Effect of lifestyle, and health consciousness on street food purchase intention using the TPB model Open Access

Street food, referred to as “informal food sector,” “on-the-go meals,” or “out-of-home eating” (Albuquerque et al., 2020; Sousa et al., 2019; Bouafou et al., 2021), plays a vital role in food consumption. Over the past three decades, street food has attracted academic interest from foundational studies in the 1990s to recent research in the 2020s (e.g. Albuquerque et al., 2020; Ferrari et al., 2021). This consideration reflects the role of street food in the lives of urban populations, especially in growing cities where hanging out culture, dual-income households, demanding work schedules, long travel times and time constraints that make home cooking difficult, restrict access to formal dining options. Therefore, street food has emerged as a convenient and affordable source of sustenance (Bouafou et al., 2021).

College students, daily wage workers, professionals, and even the unemployed (Van ’t Riet et al., 2003; Ohiokpehai, 2003) rely on street food. This segment is termed “Street Food eaters. They consume the street food based on choices of Food Quality (Peri, 2006; Zeithaml, 1988), Value (Tacardon et al., 2023), health consciousness (HEC) (Khanna et al., 2022), Convenience (Buckley et al., 2007), and Lifestyle (Wiatrowski et al., 2021). In urban areas, street food serves as a source of nourishment and plays an important part in everyday life, social connections, and cultural identity (Allen et al., 2018). Thus, as more people turn to street food, there is a need to closely understand consumer perceptions, intentions, and behaviour.

The street food market is a growing global phenomenon. As of 2020, the street food industry worldwide was estimated at $1.2 trillion and is expected to grow steadily, reaching around $1.8 trillion by 2027 with a growth rate of about 6.5% annually. In the global markets, Asia Pacific leads in terms of market share, with North America and Europe following behind (Ess Feed, 2024). Thailand, Mexico, and China are known for their diverse and vibrant street food with popular dishes like Pad Thai, Tacos al Pastor, and Jianbing (Mwove et al., 2020). Despite its strong growth potential, the context of the emerging Indian economy presents a contrasting picture. According to International Market Analysis Research and Consulting Group (IMARC) Group Estimates (2024), 2024 India's fast-food industry was estimated at around $18.6 billion. This sector is expected to grow consistently, reaching nearly $35.5 billion by 2033, with a compound annual growth rate of 7.1% between 2025 and 2033. In India, research is limited and focused mainly on street food consumers, leaving a gap in understanding behaviour and purchase intention. In India, research is focused on food safety (Sabbithi et al., 2024), hygiene (Sharma et al., 2020), and vendor livelihoods (Aman and Singh, 2024), with limited exploration of consumer behaviour dynamics. Behavioural aspects like food quality, value, HEC and lifestyle can be studied to design strategies that boost satisfaction and industry growth. The study aims to evaluate the role of these factors along with Attitude (ATT), Subjective norms (SUN) and Perceived Behavioural Control (PBC) on the Purchase Intention of street food consumers in Anand and Gandhinagar cities of Gujarat. India.

The Theory of Planned Behaviour (TPB), developed by Ajzen (1991), provides an established framework for examining consumer intentions by focussing on ATT, PBC and SUN. Although TPB has been widely used in food consumption studies across the world, its application to the Indian street food sector remains limited. To gain a broader understanding of consumer behaviour for street food, this study extends TPB by including additional factors: perceived food value (PFV), perceived food quality (PFQ), and convenience (CON). The study also assesses the effect of PBC, SUN, ATT, consumer lifestyle (COL), and HEC on street food purchase intention (SFP) in a city-specific setting of Gujarat. Through structural equation modelling, the research identified and assessed causal relationships among these factors. The findings could aid vendors in enhancing consumer experiences and guide policymakers in developing supportive regulations.

The forthcoming sections of this paper present the conceptual framework and hypotheses development followed by a discussion on questionnaire design, sample selection and methods of data collection. The subsequent parts describe the results of the study including analysis of measurement and structural models, hypothesis testing, and discussion of findings. The paper concludes with theoretical contributions, practical implications, limitations, directions for future research, and conclusions.

