Data collection and sampling

This cross-sectional, descriptive and quantitative study involved 303 street food vendors in Southern Brazil. Data was collected between March and December of 2018 in: Pelotas, Porto Alegre, Rio Grande, São Lourenço do Sul and Capão da Canoa (Rio Grande do Sul state), Canasvieiras (Santa Catarina state) and Ubatuba (São Paulo state). The recruitment criteria for food handlers was to be over 18 years old, to be involved in food handling and to work as a street food vendor, classified by the International Food Safety Network (INFOSAN) as a stationery or ambulant business⁴. A total of 388 street food vendors were approached, 85 (21.9%) of them refused to participate in the study. Food sold by street food vendors included hot-sandwiches (such as hotdogs, toasts, baurus), churrasquinho (meat skewer), corn-on-the-cobs, popcorn, crepe (a type of pancake made from wheat flour, milk and eggs), churros (tube-shaped fried doughs), cotton candy, traditional sweets, etc. Considering the absence of fixed addresses for people working in the informal sector, the street food vendors comprised a convenience sample.

A questionnaire based on the predictors of the Theory of Planned Behaviour (TPB), including an additional predictor, Optimistic bias (OB), was applied face-to-face to identify the intention of vendors to perform safe food handling behaviours. Trained interviewers administered the questionnaire. Street food handlers were invited to sign an informed consent form before participating in this study. The research was approved by Ethics Committee in Research with Human Beings of the Federal University of Pelotas (UFPel) (No. 2.673.491).

Questionnaire based on Theory of Planned Behaviour (TPB)

The questionnaire was developed to this study and comprised demographic questions, open-ended questions and questions based on the TPB, based on Ajzen et al.²¹, Clayton et al.¹³ and Weinsten¹⁷.

Demographic question included sex, age, time working with street food, educational level and training in food hygiene.

Open-ended questions aimed to assess street food vendors’ beliefs about which actions were important in safe food handling. Street food vendors were asked to describe their hygiene behaviours during food handling; to identify consequences (advantages or disadvantages of these behaviours); possible facilitators or barriers to these behaviours; and to identify sources of social pressure to perform these behaviours.

Questions based on the TPB were formulated with questions of direct measurement of TPB predictors: perceived behavioural control, subjective norm and attitude and optimistic bias. All responses were on a 7-point Likert scale, with ratings of 1 to 3 considered as “disagree/negative”, 5 to 7 considered as “agree/positive” and 4 as neutral.

Perceived Behavioural Control (PBC) was phrased as: For you to adequately carry out appropriate food safety behaviours at every appropriate occasion during your working day is: (+1 extremely difficult and +7 extremely easy); and How much control do you have over whether or not you carry out appropriate food safety behaviours at every appropriate occasion during your working day: (+1 no control and +7 complete control).

The Subjective Norm (SN) statement was phrased as: Most people who are important to you think you should carry out appropriate food safety behaviours at every appropriate occasion during your working day: (+1 strongly disagree and +7 strongly agree).

Attitude (A) was measured using a direct measure with three different response scales. Attitude was phrased as: Carrying out appropriate food safety behaviours at every appropriate occasion during your working day is: (+1 very bad and +7 very good); (+1 worthless and +7 worthwhile); (+1 negative and +7 positive).

Behavioural intention (BI) was phrased as: In the next week you intend to carry out appropriate food safety behaviours at every appropriate occasion during your working day: (+1 strongly disagree and +7 strongly agree).

To evaluate Optimistic Bias (OB) in the context of carrying out food safety behaviours, statements of the following formats were used: What is the likelihood that you carry out appropriated food safety behaviours at every appropriate occasion during your working day: (+1 very unlikely and +7 very likely). What is the likelihood that other people would carry out appropriate food safety behaviours at every appropriate occasion during their working day: (+1 very unlikely and +7 very likely). The difference in score between these two measures constitutes the measure of OB.

All questions were pilot tested among 15 street food handlers and the initial questionnaire was revised and perfected. Cronbach’s α, a test to estimate the reliability of a questionnaire, was 0.707, which is considered acceptable²².

