Effects of Instagram food posts on the intention to eat healthy and the role
of attitude towards food and perceived social norms
Elisabeth Annetta Margaretha Acket Student number: 12313564
Master Thesis
Communication Science: Persuasive Communication Supervisor: Dr. Sanne Schinkel
Word count: 8250
University of Amsterdam 28. June 2019
Abstract
Previous research on social modelling indicates that people’s food choices and the amount of food they consume can be influenced by other people or by cues from others. Although a significant amount of research has been done on social modelling of eating, research on this phenomenon exists only in an offline context. Applying the social modelling effect to the social media platform Instagram, this study investigates to what extent Instagram pictures portraying food influence the intention to eat healthy, and what role attitudes towards food and perceived social norms about food play in this regard. To test the moderated
mediation model, a 2 x 2 factorial between-subjects design was used, with healthy vs.
unhealthy Instagram food posts and healthy vs. unhealthy social norms. An online experiment with 239 participants revealed marginally significant results of the Instagram post on attitudes towards food. It appears that Instagram pictures portraying food do not directly influence someone’s intention to eat healthy. Nonetheless, the perceived positive social norms that a person holds on attitudes towards food were found to have a significant positive effect, and those positive attitudes towards food were revealed to have a positive effect on the intention to eat healthy. Further research should be conducted to investigate the possible influence that social media platforms such as Instagram could have in social modelling of eating or in creating a food norm for a healthier society.
Introduction
Nutrition is a key factor in living a healthy life. Unhealthy nutrition, for example, a diet rich in fats and sugar, correlates significantly with rising rates of obesity along with multiple health risks such as cardiovascular diseases, diabetes, and cancers (World Health Organization, 2018b). Overweight and obesity are defined as an “abnormal or excessive fat accumulation that may impair health” (World Health Organization, 2018b). Nearly 40 percent of the world’s adult population are overweight, of whom 13% are described as obese.
Worldwide, the number of people with obesity has tripled since 1975 (World Health
Organization, 2018b). Obesity has become one of the greatest health problems our society is facing today (Kuzbicka & Rachoń, 2013), which also carries economic and societal
consequences. Next to direct medical costs are indirect costs derived from disease rates, mortality costs, and loss in productivity costs. Compared to non-obese workers, obese workers demonstrably miss more working days due to illness, injury or disability. Indirect costs caused by obesity can range from $450 million in Switzerland to nearly $65 billion in the United States (Trogdon, Finkelstein, Hylands, Dellea & Kamal-Bahl, 2008). No doubt, change is needed.
The World Health Organization (2018b) claims that overweight and obesity are preventable individually. People can reduce their intake of fats and sugar; increase their consumption of fruits, vegetables, legumes, whole grains, and nuts; and engage in regular physical activity (World Health Organization, 2018b). In regions where sugar and fat intake is high, people’s dietary habits must be changed to offer them a first step towards living a
healthier lifestyle (World Health Organization, 2018b).
Multiple unsuccessful educational interventions have been attempted to broadly change dietary behaviour among children and the general population. One reason for those failures is that eating is a social practice; most interventions were knowledge-based
Ly, 2013). Obesity has been demonstrated to spread through social ties and over social networks (Christakis & Fowler, 2007). Food and eating are connected to the social environment of individuals (Cruwys, Bevelander & Herman, 2015). In the past, these
influences were limited to offline influences from family, friends, and culture. However, with the rise of social media platforms such as Instagram, people are exposed to a great deal of social information online about food and eating behaviour (Hefner & Vorderer, 2016). The effect of social modelling of eating has been studied extensively in multiple offline scenarios, but these results should be validated in an online environment (Vartanian, Spanos, Herman & Polivy, 2015; Maheshwari & Koblin, 2018). Further research is required on the complex factors that influence food choices online, as well as how they most effectively spread into society.