Globally, consumer behaviour towards the consumption of street food reflects the trend towards urbanisation and street food itself is often regarded as a reflection of cultural traditions (Soliman et al., 2024). It is crucial to understand the factors that influence consumers' choices to purchase street food for public health purposes and businesses within this ever-changing environment. The TPB is often used in social sciences to analyse consumer behaviour for food consumption (Ajzen, 1991). Integration of TPB with the intention of consumers suggests that an individual's decision to engage in a certain activity is affected by three factors, i.e. PBC, SUN and ATT. TPB has been applied in studies focused on the consumption of food like healthy eating (Fila and Smith, 2006), organic food (Dean et al., 2012; Al-Swidi et al., 2014) and fast foods (Dunn et al., 2011). As stated earlier, the TPB model has been widely used across a variety of disciplines and new variables were added to the original TPB framework by many researchers who have conducted studies on the consumption of food and purchase behaviour (Dorce et al., 2021). A study by Tacardon et al. (2023) addressed this gap by extending the TPB framework to include PFV, PFQ, and CON as additional variables influencing SFP. With regard to consuming street foods, consumers' decisions are determined by their COL (Buckley et al., 2007) and HEC (Kumar and Smith, 2017; Hoque et al., 2018; Sobaih et al., 2023). In the Indian context, consumption of street food is widespread, and an exhaustive empirical study of street food consumption remains fragmented. Most existing studies have focused on identifying the relationship among the core TPB constructs with SFP. The present study empirically offers a unique perspective by integrating COL and HEC constructs which have not been addressed in the extant literature in relation to SFP. Thus, the study of these factors, i.e. PFV, PFQ, CON, COL and HEC in addition to the core constructs of the TPB, and the nexus among these constructs, especially on SFP, evaluated under a specific geographical region, may facilitate a better understanding of behavioural intentions.

PBC reflects the individual's belief about the ease or difficulty of performing a particular behaviour (Ajzen, 1991). An individual's belief about the resource availability, opportunities and barriers that may facilitate or hinder the behaviour influences PBC (Huang and Chuang, 2004). Tsai (2009) documented that engaging in a certain behaviour required actual control over time and money. Even a favourable attitude may not lead to intention to purchase street food if the individuals perceive a low control, while consumers with a higher PBC are more likely to have such food based on hygiene, affordability or access. SUN reflects social pressure from relevant individuals or groups to either perform or refrain from performing certain behaviour (Bagozzi et al., 2000; Torres Chavarria and Phakdee-Auksorn, 2017).

Recommendations from peers and certain cultural preferences may play a crucial role in determining SFP (Tacardon et al., 2023). ATT is an individual's positive or negative assessment of engaging in the behaviour. Attitude plays a major role in consumer behaviour and is influenced by prior experiences. Beliefs, emotions, and intentions influence how individuals react to actions, ideas, or objects (Torres Chavarria and Phakdee-Auksorn, 2017). Hoque et al. (2018), suggested that consumer attitudes toward safety, taste and overall satisfaction may significantly impact their street purchase decisions. Thus, integrating PBC, SUN and ATT for purchasing street food, it may be said that an individual having a positive attitude towards street food consumption, perceiving that important others endorse of buying such food and believing that purchasing street food is within their control, are likely to result in to a stronger intention to make purchase decisions (Ajzen, 1991; Binh and Mai, 2023; Tacardon et al., 2023).

Beyond the fundamental attributes of the TPB, several other factors also play a role in shaping SFP. Evaluation of the excellence or superiority of a food product where consumers are concerned about the hygiene and food safety is represented by PFQ (Zeithaml, 1988; Tacardon et al., 2023). This is a subjective process where consumers rely on freshness, taste and appearance to assess food quality (Peri, 2006; Zeithaml, 1988). Morano et al. (2018) found that for street food vendors, service-related characteristics such as cleanliness and politeness, influenced consumers' perceptions of healthiness. This, in turn, affects perceived product quality. Consumers' PBC is improved as they are likely to have positive opinions of street food and feel more confident about their ability to select safe and high-quality food options (Rossi et al., 2018). Improved perceived quality can influence positive attitudes, conform to social norms and boost consumer confidence (Gupta et al., 2018).

