HPP Habits and Personal Preferences, SEA Social and Experiential Aspects The coefficients of determination (R²) ranged from 0.167 to 0.209, and all values were statistically significant (p < 0.05). Although these values indicate modest explanatory power, they are considered appropriate in the context of scale development and validation studies, where theoretical coherence, measurement quality, and the significance of structural relationships are more critical than achieving high R² values. Regarding predictive validity, all Stone–Geisser’s Q² values were positive, providing evidence of the model’s predictive relevance for the endogenous factors. In addition, predictive error indicators, including root mean square error (RMSE) and mean absolute error (MAE), showed acceptable magnitudes across all factors, indicating consistency and stability in the model’s out-of-sample predictive performance. Overall, these findings support the adequacy of the structural model, confirming its robustness, predictive relevance, and suitability for interpreting the relationships among the determinants of out-of-home food consumption. Measurement invariance and multi-group analysis The structural invariance analysis aimed to examine whether the path coefficients (β) of the proposed model differed significantly across sociodemographic subgroups. Multi-group analysis was conducted using the Henseler’s MGA test (HT) and the permutation test (PT) to compare structural paths across groups, as reported in Table 8. Table 8Invariance analysis results – gender, marital status, and Brazilian RegionRelationshipGenderΔβF−Mp-valueβF (p-value)βM (p-value)HTPTH8-aEA → Con0.336 (< 0.001)0.124 (0.316)0.2110.1080.084H8-bEA → HPP0.241 (< 0.001)0.326 (0.001)-0.0850.4380.478H8-cEA → SEA-0.254 (< 0.001)0.056 (0.794)-0.3100.2130.234H8-dHPP → SEA0.296 (< 0.001)0.253 (0.031)0.0430.7870.695H8-eHPP → Con0.147 (0.037)0.157 (0.263)-0.0100.8840.942H8-fSEA → FQ0.226 (0.001)0.230 (0.031)-0.0040.9200.979H8-gCon → FQ0.068 (0.384)0.170 (0.141)-0.1020.3960.452 Brazilian Regions Δβ So−OR βSo (p-value)βOR (p-value)H10-aEA → Con0.253 (< 0.001)0.274 (0.049)-0.0210.7890.853H10-bEA → HPP0.252 (< 0.001)0.343 (0.008)-0.0910.4060.451H10-cEA → SEA-0.181 (0.018)-0.122 (0.507)0.0600.7980.721H10-dHPP → SEA0.286 (< 0.001)0.288 (0.011)-0.0020.9450.983H10-eHPP → Con0.165 (0.015)0.121 (0.421)0.0440.8150.725H10-fSEA → FQ0.220 (0.010)0.475 (< 0.001)-0.2550.1020.106H10-gCon → FQ0.146 (0.047)0.088 (0.547)0.0580.7510.687 Marital Status Δβ Ma−Ot βMa (p-value)βOt (p-value)H9-aEA → Con0.326 (< 0.001)0.198 (0.005)0.1270.2180.232H9-bEA → HPP0.205 (0.029)0.317 (< 0.001)-0.1120.2540.283H9-cEA → SEA-0.341 (< 0.001)-0.103 (0.979)-0.2380.1020.111H9-dHPP → SEA0.254 (< 0.001)0.319 (< 0.001)-0.0640.4910.495H9-eHPP → Con0.016 (0.867)0.272 (0.001)-0.2570.1230.127H9-fSEA → FQ0.309 (< 0.001)0.176 (0.0400.1340.4960.567H9-gCon → FQ0.131 (0.126)0.152 (0.101)-0.0210.8520.855β path coefficient, Δβ difference, Con Convenience, EA Economic Aspects, FQ Food Quality, HPP Habits and Personal Preferences, SEA Social and Experiential Aspects, HT Henseler Test, PT Permutation Test, F Female, M Male, Ma Married, Ot Others, So South, OR Other Regions Table 8Invariance analysis results – gender, marital status, and Brazilian RegionRelationshipGenderΔβF−Mp-valueβF (p-value)βM (p-value)HTPTH8-aEA → Con0.336 (< 0.001)0.124 (0.316)0.2110.1080.084H8-bEA → HPP0.241 (< 0.001)0.326 (0.001)-0.0850.4380.478H8-cEA → SEA-0.254 (< 0.001)0.056 (0.794)-0.3100.2130.234H8-dHPP → SEA0.296 (< 0.001)0.253 (0.031)0.0430.7870.695H8-eHPP → Con0.147 (0.037)0.157 (0.263)-0.0100.8840.942H8-fSEA → FQ0.226 (0.001)0.230 (0.031)-0.0040.9200.979H8-gCon → FQ0.068 (0.384)0.170 (0.141)-0.1020.3960.452 Brazilian Regions Δβ So−OR βSo (p-value)βOR (p-value)H10-aEA → Con0.253 (< 0.001)0.274 (0.049)-0.0210.7890.853H10-bEA → HPP0.252 (< 0.001)0.343 (0.008)-0.0910.4060.451H10-cEA → SEA-0.181 (0.018)-0.122 (0.507)0.0600.7980.721H10-dHPP → SEA0.286 (< 0.001)0.288 (0.011)-0.0020.9450.983H10-eHPP → Con0.165 (0.015)0.121 (0.421)0.0440.8150.725H10-fSEA → FQ0.220 (0.010)0.475 (< 0.001)-0.2550.1020.106H10-gCon → FQ0.146 (0.047)0.088 (0.547)0.0580.7510.687 Marital Status Δβ Ma−Ot βMa (p-value)βOt (p-value)H9-aEA → Con0.326 (< 0.001)0.198 (0.005)0.1270.2180.232H9-bEA → HPP0.205 (0.029)0.317 (< 0.001)-0.1120.2540.283H9-cEA → SEA-0.341 (< 0.001)-0.103 (0.979)-0.2380.1020.111H9-dHPP → SEA0.254 (< 0.001)0.319 (< 0.001)-0.0640.4910.495H9-eHPP → Con0.016 (0.867)0.272 (0.001)-0.2570.1230.127H9-fSEA → FQ0.309 (< 