The social media platform Instagram is the largest picture-sharing platform in the world, with 1 billion users, and therefore one of the most popular (Statista, 2018). This
platform is also well-known to many users who like to share their food preferences and eating habits (Vaterlaus, Patten, Roche & Young, 2015). Food and eating experiences are connected to our social lives and therefore are shared on Instagram (Holmberg, Chaplin, Hillman & Berg, 2016). Researchers have used the platform to monitor the connection between posts of individuals at fast-food restaurants, obesity, and the perception of people’s eating habits, but these studies used Instagram only as a tool to find the locations of restaurants and connect these to obesity rates and people’s perception of the food (Mejova, Haddadi, Noulas & Weber, 2015). Additionally, studies have investigated the food-related content on Instagram (Holmberg et al., 2016; Hu, Manikonda & Kamghampati, 2014). More than 10 percent of all photographs posted on Instagram are related to food (Hu et al., 2014). Food is represented in various ways on Instagram but a significant number of the posts are of unhealthy food. Most of the food-related pictures are accompanied with hashtags that express the palatability of the food (Holmberg et al., 2016). Interviews with young adults indicated that social media
influence their health behaviour, but little is known about what effect an Instagram post portraying food might have on eating choices and health outcomes (Vaterlaus et al., 2015). Because of the high levels of engagement and the use of visual content, which is processed differently and stays longer in people’s memories, Instagram appears to be an ideal platform to reach people (Covello, 2003). It has been proven particularly useful in establishing meaningful and interactive communication with the public during a global health crisis (Guidry, Jin, Orr, Massner & Meganch, 2017). Therefore, the influence of Instagram food posts on people’s food choices deserve to be studied.
This study thus focuses on the influence that Instagram can have on a person’s eating habits. The theory of social modelling includes this social aspect of eating behaviour and indicates that people are influenced by other people’s eating habits. The behaviour of a person is explained along the theory of planned behaviour; it is based on the theory of reasoned action and describes that intention is a key determinant for behaviour (Ajzen, 1991).
Therefore, is it important to first create an intention that, in turn, will lead to behaviour. Two major determinants for intention are attitudes and perceived social norms (Ajzen, 1991). Attitudes explain a person’s preference for certain behaviour; norms are a social pressure that derives from expected and performed behaviour by someone’s peers. Thus, these two aspects are expected to influence how Instagram food posts affect the intention to eat healthy.
This study attempts to determine the extent to which Instagram posts portraying food (healthy vs. unhealthy) can affect dietary intentions and to determine the role played by perceived social norms and attitudes towards food. The findings provide new insights into the use of Instagram and examine the effect of social modelling in an online environment.
Additionally, this study provides a new viewpoint on perceived social norms and attitudes towards food. The findings can be used by communication managers to target the population in an effective way to create healthier food intentions. These intentions could lead to healthier
behaviour and an even healthier society by reducing obesity. Consequently, the following research question is formulated for this study:
How do Instagram posts portraying healthy and unhealthy food influence the intention to eat healthy, and what role do attitudes towards food and perceived social norms about food have in this effect?
Theoretical Framework Social modelling and Instagram
Food and dietary choices are strongly connected to our social environment. In an offline context, social connections with food have been studied extensively and reveal that the amount of food we eat can be strongly influenced by the people surrounding us (Higgs & Thomas, 2016). When someone in a group eats a lot, the others in that group tend to increase their food intake as well. On the other hand, when another person eats very little, people in such a group tend to eat smaller quantities of food. This effect is known as social modelling and is considered to be a resilient phenomenon that can influence what and how much people actually eat (Cruwys et al., 2015; Vartanian et al., 2015). The appropriate amount to eat is generally unclear, so most people rely on others to determine how much to eat themselves (Herman, Roth & Polivy, 2003; Vartanian et al., 2015). Social modelling happens when specific behaviour is performed by others, even if no person is present and there are only environmental cues or there is textual information (Vartanian et al., 2015; Higgs &Thomas, 2016).