PFV refers to the trade-off consumers make between the benefits in terms of quality, taste, convenience and sacrifices in terms of price and time associated with a food product (Zeithaml, 1988; Tacardon et al., 2023). This assessment is psychological and subjective, which is influenced by the perceptions of an individual regarding the benefits and costs (Zeithaml, 1988). In addition to the economic aspect, perceived value also includes functional, social and emotional dimensions (Chen and Chang, 2012). Consumers are more likely to prefer street food that they perceive as offering good value for money (Tacardon et al., 2023). This perception can lead to a sense of control over their purchasing choices, develop a favourable attitude towards street food consumption and lead them to believe that society would also view it favourably (Hoque et al., 2018).

Several studies have emphasised that CON significantly influences consumer behaviour and purchase intentions, particularly in the context of street food consumption. Anderson (1971) identified that convenience-oriented consumers prioritise time-saving and effort-reducing aspects when making purchase decisions. Subsequently, in a study by the same author in 1972, convenience has been defined as a multidimensional concept that includes task simplification beyond time and effort. Many individuals prefer street food for its ease of access, speed of service and the fact that such food requires no preparation (Berry et al., 2002). Gupta et al. (2018) found that both PFV and CON strongly influenced consumer attitudes and behavioural intentions. In alignment with this, Naufal and Nelloh (2021) noted that convenience has a significant role in shaping millennials' and Gen Z's purchase intentions by studying preloved apps. These findings can be attributed to street food consumption for affordability and speed. Brown (1990) proposed that marketers must understand and address the factor of convenience to align product offerings with consumers' lifestyle needs. Apart from attitude formation, convenience also influences SUN and PBC, the other critical components of the TPB, as shown by (Ham et al., 2015). Further, Chowdhury (2023) emphasised that convenience not only directly affects consumer behavioural intentions but also operates through mediators such as attitude, particularly in online food delivery scenarios. Moreover, the widespread availability and social acceptance of convenient street food options might influence attitude and SUN (Berry et al., 2002). Therefore, it was hypothesised that

COL includes factors related to how people live their lives, their activities, interests and opinions. Food preferences and purchasing behaviour can differ according to lifestyle (Plummer, 1974). A study by Buckley et al. (2007), found that “Kitchen Evaders”, i.e. those who wanted to minimise the time spent on cooking, andConvenience-Seeking Grazers' segments, i.e. those who preferred ready-to-eat food as they were easily available and offered time-saving options, were more often driven by the desire for convenience. Many studies have explored and found that HEC affected the purchase intention directly and indirectly via the three core constructs of the TPB (Tacardon et al., 2023; Lai et al., 2020; Wang and Wang, 2021). Khanna et al. (2022) studied perceived risk as a variable affecting purchase intention, suggesting an indirect link to HEC within the TPB model in relation to street food consumption. Health-conscious people may choose healthier food choices over street food because of hygienic concerns (Rozin et al., 1999). Considering the effect of COL and HEC on SFP, it was hypothesised that

The research model is presented in Figure 1.

Using the rigorous literature review, the researchers identified antecedents affecting the SFP intention among young consumers. In addition to the core constructs (ATT, SUN, PBC) of the TPB theory, COL and HEC were also examined in the study. In addition, the researchers explored how factors such as PFV, PFQ, and CON, influence ATT, SUN, and PBC.