0.001)0.176 (0.0400.1340.4960.567H9-gCon → FQ0.131 (0.126)0.152 (0.101)-0.0210.8520.855β path coefficient, Δβ difference, Con Convenience, EA Economic Aspects, FQ Food Quality, HPP Habits and Personal Preferences, SEA Social and Experiential Aspects, HT Henseler Test, PT Permutation Test, F Female, M Male, Ma Married, Ot Others, So South, OR Other Regions Invariance analysis results – gender, marital status, and Brazilian Region β path coefficient, Δβ difference, Con Convenience, EA Economic Aspects, FQ Food Quality, HPP Habits and Personal Preferences, SEA Social and Experiential Aspects, HT Henseler Test, PT Permutation Test, F Female, M Male, Ma Married, Ot Others, So South, OR Other Regions With respect to gender, the results indicate that none of the seven hypothesized relationships exhibited statistically significant differences between female and male respondents. All p-values associated with both the Henseler’s MGA and permutation tests were above the 0.05 significance threshold, indicating that the structural path coefficients are invariant across gender groups. These findings suggest that the effects specified in the structural model operate similarly for women and men. Regarding marital status, no statistically significant differences were observed in any of the seven structural relationships when comparing married respondents with those classified as other marital statuses. All comparisons yielded non-significant p-values (p > 0.05) in both tests, providing evidence of structural invariance across marital status groups. This result supports the equivalence of the structural relationships regardless of respondents’ marital situation. Similarly, the analysis comparing respondents residing in the Southern region of Brazil with those living in other Brazilian regions revealed no statistically significant differences in the structural path coefficients across all seven tested relationships. The absence of significant differences (p > 0.05) indicates that the structural effects proposed in the model are consistent across regional contexts. Overall, the invariance analyses across gender, marital status, and geographic region demonstrate that the proposed model is structurally stable and homogeneous across the examined subgroups. These findings confirm the robustness of the theoretical model and justify the interpretation of the structural relationships based on the full sample, without the need for subgroup-specific estimations. Table 9 presents the results of the hypothesis testing for the proposed structural model, based on the estimated path coefficients (β), standard errors, t-values, and significance levels. All seven hypotheses were statistically supported (p < 0.05), indicating significant relationships among the factors included in the model. Table 9Structural model hypotheses testing resultsRelationshipβSDStat. tp-valueResultH1EA → Con0.2580.0544.764< 0.001SupportedH2EA → HPP0.2690.0554.938< 0.001SupportedH3EA → SEA-0.1750.0652.6920.007SupportedH4HPP→ SEA0.2870.0495.865< 0.001SupportedH5HPP → Con0.1480.0582.5450.011SupportedH6SEA → FQ0.2310.0504.591< 0.001SupportedH7Con → FQ0.1200.0592.0230.043Supportedβ path coefficient, SD Standard Deviation, EA Economic Aspects, Con Convenience, HPP Habits and Personal Preferences, SEA Social and Experiential Aspects, FQ Food Quality Table 9Structural model hypotheses testing resultsRelationshipβSDStat. tp-valueResultH1EA → Con0.2580.0544.764< 0.001SupportedH2EA → HPP0.2690.0554.938< 0.001SupportedH3EA → SEA-0.1750.0652.6920.007SupportedH4HPP→ SEA0.2870.0495.865< 0.001SupportedH5HPP → Con0.1480.0582.5450.011SupportedH6SEA → FQ0.2310.0504.591< 0.001SupportedH7Con → FQ0.1200.0592.0230.043Supportedβ path coefficient, SD Standard Deviation, EA Economic Aspects, Con Convenience, HPP Habits and Personal Preferences, SEA Social and Experiential Aspects, FQ Food Quality Structural model hypotheses testing results β path coefficient, SD Standard Deviation, EA Economic Aspects, Con Convenience, HPP Habits and Personal Preferences, SEA Social and Experiential Aspects, FQ Food Quality As shown in Table 9, Economic Aspects played a central antecedent role in explaining out-of-home food consumption. Specifically, Economic Aspects exerted positive effects on Convenience and Habits and Personal Preferences, while showing a negative effect on Social and Experiential Aspects. These findings suggest that economic conditions simultaneously facilitate convenience-oriented choices and shape habitual consumption patterns, while