Most of the studies regarding social modelling examine the amount of food consumed. Nonetheless, Prinsen, Ridder, and Vet (2013) found that environmental cues can additionally influence the food choices of a person. A selection of healthy and unhealthy snack bars was presented to a study participant, with three empty wrappers left by previous participants in sight. As expected, each participant chose the snack that was coherent with the choice of
previous subjects, suggesting that social modelling has an effect on food choice, even when no person is physically present.
Although most of the studies on social modelling and food portray unhealthy foods, a modelling effect has been found in healthy foods as well. Salvy, Kieffer, and Epstein (2008) found that children who were presented with healthy and unhealthy food options were influenced by others present in the room. When another person ate a healthy food snack, the youthful participants chose the same. Overall, multiple studies suggest that there is general support for the observation that social modelling shapes people’s food intake as well as healthy or unhealthy food choices (Cruwys et al., 2015). Since most of the studies were performed in an offline environment, it is unclear whether these results would be replicated online. As social modelling of eating has been found to influence people’s food choices, it is plausible that the exposure to a post on Instagram might influence a person’s intention to eat similarly, since such a social media platform creates a space where users are able to follow friends, peers, and people they admire (Cruwys et al., 2015). Social interaction, keeping up with peers, and seeing what other people have to say with pictures are the main motivations to use Instagram (Lee, Lee, Moon & Sung, 2015). Therefore, this platform can mimic the
influence of social modelling in an offline context and can lead to people to adjust their food choices to the ones of others. Seeing a healthy food picture that a peer has posted on
Instagram can influence a person to eat similarly, in other words, healthy, as well.
A majority of the studies regarding social modelling measured food choice as their behavioural outcome. As mentioned earlier, intention is a key determinant for actual behaviour (Ajzen, 1991), therefore, it is plausible to assume that social modelling can influence someone’s intention to eat healthy.
Consequently, this study proposes the first hypothesis as follows: Exposure to an Instagram post portraying healthy food will lead to a greater intention to eat healthy than exposure to an Instagram post portraying unhealthy food (H1).
Attitudes towards food
Another important feature of the social aspects of food is the attitude towards food. Ajzen (1991, p. 188) defines attitude as “the degree to which a person has a favourable or unfavourable evaluation of the behaviour in question”. Little research has been conducted on the influence that Instagram can have on attitudes towards food. Nonetheless, some research has examined the portrayal of junk food advertisements on television. Dixon, Scully,
Wakefield, White and Crawford (2007) discovered that the exposure to only unhealthy junk food on television led to more positive attitudes towards junk food. In a study by Dixon et al. (2007), heavy television watchers perceived that consumption of unhealthy foods was
prevalent among peers with whom they identified. This regular reminder on television normalised the consumption of unhealthy foods, leading to more positive attitudes towards the junk food portrayed. Additionally, the researchers examined whether this effect was similar when only healthy food advertisements were aired on television. An even stronger effect occurred for healthy food: Exposed to a healthy food advertisement, people had more positive attitudes towards healthy food. These findings are in line with the cultivation theory, which claims that people who watch a lot of television are inclined to have beliefs similar to those of the television characters (Gerbner, Gross, Morgan & Signorielli, 1994).
Given that this strong influence on attitudes towards food was triggered by television exposure, it can be assumed that similar influence will occur from the exposure to social media platforms such as Instagram. Since hours devoted to television-watching, especially among younger audiences, have decreased over the past decade with the rise of social media platforms, it is expected that the consumption of photographs and videos on Instagram can influence people’s intention to eat healthy or unhealthy in a similar manner in which television did (Maheshwari & Koblin, 2018).
Neuroscience research indicates that appetite-related areas in the brain activity can be affected after one sees a picture of food (Beaver, Lawrence, van Ditzhuijzen, Davis, Woods &
Calder, 2006). Since appetite is related to taste, and taste is one of the strongest predictors of attitudes towards food, this might explain why a picture can influence someone’s attitudes towards food (Roininen, Lähteenmäki & Tuorila 1999; Roininen & Tuorila, 1999). Therefore, it is expected that the portrayal of healthy food instead of unhealthy food on Instagram could lead to a more positive attitude towards healthy food.