Since the study employed a primary research approach, the development of the questionnaire was essential for data collection. The questionnaire was designed based on a review of the literature using a similar latent construct to those proposed in the present research. The adapted questionnaire was reviewed by two academic experts, one in the field of research and the other in marketing, to ensure face and content validity. They evaluated the entire questionnaire for content relevance and language. Their suggestions such as minimising the demographic questions and rewording certain statements, were duly incorporated. Subsequently, a pilot test was conducted with 30 respondents to assess the reliability and accuracy of the questionnaire. Data from the pilot study were excluded from the final study. Based on the feedback received from this study, necessary modifications, such as the removal of ambiguous items, were made to ensure respondents' engagement in the survey. To facilitate a better understanding of the concept of street food, illustrative examples such as fast food, fixed meals, and snack items were provided. In the Google Form and the printed form, an image of a hawker selling food from a handcart was shown to help respondents visualise the context.

Close-ended questions were used in the questionnaire. The questionnaire comprised sections on the constructs and demographic details. 9 constructs and 46 items were used in the study. Street Food Purchase Intention (SFP) with 7 items; PFV with 5 items; PFQ with 5 items; PBC with 5 items; SUN with 6 items; Attitude (ATT) with 4 items; Convenience (CON) with 5 items were adapted from (Tacardon et al., 2023). Consumers' Lifestyle (COL) with 4 items, and HEC with 5 items were adapted from (Hoque et al., 2018). The language of the questionnaire was kept simple and lucid, to ensure that the respondents could easily understand, and complete the survey efficiently. An itemised five-point Likert scale ranging scale ranging from 1 = Strongly Disagree to 5 = Strongly Agree was used to measure the responses.

The study population comprised young consumers of street food in the Anand and Gandhinagar cities. Individuals aged 18 years and above, primarily in the age group of 18–24 years were targeted for the study. A non-probability sampling technique was employed in the absence of a sampling frame (Etikan et al., 2016). A single-cross-sectional research design was used to collect data from the respondents in a single time period (Creswell and Creswell, 2017).

The convenience sampling technique was adopted as the data were collected from the respondents willing to participate in the survey (Etikan et al., 2016). The convenience sampling is best suited for generalising the findings of a homogenous group of respondents (Jager et al., 2017). The present study consisted of a sample of those respondents who were consuming street food, so the findings of this study can be generalised to a homogenous group of street food consumers. This sampling technique minimally impacts relationships among the variables, and does not significantly impact the path estimates, while testing the TPB model (Hair et al., 2019a, b; Sarstedt et al., 2016). This sampling method is used because it is easy to implement, less time-intensive, and less expensive. But, the major limitation of the convenience sampling is that the findings of this study would be applicable only to a similar sample, as identical to the one used in the present study (Bornstein et al., 2013).

The online survey was conducted through Google Forms between December 2024 and March 2025. The questionnaire link was shared with various groups and was also sent directly to prospective respondents in Anand and Gandhinagar cities. It was ensured in the online survey, that if the respondents were consuming street food, then only they filled the form, otherwise an exit option was available to them. Along with the online survey, a pencil and paper-based survey was also administered for wider data collection (Hair et al., 2019a, b). Before circulating the questionnaire, respondents were asked whether they purchased the street food and they were willing to spend 10 min to complete the survey, justifying the convivence sampling method, despite its limitations. A total of 386 responses were collected from Gandhinagar and Anand cities, primarily from young respondents aged 18–24 years, with the majority being males. The online questionnaire targeted respondents with internet access.

Adherence to research ethics ensured that participants provided genuine responses and allowed the researchers to analyse the data objectively and publish findings (Dixon and Quirke, 2018). For this study, the minimum required sample size, was determined through power analysis using G*Power software, indicating a requirement of 262 responses based on a significance level of 0.05, a test power of 0.95, and a small effect size of 0.05 (Faul et al., 2009). The researchers distributed 500 questionnaires to ensure an adequate number of valid and useable responses. Of these, 90 participants did not return them despite multiple follow-ups, and 24 provided incomplete responses. Consequently, 386 fully completed questionnaires were obtained, resulting in a response rate of 77%.