constraining or reshaping social and experiential dimensions of eating outside the home. Habits and Personal Preferences also demonstrated positive and significant effects on both Social and Experiential Aspects and Convenience, reinforcing the role of routine-based and preference-driven behaviors in structuring out-of-home eating practices. Furthermore, both Social and Experiential Aspects and Convenience were positively associated with Perceived Food Quality, indicating that quality perceptions are influenced not only by functional considerations but also by experiential and contextual factors. Overall, these results provide empirical support for the proposed theoretical structure and demonstrate coherent relationships among the latent factors. The findings validate the structural model and confirm the adequacy of the SDF-OHFC scale as a robust instrument for analyzing the determinants of out-of-home food consumption in Brazil. By capturing economic, behavioral, and experiential dimensions simultaneously, the scale offers a comprehensive framework for understanding food-related decision-making across diverse urban and regional contexts. Discussion The findings of this study confirm that out-of-home food consumption in the Brazilian context is a multidimensional phenomenon shaped by the interaction of economic, practical, social, experiential, and individual factors. The validation of the SDF-OHFC reinforces the relevance of integrative analytical approaches for understanding contemporary food choices in urban contexts characterized by time scarcity, economic pressure, and diversified food environments [20, 35]. The structural results indicate that Economic Aspects play a central antecedent role, exerting significant effects on Convenience, Habits and Personal Preferences, and Social and Experiential Aspects. These findings are consistent with previous studies emphasizing price sensitivity, disposable income, and perceived value as important determinants of eating outside the home, particularly in contexts marked by socioeconomic inequalities [14, 50]. The negative relationship between Economic Aspects and Social and Experiential Aspects suggests that stronger economic constraints tend to orient food consumption toward more functional and utilitarian practices, reducing the emphasis on symbolic and experiential dimensions of eating out. In contrast to traditional interpretations that associate eating out primarily with social distinction and cultural capital [33], the present findings suggest a reconfiguration of this behavior in contemporary urban routines, in which Convenience emerges as a central dimension linking structural constraints to individual decision-making. Consistent with previous literature, convenience reflects not only speed and accessibility but also an adaptive response to time scarcity, work intensity, and competing daily demands [5, 11, 34]. In this sense, convenience functions as a mechanism through which individuals manage occupational, family, and social pressures within contemporary food environments. The significant effects of Habits and Personal Preferences on both Convenience and Social and Experiential Aspects reinforce sociocultural perspectives that conceptualize eating practices as socially constructed and embedded in routines and cultural repertoires [33, 51]. These findings suggest that food choices are shaped not only by individual preferences but also by repeated exposure, socialization processes, and contextual conditions associated with everyday life. The results related to Perceived Food Quality further highlight the relational character of out-of-home food consumption. Perceptions of food quality were positively influenced by both Convenience and Social and Experiential Aspects, indicating that quality evaluations are not based exclusively on nutritional attributes but also on atmosphere, service, social interaction, and broader consumption experiences [24, 36, 52]. This finding reinforces the importance of considering experiential and contextual dimensions when examining eating-out behaviors and food environments. By integrating economic, behavioral, and experiential dimensions within a single framework, the SDF-OHFC addresses an important gap in the literature, where validated multidimensional instruments remain limited, particularly in middle-income countries undergoing nutritional and urban transitions. From an applied perspective, the findings suggest that pricing strategies, service accessibility, and