Data analysis

Questionnaire responses were coded and data entered into the Statistical Package for Social Sciences, (SPSS) Version 18. Descriptive statistics were employed to examine the characteristics of respondents’ overall responses to demographic questions, and Chi-square to compare frequencies between groups. Qualitative data from open-ended questions were evaluated by content analysis, defined as “a research technique for making replicable and valid inferences from texts or other meaningful matter to the contexts of their use”²³. Responses to the questions based on the TPB were analysed using descriptive statistics and correlations between TPB components tested the assumptions of the model. Multiple linear regression was used to identify the main drivers of intention.

Demographic data

Table 1 presents the demographic characteristics and training participation rate of 303 street food vendors whom completed the questionnaire. The majority of street food vendors were female (56.1%), had high school-level education (45.5%) and most reported that they had training in food hygiene (77.9%), with an average time of 20.4 (SD= 34.6) months since they were last trained. The time working as a street food vendor varied between one and 732 months (M= 123.5; SD= 29.7).

Open-ended questions

Table 2 shows results of open-ended questions to assess street food vendors’ beliefs about which actions are important in safe food handling. Street food vendors described their hygiene behaviours during food handling, advantages and disadvantages related with these behaviours, possible facilitators and barriers to carrying out food safety behaviours and whom would approve or not of their behaviours.

Handwashing (57.7%), cleaning the shelf they used to prepare food (40.5%), food temperature control (38.9%) and using alcohol hands gel (36.9%) were the hygiene behaviours described by street vendors. Advantages and disadvantages related to consumers, barriers and facilitators related to water supply and toilets nearby were identified as important factors influencing street food vendors’ behaviour. When food handlers did not have water supply and toilets nearby it was difficult for them to perform food safety behaviours such as handwashing and cleaning the shelf they used to prepare food. Consequences related to consumers could be a motivating factor to perform adequate food safety behaviours. Also, consumers were the main result related to source of social pressure to carry out food safety behaviours. Training and education level were found to impact on street food vendors’ recognition of advantages and disadvantages, noted barriers and facilitators, and the source of social pressure to carry out these behaviours.

TPB based questions

Table 3 shows the results of street food vendors’ intention to perform food safety behaviours. The results showed that the majority of street food vendors had a positive perceived behavioural control, subjective norm, attitude and behavioural intention. When OB was evaluated, 99% of street food vendors had a positive response of their own likelihood of carrying out appropriate food safety behaviour each working day (M= 6.77, SD= 0.58). However, when questioned about other street food vendors likelihood of carrying out appropriate food safety behaviours, they believed (62.1%) that other vendors were less likely to carry out the same behaviour (M= 4.89, SD= 1.91), confirming OB (p<0.001).

No correlation was found between TBP components, thus the assumptions of the TPB model were not confirmed in this study (adj r2: 0.004, p= 0.284). Perceived behavioural control (β= 1.46, p= 0.93), subjective norm (β= 0.076, p= 0.145), attitude (β= 0.005, p= 0.192) and OB (β= 0.044, p= 0.45) did not significantly influence behavioural intention.

Demographic data

The majority of street food vendors were female, which is consistent with previous studies^7,8,24,25. Desire for an autonomous income that is flexible with their household responsibilities can be associated with the majority of vendors being women in the informal sector²⁵. Usually street food vendors have elementary school as the highest educational level achieved^7,8,24,25, however in this study it was found that the majority of respondents had high school-level education, which may be associated with increased unemployment in the formal work sector. It is known that during economic crises the informal work sector increases due to a reduction in formal employment, therefore, people with higher educational levels look for a new income in the informal work sector²⁶. Street food sector plays a critical role as employment opportunities due to the low initial investment and to the low skills required to start working, where unemployed populations turn to it as a means to earn a living^1,27. Furthermore, food handlers with higher education levels have also been associated with better hygiene behaviours during the working day at food services^28,29,30.