Table 1 indicates the demographic profile of the respondents. Among them, 53% were male and 47% were female. The majority were post-graduates (52%). Regarding income, 40% earned less than ₹5000, while 34% reported income above ₹30,000.48% of the respondents stayed either at a hostel or as paying guests. All the respondents were street food consumers, and most of them (93%) preferred eating such food with companions or in groups. In terms of frequency of eating, the majority consumed street food once or twice a week (63%). 93% of the respondents ate either fast food, a fixed meal, or snacks.

The responses were assessed for normality using the WebPower analysis. The univariate skewness ranged between −3 and +3, and the univariate kurtosis ranged between −11 and +11. Mardia's multivariate skewness and kurtosis indicated a p-value of 0.00(<0.05). The Shapiro–Wilk normality test also confirmed that the data were not normally distributed. Given the relatively small sample size, non-normal data, and the study's objective to predict the Dependent Variable (DV) based on Independent Variable (IDVs), Partial Least Square-Structural Equation Modelling (Partial least squares-structural equation modelling (PLS-SEM)) was employed instead of Covariance-Based-Structural Equation Modelling (Sarstedt et al., 2016).

A random variable was introduced to check the potential response disengagement and to examine the presence of common method bias (CMB). The statistics for the same are presented in Table 2.

After executing the PLS-SEM algorithm, the inner Variance Inflation Factor (VIF) values were reported to be less than 3. To further confirm the absence of CMB, full collinearity assessment was conducted, which revealed that all inner VIF values were below the recommended threshold of 3.3 (Kock, 2015). Thus, the collected data were deemed suitable for assessing the street food purchase intention.

SEM is a multivariate data analysis technique (Ahmed et al., 2024) that integrates factor analysis and path analysis to simultaneously test relationships among latent and observed variables (Kline, 2016). The technique examines the associations between the items and the constructs, and validates the relationships among the constructs (Byrne, 2016). It is widely used in the marketing field to understand customer behaviour (Sarstedt et al., 2022). Tacardon et al. (2023) used SEM to study the impact of different latent variables on the street food purchase intention. In this study, the effect of IDVs such as PFQ, PFV, CON, PBC, SUN, ATT, COL, and HEC on SFP (DV) was studied.

It consisted of the Confirmatory Factor Analysis (CFA). Under CFA, the reliability, statistics, factor loadings, Average Variance Extracted (AVE), and discriminant analysis were performed. The results are presented in Table 3.

The items for the construct loadings and cross-loadings were also checked. Some of the items, such as PFV1, PFQ4, PBC1, PBC2, PBC4, ATT1, and SFP, were deleted due to strong inter-correlations of the items leading to a discriminant validity issue.

Internal consistency was demonstrated by Cronbach's alpha exceeding 0.70 for all constructs. Furthermore, the Rho A statistic also surpassed the 0.70 threshold. The Rho C value for each construct was greater than 0.80. These results confirmed that Cronbach's alpha and composite reliability values were within the acceptable parameters, as suggested by Nunnally and Bernstein (1994). Convergent validity was established, as the AVE for each construct was above 0.50, consistent with Hair et al. (2019a, b). All indicator loadings were found to be at or above 0.40.

To ensure constructs were distinct, discriminant validity was assessed, and the results are presented in Table 4. This evaluation aimed to confirm that the constructs were unrelated to each other. Comparisons within the same construct were categorised as monotrait, whereas comparisons between different constructs were classified as heterotrait.

According to Henseler et al. (2015), for distinct constructs, the Heterotrait-Monotrait (HTMT) ratio (HTMT) should not exceed 0.85. All calculated HTMT values remained below 0.85, indicating the absence of discriminant validity problems. A bootstrapping procedure, using 5,000 resamples, was employed to validate the HTMT ratios. The values presented in parentheses represent confidence intervals, showing the range of HTMT values obtained across the bootstrapped subsamples. Critically, none of these confidence intervals included the value 1, further supporting discriminant validity. The composite mean value was above 3, and the Standard Deviation (SD) value was above 0.5. The reliability, validity, and loadings all adhered to the benchmark values. Thus, the convergent and discriminant validity of the model was confirmed.