The fact that anyone can create, comment, like, and share on social media makes these platforms highly interactive in contrast to television, where a programme is less interactive. The possibility of interaction on Instagram increases the scope for influence on subjects by peers and therefore might have an even stronger effect than television advertisements. Seeing what peers post and experiencing the full dialogues of all users of the platform can feel closer and more realistic to a platform user than a television programme does.
Multiple studies have demonstrated that a positive and therefore healthy attitude towards food influences people’s eating habits such that people have a healthier diet and make healthier food choices than people with a negative attitude towards healthy food do (Lê et al., 2013; Cooke & Papadaki, 2014; Hearty, McCarthy, Kearney & Gibney, 2007). As all these studies examined health behaviour or food choices as their outcomes, and the theory of planned behaviour claims intention is a determinant of behaviour, it is expected that positive attitudes towards food will lead to a higher intention to eat healthy (Ajzen, 1991; Fishbein & Capella, 2006).
Considering the above, it is expected that attitudes towards food will play a mediating role in the influence of Instagram posts on the intention to eat healthy. Therefore, the
following is assumed: The effect of Instagram posts portraying food (healthy versus
unhealthy) on the intention to eat healthy is mediated by attitudes towards food, in a way that healthy posts lead to more positive attitudes towards food (H2), which leads to a higher intention to eat healthy (H3).
Social norms
Social norms are an important determinant also used in the theory of planned
behaviour to understand intention (Ajzen, 1991). They are described as forms of boundaries set by the social environment to guide intention and behaviour (Ajzen, 1991; Cialdini & Trost, 1998). In literature, two different types of norms are distinguished: descriptive and injunctive norms (Cialdini, Reno, & Kallgren 1990; Cialdini, Kallgren & Reno, 1991). Descriptive norms are derived from what other people do, and injunctive norms specify what other people approve or disapprove of and therefore demonstrate what should be done (Cialdini et al., 1991; Cialdini & Trost, 1998). In the context of food, when a food norm is regularly communicated in a person’s surroundings, this could imply that such behaviour is socially expected and therefore creates a social norm that guides and influences someone’s intention. Social norms can thus influence people’s intention to behave a certain way or not.
Considerable research has been done in the field of the manipulation of social norm messages and the influence that these can have on people’s eating choices. Two experiments that attempted to manipulate participants with social norm messages into eating either burgers or salad and fruit had differing results (Mollen, Rimal, Ruiter & Kok, 2013; Stok, de Ridder, Vet & de Wit, 2014). In both experiments, the descriptive norm message had more of an effect then the injunctive norm message. However, in the experiment by Mollen et al. (2013), the injunctive norm message (“have a salad for lunch") can be criticised for being ambiguous.
Both studies attempted to manipulate the social norm with just one sentence. Stok et al. (2014) mention that this might not lead to strong influence. They add that if a social norm is repeated, it, in turn, could lead to becoming a stronger social norm, which should have a strong influence on people’s eating intentions and habits. These differing results indicate that manipulating social norms is only temporarily effective, if at all, in influencing people’s intention, and that a social norm is a construct that shapes someone’s intentions over time.
Compared to social modelling and the influence on people’s eating behaviour, a social norm does not seem to be useful when manipulated, since its effect is rather short-term. People’s food choices are determined and formed by norms other people provide for daily life, longer-term. These findings above illustrate that it might be better not to manipulate social norms but to rather see them as an overall embodied construct that every individual has developed over time. Since multiple studies have examined the perceived social norm (a combination of injunctive and descriptive norms), for this study, the term perceived social norm refers to both a descriptive and an injunctive norm (e.g., Stok, 2014; Thompson, Bachman, Watson, Baranowski & Cullen, 2008, Wood Baker et al., 2003). This study examines the total embodied social norm and investigates the influence that a person’s perceived social norm can have.