In relation to food hygiene training, Brazilian legislation requires training in food hygiene for people that are involved in food handling¹², which may explain the high percentage of street food vendors with training in this study. Although results suggest compliance with Brazilian legislation, the average time since street food vendors had trained in food hygiene was more than one year and a half, on average. In Rio Grande do Sul state, Southern Brazil, where data were collected, the state legislation demands that food handlers must have annual training in food hygiene³¹. Refresher training is essential to prevent food safety failures that can be a result of poor knowledge and food safety practices of food handlers³².

Other studies carried out in Brazil have shown a high percentage of street food vendors with formal training^11,33. On the other hand, studies carried out in other countries such as India⁵, China⁶, Haiti⁷ and Vietnam⁸, where training in food safety is not mandatory, have found that the majority of food handlers do not have training. Although the average time between trainings was not adequate, the number of trained people is higher in Brazil when compared to other developing countries. However, training and education does not always result in adequate food safety behaviour. Moreaux et al.³⁴ investigated the food safety attitude and behaviour gap and results showed that formal education and food safety training did not bridge this gap, but there was a link between consumers’ dissatisfaction and behaviour.

Open-ended questions

Advantages and disadvantages related to consumers were important factors influencing behaviours of street food vendors. Gustatory attributes, affordable and accessible food are the main reasons reported by consumers for purchasing street food in previous studies^4,8,35,36, but food hygiene has also been reported by consumers as a concern when purchasing street food. Asiegbu et al.³⁵ results showed 73% of 402 consumers were either always or often concern about bacteriological safety of street food, 53% and 63% strongly disagree or disagree, respectively, with statements that “they did not always think about food safety” and “they were certain they could not get foodborne diseases” when they ate street food. An important factor that might justify the concerns of street food vendors related to consumer pressure to maintain food hygiene is that consumer opinions are now widely accessible with online reviews, where they transmit positive and negative information about their experiences with street food³⁷.

“Produce quality food” and “food will not cause foodborne disease” were also reported as an advantage of performing food safety behaviours. If street food vendors perceived that producing quality food and that food will not cause foodborne disease, might lead them to get more consumers and improve consumer satisfaction, they may be influenced to give more attention to producing safe food.

Water supply and toilets nearby were identified by street food vendors as facilitators and barriers to carrying out food safety behaviours. It was previously reported that the street food sector has a lack of potable water, toilets, waste disposal and general infrastructure, affecting the performance of food safety behaviours^1,9. Kothe et al.³³ found only one of 20 street food point-of-sale outlets with potable water in Porto Alegre and 100% of street food vendors did not follow correct hand hygiene practices. Auad et al.³⁸ found lack of water supply in 11 of 40 street food trucks in Federal District of Brazil, it was also reported by them that 55% of street food vendors had training in food hygiene and questions related to handwashing knowledge had the highest success rate between vendors, but, even with training and handwashing knowledge, when hygiene good practices were observed they found inadequacy related to handwashing.

Handwashing was identified as a food safety behaviour performed during the workday of street food vendors, but if there is no water supply and toilets/facilities nearby, they cannot perform handwashing properly. Adequate handwashing is extremely important since hands are responsible for spreading microorganisms to different places leading to an increased incidence of contaminated food, potentially causing foodborne diseases^1,4. Without proper facilities, street food handlers become likely vectors of food contamination¹¹. If water supply and toilets/facilities nearby were available to street food vendors it would make handwashing and cleaning the materials used to prepare food easier, they would have better control of that behaviours described as being performed by them during their workday.

Another barrier reported by street food vendors was the weather (e.g., heat, rain, wind, air pollution). Many street food vendors handle food in unsafe environmental conditions, were food can become contaminated by the open space where food handling happens due to dust, pollution and animals/pests, whereby pathogenic microorganisms may be in contact with street food^1,3,4,10. Despite noted barriers, it was also found that 27.3% of street food vendors believed that there was nothing that would make it easier for them to perform food safety behaviours. If street food handlers do not perceive the lack of water supply, not having toiles/facilities nearby and the weather as a barrier in street food sector, they can fail to adopt protective behaviours to avoid contamination.