R-Square, effect magnitude, and the predictive power of the model were assessed through a structural model. The collinearity among the constructs was assessed through the Inner VIF values for all the IDVs.

As per Table 5, all the inner VIF values were less than 3, indicating the absence of a multicollinearity issue (Kock, 2015). A bootstrapping on 5,000 samples was run to estimate the path coefficients. The results are depicted in Figure 2.

The regression values were derived from the path coefficients, representing the relationships between one construct and another. All the path coefficients except for HEC to SFP were positive, indicating a direct relationship. A negative path coefficient between HEC and SFP suggests that as HEC increases, street food purchase intention decreases. Bootstrapping results revealed that the relationship between all the constructs (p-value) was less than 0.05, indicating a significant relationship. However, the relationship between PBC to SFP, and COL to SFP were not significant (p-values>0.05). Moreover, the bias-corrected confidence interval reported the presence of 0, confirming that the hypothesis was rejected and unsupported. All the hypothesised relationships in the model may be inferred from Table 6.

The results of IDVs explaining the DVs are in Table 7.

It may be inferred that the IDVs explain the DVs at least more than 40%. The values of the adjusted R-Square further confirm that the results of the R-square are not spurious.

As per Hair et al. (2019), an F-square value less than or equal to 0.02 but not exceeding 0.15 indicates a weak effect, while values between 0.15 and 0.35 represent a moderate effect. These interpretations are reflected in Table 8.

The Q-Square (cross-redundancy) values were reported to be 0.623 for SFP, 0.513 for PBC, 0.469 for SUN, and 0.623 for ATT, demonstrating that the model possesses a strong predictive power (Sharma et al., 2023). The value of the saturated Standardised Root Mean Square Residual was 0.068, which was below the benchmark value of 0.08, indicating a good model fit.

This study analysed the relationship between various IDVs and street food purchase intention by applying an extended TPB model. 14 hypotheses were tested using the SEM statistical tool. The research outcome presented significant support for 12 out of 14 hypotheses, offering practical insights into how COL, HEC, PFQ, PFV and CON influenced SFP both directly and indirectly via TPB constructs. The findings suggested that PFQ emerged as a strong predictor influencing the three primary TPB variables, i.e. ATT, SUN and PBC. It had a significant impact on ATT (β = 0.278, p = 0.000), SUN (β = 0.460, p = 0.000) and PBC (β = 0.362, p = 0.000). This confirmed that when consumers perceived street food as hygienic, safe and well-prepared, they had favourable attitudes, perceived social endorsement and felt more confident in their decision for street food. These results were consistent with earlier studies by Tacardon et al. (2023) and Ong et al. (2025), which emphasised the significance of food quality in determining cognitive and normative beliefs. Further, it was observed that PFV had a significant impact on SUN (β = 0.189, p = 0.021) and ATT (β = 0.287, p = 0.000) but did not significantly affect PBC (β = 0.121, p = 0.187). This showed that the value-for-money offerings influenced SUN and positive evaluation, but it did not increase the confidence of the consumers in having the street food. This result partially contradicts Tacardon et al. (2023), who found stronger links of food value with ATT and SUN. In the context of street food, value may be perceived differently and may not necessarily translate into a sense of control. CON had a significant impact on all three TPB antecedents, i.e. PBC (β = 0.294, p = 0.000), SUN (β = 0.165, p = 0.035), and ATT (β = 0.375, p = 0.000). These findings were in alignment with studies conducted by Gupta et al. (2018), Chowdhury (2023) and Naufal and Nelloh (2021), who reported convenience as a key determinant in food choice behaviour. However, as highlighted by Morano et al. (2018), eating such food may be risky due to factors like containment, preparation and cooking. Hence, consumers should not consider having street food as a replacement for home food on a regular basis. Among the three core TPB constructs, the results of the study revealed that the social influence (β = 0.421, p < 0.000) and personal attitude (β = 0.241, p < 0.010) were the most significant factors resulting in purchase intention for street food. On the other hand, PBC (β = 0.130, p < 0.073), i.e how easy or difficult it is for the consumers to buy such food, did not significantly influence their intention. Thus, the street vendors should focus on making such food socially appealing and its positive evaluation rather than just emphasising the accessibility or availability of street food. Binh and Mai (2023) and Didarloo et al. (2022) also identified SUN as having the highest impact on purchase intention. However, the findings of the study differed from those of Tacardon et al. (2023), where PBC also significantly influenced SFP.