The offline behaviour of people’s surroundings shapes the perceived norm of every individual and could explain food choices (Cruwys, 2015). A meta-analysis by Stok (2014) discovered that there is a relation between perceived social norms of a peer group and eating behaviour. Additionally, researchers examined whether the peers’ encouragement of healthy eating and discouragement of unhealthy eating was related to the intention to eat healthy. Stok (2014) found that especially the encouragement of healthy eating by peers predicted a food intake of more healthy than unhealthy food. This finding indicates that peer-subjective norms appear to influence food choices and play an important role in eating habits. What friends, family members, and other peers approve or disapprove of, and observing the behaviour of those peers, can lead to social pressure or perceived social norms to engage or not engage in specific behaviour. When people believe that the most respected others in their surroundings expect them to demonstrate a specific behaviour, such a perceived social norm will pressure a person to engage in that behaviour (Ajzen & Fishbein, 1980; Fishbein & Ajzen, 1975).
The elaboration likelihood model (Petty & Cacioppo, 1986) also indicates that persuasive information can be processed differently. Persuasive information is either
processed in-depth through a central route or, without complex processing, through a peripheral route. When a person considers information as necessary, truthful, and highly credible, he or she will process the information through the central route (Bredahl, Grundert, & Frewer, 1998). Consequently, when people see an Instagram post that agrees with their norms, it can be seen as highly credible and can be processed more consciously. However, when this Instagram picture does not agree with a person’s norms, this information will be processed peripherally and might not lead to the desired persuasive effect.
Instagram could lead to a momentary social modelling effect and therefore could influence a person’s intention to eat similarly. But the underlying norm a person experiences in his offline surroundings from his peers can intervene here. The behaviour being modelled might not match up with that person’s perceived social norm of his or her daily surroundings and peers and might not be processed as expected.
In this study, it is expected that an Instagram food post could affect someone’s intention to eat healthy but it is also expected that social pressure, from a perceived social norm, can strengthen or weaken the influence an Instagram post has on the intention to eat healthy (see Figure 1). On these grounds, the present paper assumes that the effect of
Instagram posts portraying food (healthy vs. unhealthy) on the intention to eat healthy will be moderated by perceived social norms (healthy vs. unhealthy) in such a way that the Instagram pictures showing healthy food will lead to an increased intention to eat healthy, and that this effect is even stronger for people who hold a healthy perceived social norm, compared to people with an unhealthy perceived social norm (H4).
Based on the findings above, a social norm is an embodied construct and can influence people’s eating habits. Wood Baker, Little, and Brownell (2003) found that social norms do not exert influence directly on intention, but rather via attitudes. In particular, adolescents who had perceived unhealthy social norms, meaning that their parents and their peer group did not care about their eating habits and that their peers were not eating healthy themselves,
were less likely to have positive attitudes towards food and instead demonstrated little intention to eat healthy. The elaboration likelihood model additionally deals with the impact of information on attitude change (Petty & Cacioppo, 1986; Bredahl, et al.,1998). Persuasive information can be processed differently and affect a person’s attitude in a way that, when persuasive information does not fit with someone’s norms, it might not seem credible and therefore will be processed peripherally. On the other hand, when someone has a healthy perceived social norm, and sees an Instagram post portraying healthy food, this match in perception could lead to more conscious processing of the Instagram post, probably leading to a positive attitude towards food. Consequently, a social norm can intervene in the effect that an Instagram post might have on someone’s attitudes towards food, in a way that when the norm matches the Instagram post, the healthy or unhealthy message will be processed more consciously, leading to the desired social modelling effect. This study expects that the portrayal of food on Instagram can influence the attitudes towards food but the perceived social norms of a person can strengthen or weaken that influence (see Figure 1).
More specifically, it is proposed that the effect of Instagram posts portraying food (healthy vs. unhealthy) on attitudes towards food will be moderated by perceived social norms (healthy vs. unhealthy) in such a way that the Instagram posts portraying healthy food will lead to a positive attitude towards food, and that this effect is stronger for people who hold a healthy perceived social norm, compared to people with an unhealthy perceived social norm (H5).