Consumers were reported by great majority of street food vendors as individuals who would approve of them carrying out safe food handling practices. As previously discussed, street food vendors recognized consumers’ importance in the success of their business, since “getting more consumers” and “losing consumers” were reported as an advantage of performing food safety behaviours and a disadvantageous consequence of poor food safety behaviour, respectively. Auad et al.³⁸ interviewed 133 food truck consumers; 78.2% and 83.4% of them always consider food hygiene and street food vendor personal hygiene when eating from food trucks, respectively. They also point out that having a hand sink with hand soap and paper towels available for the food handlers for handwashing is extremely important. Consumers concern about food and food handler hygiene and proper facilities to perform food safety behaviours could be used as a motivating factor to street food vendors to perform these behaviours.

Sanitary surveillance was mentioned by only 16.8% of street food vendors as a system that would approve their food safety behaviours. Food services in Brazil are regulated by sanitary legislation¹², however the street food sector has fewer facilities and accessibilities (lack of water supply, proper facilities, food handling environment) when compared to restaurants and other food services. Previous studies have proposed improvements in governmental regulation and sanitary surveillance, with rigid supervision and frequent inspections, as well as support for providing food safety training, as a way to encourage good behaviours among street food vendors^11,33. A specific law for street food vendors that address food safety considering particularities vendors face during their work day could also assist performance of food safety behaviours.

TPB predictor-based questions

When assessing perceived behavioural control of food handlers, it was found that most of them believed that it was easy for them to perform food safety behaviours and that they have control of food safety behaviours. Food handlers must be aware of barriers and the real control they have around food safety behaviours in the street food sector. If they believe they have more control over a behaviour than they actually have, this belief can lead them to be more optimistic about their real chances of avoiding an unwanted result and reaching a desired result. Food handlers do not perceive risk if they are confident that they are controlling it^39,40.

The majority of street food vendors had a positive subjective norm, a predictor that assesses perceived social pressure to perform or not a behaviour. According to Hofstede⁴¹ and Hofstede et al.⁴² in the National Culture model, Brazilian society has a power distance index of 69, which reflects a society that believes hierarchy should be respected. Individuals feel less inclined to act according to their personal preferences and attitudes, as well as being concerned to comply with the opinions of others⁴³. As previously discussed, street food vendors pointed to consumers as a group who would be interested in their food hygiene related behaviours, and pointed to “getting more consumers” as an advantage of performing good hygiene practices. Consumer satisfaction and opinion could be used as a motivating factor in food training for street food vendors to improve their food safety behaviours.

Behavioural intention is used to evaluate how hard people are willing to try and how much effort they plan to exert to perform a certain behaviour. Individual intention to perform a behaviour is assumed to be directly related to actual behaviour¹⁵. The results showed that 99.3% of street food vendors strongly agreed that they intended to perform food safety behaviours. Results also showed a positive attitude from majority of street food vendors towards food safety behaviour, they believed that carrying out food safety behaviours during their workday to be very good, worthwhile and positive. This positive attitude showed an optimistic view by street food vendors. Brazilian society is classified as a collectivist group^41,42, which leads to high optimism because reaching a goal does not depend only on one person, but depends on collective work. For this reason, to perform a behaviour, it is necessary to believe that results will be positive, even if it does not depend on only one person. In addition, people usually overstate their own importance and their own control over an event and this, in part, may underlie OB^39,40. Even if there is a poor infrastructure and weak support by the Brazilian government, street food handlers have a positive attitude towards performing food safety behaviours, resulting in a positive view in relation to food safety behaviour performance⁴⁴.

OB was identified among street food vendors and has previously been reported in Brazilian food handlers by Da Cunha et al.¹⁴, Rodrigues et al.¹⁶, Rossi et al.¹⁸, Da Cunha et al.⁴⁵, and Andrade et al.⁴⁶. Since food handlers believe that others are worse than themselves and bad things will not happen to them, they can, again, overlook some protective behaviours, increasing the risk of contaminating food. People perceive a personal invulnerability, but fail to extended this perception of invulnerability to others, believing that negative events that happen to others will not happen to them, leading them to underestimate their personal risk, and thus involving themselves in risky behaviours^17,47.