One more interesting finding of the study is related to HEC, which negatively impacted SFP (β = −0.118, p = 0.046), indicating that health-conscious individuals may not like having street foods because of perceived health risks. The outcome of the study aligned with the findings of Kokthi et al. (2022), who found that awareness of the food affects the food choices of individuals. However, it differed from the findings of Hoque et al. (2018), who found that HEC did not influence the intention significantly. Furthermore, COL did not significantly influence SFP (β = 0.108, p = 0.220), suggesting that lifestyle choices of individuals may not directly translate into purchasing street food. It may often be driven by situational factors.

Thus, overall, the present study found SUN and ATT as the major contributors to SFP, especially in the Indian context for the younger generation. This finding could be identified with a few studies that have examined the food preferences of the same target audience in the Indian context. Studies by Beniwal and Mogra (2023) demonstrated that current social trends and peer influence significantly affected young consumers' food choices. Within the close social setting of a college, students often seek approval or match their choices with their peer group. Additionally, in this age, individuals often positively appreciate the taste, affordability, and convenience of street food, consistent with findings by Subhalakshmi and Dhanasekar (2018). These social preferences often take precedence over factors related to PBC, which in this study did not significantly affect the SFP. Further, in India, street food is easily available and affordable. Even if youngsters acknowledge the health risks associated with street food, they are more influenced by acceptance in their social circle and their favourable personal attitude.

The findings of this study may be interpreted with a few contextual boundaries. As the primary data for the study were collected using convenience sampling, the findings may not be generalisable to the entire population of all street food consumers or geographic regions. However, the study offers insights indicating street food purchase intention in urban settings. Moreover, as the study consisted of a homogenous group of street food consumers, the findings of this study can be easily generalised to a population having a comparable socio-economic context. The TPB model suggests that intention is driven mainly by cognitive evaluations of consuming street food, rather than demographic attributes, implying that the direction of relationships among lifestyle, HEC, and purchase intention is likely to remain stable across demographic dimensions including age and income groups, although its effect and magnitudes may vary (Ajzen, 1991).

This study proved that SEM is a powerful statistical tool for analysing the relationships among the predictors of street food purchase intention. The research applied the TPB framework and extended it with COL, HEC, PFQ, PFV and CON as contextual factors influencing SFP. The findings of the study suggested that SUN and ATT of an individual largely influenced SFP, whereas PBC did not significantly affect it. These findings challenged the conventional assumption of the TPB model, especially where situational factors dominate eating street food. Further, it was found that HEC negatively impacted SFP, suggesting that health-conscious individuals avoided eating street food. This finding may guide the street vendors' practices as well as public health interventions. COL did not impact SFP significantly. This was not in alignment with previous research on COL. The study also highlighted the importance of PFQ and PFV in affecting ATT and SUN significantly. This suggests that individuals gave higher preference to quality while forming perceptions about control and social pressure. Meanwhile, PFV did not significantly affect PBC, suggesting that consumers perceived value differently when convenience and quality were at play. The study confirmed the significant impact of CON on PBC, SUN and ATT. This could be interpreted as while quality and value were important for consumers, CON drove the actual consumption of street food. This reinforced that street food vendors need to maintain ease of access and quick service. This study extends the extant literature by highlighting the interplay between HEC, lifestyle and street food purchase intentions. It also adds depth to the theoretical understanding of the extended TPB framework, which can also be applied to other contexts in emerging markets, such as fast-food services. In the context of street food consumption, this study successfully integrates existing research that independently examined and validated the importance of quality, value, convenience, HEC and lifestyle.