Figure 1. Moderated mediation model of the proposed hypotheses
Method Sample
Data were collected via an online experiment which was available between 17th of April and 2nd of May 2019. Participants were recruited via Facebook, Instagram, and private messaging. To recruit people with a perceived healthy and unhealthy social norm, the link was shared amongst different platforms, including platforms attracting fast-food lovers. An Instagram account was not necessary but participants needed to know what Instagram is and how it works, since participants should be able to identify the pictures as posted on the platform Instagram.
A total of 305 people were recruited for this study, from which 49 participants were removed because they did not finish the entire study, 13 due to missing values, and four because they indicated not to be familiar with Instagram. This led to a total sample of 239 participants between the ages of 18 and 66 (M=28.83, SD=10.32). The majority (69.5%) identified as female and more than half had a high educational degree, such as a bachelor’s degree or higher (64.4%) (see Table 1). The sample was highly international, with participants from 28 countries. Most of the participants said they had spent most of their lives in
Instagram posts portraying food
(healthy vs. unhealthy) Perceived social norms about food
(Healthy vs. unhealthy) Attitude towards food Intention to eat healthy H4 H1+ H2+ H5 H3+
Switzerland (46%), followed by the Netherlands (22.2%), Germany (5%), and the United States (3.3%) (see Appendix A). The majority of the participants (84.9%) said they have their own Instagram account, and nearly half of the participants (42.3%) use Instagram several times per day or every day (20.5%).
Table 1. Background characteristics of participants Overall sample (N=239) Healthy food posts (N=122) Unhealthy food posts (N=117) Age 28.83 (0.50) 28.80 (10.71) 28.86 (9.99) Gender Male 30.5% (73) 69.5% (166) 27.9% (34) 33.3% (39) Female 72.1% (88) 66.7% (78) Education
Less than high school diploma 1.3% (3) 16.7% (40) 17.6% (42) 43.1% (103) 19.7% (47) 1.7% (4) 0.8% (1) 1.7% (2)
High school degree or equivalent 13.9% (17) 19.7% (23)
Matura/Abitur or similar 15.6% (19) 19.7% (23)
Bachelor’s degree (e.g., BA, BSc) 48.4% (59) 37.6% (44)
Master’s degree (e.g., MA, MSc) 20.5% (25) 18.8% (22)
Doctorate (e.g., Ph.D., Ed.D.) 0.8% (1) 2.6% (3)
Design
To test the moderated mediation model (see Figure 1), a 2 x 2 factorial between-subjects design was used with healthy vs. unhealthy Instagram food posts and healthy vs. unhealthy social norms. Social norms is a quasi-experimental variable that would act as a moderator, attitude towards food was seen as a mediator in this design, and the intention to eat healthy functioned as the dependent variable in this model.
Procedure
Participants were recruited via links posted on Facebook, Instagram, and WhatsApp. By clicking on the link, participants entered the online experiment, where they were informed that the survey is about Instagram and food. Participants were randomly assigned to one of the two conditions (healthy or unhealthy food posts). One filter question assessed whether the participant knew what Instagram is. Participants who answered with "no" were immediately excluded from the experiment. Instagram usage and past dietary habits were assessed at the beginning. Next, the demographic elements of gender, age, country of origin, and education were assessed. Subsequently, the participants saw one of the Instagram feeds (healthy or unhealthy). Afterwards, attitudes towards food and the intention to eat healthy were questioned. Then, the manipulation check and control variables were presented to the
participants and after that, the perceived social norms were assessed. Finally, the participants were thanked for their contribution and the experiment ended.