the organization of urban food environments may substantially influence food-related decisions, especially among individuals facing economic and occupational pressures. The influence of social and experiential aspects on perceived food quality also highlights the importance of food environments that foster positive dining experiences and supportive social interactions. Together, these findings indicate that interventions aimed at improving dietary practices in out-of-home contexts should adopt integrated approaches that combine affordability, accessibility, service organization, and healthier food environments. From a methodological perspective, the observed structural invariance across gender, marital status, and Brazilian regions provides evidence of the psychometric robustness and stability of the SDF-OHFC across different sociodemographic groups. These results strengthen the scale’s potential for use in future comparative studies and applied public health research. Some limitations should be acknowledged. First, the cross-sectional design does not allow causal inferences regarding the relationships among the constructs. Second, the use of self-reported measures may introduce response and social desirability biases. Third, data collection through online recruitment and non-probabilistic sampling may have favored the participation of individuals with greater internet access and higher educational levels, limiting the representativeness of the sample. In addition, the strong concentration of respondents from the Southern region of Brazil may restrict the generalizability of the findings to other regional and cultural contexts. Future studies should replicate the validation of the SDF-OHFC using probabilistic and more geographically balanced samples, as well as longitudinal designs capable of examining changes in eating-out behavior over time. Conclusion This study aimed to propose and validate the Scale of Determinant Factors in Out-of-Home Food Consumption (SDF-OHFC), structured around five interrelated dimensions: economic aspects, convenience, habits and personal preferences, social and experiential aspects, and perceived food quality. The results provide empirical support for the proposed hierarchical structure, demonstrating satisfactory psychometric properties and confirming the theoretical coherence of the scale within the framework of contemporary food environment research. Overall, the findings reinforce the notion that out-of-home food consumption extends beyond purely nutritional or economic considerations. Food-related decisions in out-of-home contexts are shaped by a complex interplay of economic constraints, everyday practices, social interactions, experiential meanings, and individual habits. In particular, the negative association between Economic Aspects and Social and Experiential Aspects highlights how increased economic constraints tend to reorient eating-out behavior toward more functional and utilitarian patterns, reducing its symbolic and experiential dimensions. From a methodological perspective, the SDF-OHFC demonstrated satisfactory reliability, as well as convergent and discriminant validity, within the analyzed sample. However, the generalizability of these findings should be interpreted with caution, given the concentration of the sample in the Southern region of Brazil and among individuals with higher educational attainment. From an applied perspective, the findings offer actionable insights for public policies and interventions aimed at improving dietary practices in out-of-home contexts. In particular, strategies that address affordability, accessibility, and time-related constraints may help promote healthier food choices, especially among populations facing economic and occupational pressures. In addition, the SDF-OHFC provides a practical tool for policymakers and practitioners to diagnose consumption patterns and design targeted interventions within urban food environments. Future research should replicate and validate the SDF-OHFC in different countries, regions, and population groups, further examining its measurement invariance across diverse social and territorial contexts. Additionally, future studies may explore the relationships between the scale dimensions and indicators of diet quality, health outcomes, sustainability, and food security, as well as adopt longitudinal designs to capture changes in out-of-home food consumption patterns in response to social, economic, and policy transformations.