This research offers practical implications for street food vendors, urban planners, and policy regulators in India. The strong influence of PFQ on purchase intention suggests that vendors must focus on safe food handling, visible cleanliness, use of fresh ingredients, and overall hygiene to win customer trust. Measures such as clean utensils, using gloves, covered food displays, and proper waste disposal can significantly impact consumer perception and intention towards purchase. The finding that convenience strongly influences intention, even in the presence of health risks, indicates that vendors should maintain accessibility without compromising hygiene and safety. There is a need to spread awareness about safer and healthier street food options due to the limited role of lifestyle and the negative association between health-conscious people and street food. This can help attract more health-aware customers. The government can support this by enforcing better hygiene rules, promoting efforts like “Clean Street Food Hubs,” and making the licensing process easier through digital tools. Urban Planners should also provide clean water, proper vending areas, and sanitation. Giving rewards to vendors who follow hygiene standards can motivate others to improve. These steps together can make street food safer and trusted by the public.

The study offers useful insights, but it has a few limitations too. It mainly targeted young respondents from the cities of Anand and Gandhinagar in Gujarat, which may fail to represent the wider street food consumer base across the country. Since the study used convenience sampling and collected data only once, the results may not apply to a wider population and do not show how behaviour might change over time. Respondents may also be inclined to give socially acceptable answers, giving a blurred picture. Future studies should aim to include a broader mix of age groups, regions, and income levels. Using longitudinal methods could help track shifts in behaviour, while tools like interviews or focus group discussions may provide a deeper understanding. Further exploration can include factors like trust in vendors, food safety knowledge, past behaviour, cultural background, sensory experience, and digital habits to better explain what drives people to choose street food (Zulmi and Suzianti, 2021). As the consumers often visit the same vendor to buy the street food, even the street food patronage behaviour could be the scope for future study (Esiti and Ayodele, 2025).

Apart from the TPB model, other theoretical frameworks can be integrated or applied separately to study the intention of street food consumption. The Health Belief Model and Protection Motivation Theory, may be used in examining the role of perceived health risks and coping responses in the intention to consume street food (Rosenstock, 1974; Rogers, 1983). Further, Consumer Culture Theory may also be considered to study how street food consumption is shaped by various local cultural dimensions rather than purely rational decisions (Arnould and Thompson, 2005).

This study developed an extended TPB framework and analysed the complex relationship of core domains on SFP in addition to COL and HEC. It also checked the indirect effects of PFQ, PFV and CON on SFP. It was found that PFQ of the street food had the highest impact on the core constructs, suggesting that when the street food consumers of Anand and Gandhinagar perceived, such food as safe and hygienically prepared, they developed a favourable attitude, felt socially validated and gained confidence in their purchase decision. The study also revealed that perceived value did not necessarily give a sense of control but influenced social acceptance and positive evaluation. Convenience has also emerged as an important driver of food choices despite the potential health risks.

The study highlighted an interesting fact that HEC showed a negative relationship with SFP, suggesting that such consumers were less inclined to have street food due to health concerns. Further, it was revealed that COL did not affect the SFP significantly, suggesting that situational factors were more important in driving the choices of young consumers. Among the core TPB constructs, PBC had a limited impact as compared to SUN and ATT. These findings throw light on the socio-cultural context found in Anand and Gandhinagar cities, where the vendors need to focus on social appeal and positive evaluations to increase penetration among young consumers. In India, street food vendors are found everywhere. Hence, rather than availability, quality is of higher concern. If vendors focus on improving quality, perceived risk factors related to health can be ensured, including maintenance of hygiene standards, sourcing fresh ingredients, safe food handling practices, proper waste management, and display quality certification, etc. Thus, the study serves as a reference point for researchers to build on and extend beyond food and beverage consumption.