Da Cunha et al.¹⁴ reported OB in several food services in Brazil, were food handlers believe to be less likely to spread foodborne diseases to consumers and family and/or friends than their peers. Persons working in food services, food handlers working in restaurants, hospitals and schools had a higher OB when compared to food handlers whom work with street food. Da Cunha et al.⁴⁵ also identified OB about food business risk, in that food handlers believe that their colleagues are less likely to cause foodborne diseases than food handlers from another food service such as street food sector. These results can be associated with environmental empowerment, since street food is commonly associated with poor quality due to the environment and perceived inappropriate food handling, it can provide a positive outlook regarding risks when food handlers from other food services compare themselves and their peers with street food vendors^5,14. Rodrigues et al.¹⁶ were the first to include OB as a predictor in the TBP. The study with foodservice operators from the UK and Brazil found OB in both countries, but only in the UK was OB a significant predictor to intention to perform food safety behaviours. Also, perceived behavioural control was not significant to intention in Brazil, showing that Brazilians are less likely to perceive external barriers and to perform food safety behaviours to prevent these barriers.

Food handlers with OB choose and decide to take certain risks and avoid others. Beliefs, knowledge, experience and other characteristics can induce this choose¹⁴. Food handlers in the street food sector usually have inadequate sanitary facilities and the food is often exposed to possible contaminants due to environment were food is handled^1,5. These conditions can lead street food vendors with OB to decide which risks to take and which ones he/she will avoid, however, to neglect any protectives measures in street food handling environment is dangerous, increasing foodborne diseases risk. As previously discussed, even with noted barriers beyond the control of street vendors, they believed they had control and it was easy for them to perform food safety behaviours. The belief of having more control of events than they really have leads to OB and, consequently, to neglectful attitudes and food safety practices^14,39,47.

No correlation was found between TBP components. This may be the result of the highly positive responses to all TPB-related questions. The optimism of Brazilians and the importance they give to the opinions of others, may be related to social desirability bias, and may account for the consistently high scores of analysed items, which in turn may account for the lack of correlation found between TBP components. Social desirability bias is the tendency of research subjects to give answers they believe are more socially desirable or culturally acceptable, rather than give answers that express their real opinion about the topic⁴⁸. Since the Brazilian power distance index typifies individuals who are concerned with complying with the opinions of others^41,42,43, this may be related with them giving the response they believe would be the more acceptable, instead of the reality they face during their workday.

Providing training that highlights the risk and consequences related to OB, positive perceived behavioural control, subjective norm and attitude, and focusing on desirable consequences related to consumers is a possible way to promote food safety behaviour in the street food sector.As previously reported by Moreaux et al.³⁴ there is a gap between food safety attitudes and behaviours that was not bridged by formal education and food safety training, but there was a link between consumer dissatisfaction and behaviour. Desirable consequences can positively influence street food vendors attitudes. As identified by some street vendors, if they realise that food safety behaviours lead them to “get more consumers” and “consumer satisfaction” they can give more attention to food safety behaviours. Including these possible consequences in food safety training may act as a behavioural motivator. Training may be a way to reduce OB, and thus improve good practices, but is important to note that food safety cannot be achieved if the basic infrastructure required to perform food safety behaviours is not available. In addition, it is important to have retraining or refresher training about food safety in order to maintain food handlers’ knowledge and behaviours. Unless these barriers are addressed, we will continue to find gaps in food safety in the street food sector.

Limitations

The study highlighted challenges due to street food vendors’ limited time to respond and their understanding of the questions, although questions were piloted to maximise understanding. The questionnaire form of interview was chosen in order to obtain more accuracy, more complete questionnaires and to interfere as little as possible with the working day of street food vendors. There is a possible limitation in the use of face-to-face interviews in the form of social desirability bias. This is the tendency of research subjects to answer in a way that he/she believes to be more socially acceptable, interfering in behaviour-related reports of street food vendors. Also, the study did not cover all Brazilian regions, other regions might have different results. In-depth qualitative approaches may also have exposed additional beliefs and motivations, but would have required vendors to sacrifice more of their time.