Stimulus materials
Instagram posts portraying food. The independent variable Instagram food posts were manipulated at two levels: healthy vs. unhealthy food posts (see Figures 2 and 3). Participants were randomly exposed to different stimulus material consisting of Instagram posts. Both groups were shown four Instagram posts; the first picture showed a meal (salad or burger), the next picture a snack (vegetable sticks or chips), the third picture something sweet (fruit platter or doughnuts), and the last picture portrayed a beverage (water with lemon or Coca Cola). The four food categories were chosen to provide a broad idea of food possibilities so the results could be tracked back to the entire category of healthy and unhealthy food (Dixon et al., 2007). All pictures looked equal with regard to food quality, and there was no person in the picture, since these factors should not influence the decision-making. All of the pictures were taken from Instagram and had various captions mentioning exclamations such as
“Yummy lunch today!” with three popular hashtags next to it (Chapman, 2019). All meals, snacks, sweets, and beverages in the healthy and unhealthy conditions had the same caption, were posted by the same fictional person under a fictional Instagram name, and had identical hashtags. The hashtags were neutral and did not indicate a sort of norm or expectation. The only element that differed between the conditions was the picture itself.
As in a normal Instagram feed, there were pictures of the people who either posted a story or the food pictures. However, these people were unrecognisable, had given their consent that their picture could be used in this study, or were taken from an online news article (Haylock, 2017).
Figure 3. Stimulus material, unhealthy Instagram posts
Pilot study
A pilot study was conducted to test the questionnaire and to determine how the
participants rated and perceived the Instagram posts as stimulus material. Participants (N=25) were recruited via private messaging and filled in the survey. The pilot questionnaire asked some additional questions to measure perceived healthiness, attractiveness, appeal, quality, and colourfulness of the posts on a scale from 1 to 7.
Manipulation of the Instagram post was successful since participants perceived the healthy Instagram post as significantly healthier (Mhealthy=6.00, SD=1.29) than the unhealthy Instagram post (Munhealthy=2.08, SD=0.49; t(20)=16.94, p < .000). Additionally, there was no significant difference in perceived attractiveness (Mhealthy=5.92, SD=1.32; Munhealthy=5.00, SD=1.87; t(24)=-1.45, p=.159), appeal (Mhealthy=5.92, SD=1.26; Munhealthy=5.08, SD=1.85; t(24)=-1.37, p=.185), quality (Mhealthy=6.20, SD=0.84; Munhealthy=5.83, SD=1.17; t(9)=-0.59, p=.573) and colourfulness of the post (Mhealthy=6.80, SD=0.45; Munhealthy=6.33, SD=0.82;
t(9)=-1.14, p=.285). Participants additionally indicated they had read the captions underneath the Instagram posts but none of them claimed to have felt persuaded or influenced by the captions in any way.
Measures
Perceived social norms about food. Perceived social norms were a nominal moderator of the influence of Instagram posts on attitudes towards food. Adapted from Fishbein and Ajzen (2010), the behaviour in question needed to be defined in terms of TACT: target, action, context, and time elements. To measure perceived social norms, the four most important food behaviours recommended by the World Health Organization (2018a) were chosen: eating at least five portions of vegetables and fruits (like 400g) each day and drinking two litres of water every day as healthy behaviour, and eating fried food regularly and
drinking more than half a litre of sugary sweetened drinks every day as unhealthy behaviour (World Health Organization, 2018a).
Eight items assessed the perceived social norms about food. Four items adapted from Fishbein & Ajzen (2010) and Prentice (2008) measured perceived injunctive norms. For this, the four statements mentioned earlier were reformulated, for example, as “People who are important to me are expecting me to drink two litres of water every day”. The four statements were assessed on a seven-point scale ranging from 1=strongly agree to 7=strongly disagree. Four additional items measured descriptive norms; here, the same four statements from above were taken and reformulated into items such as “Most people who are important to me drink two litres of water every day" and measured on a seven-point scale ranging from 1=strongly agree to 7=strongly disagree. The healthy norm items were recoded to ensure that the direction of the scale was similar to the unhealthy norms. Consequently, higher numbers indicated a healthier social norm and lower numbers an unhealthier social norm.
An exploratory factor analysis, with oblimin rotation, indicated that the scale was loading on three factors, explaining 71.67% of the variance. These three factors divided the healthy and unhealthy social norms from each other and, additionally, injunctive healthy norms from descriptive healthy norms. These results indicated a difference between how healthy and unhealthy norms were perceived. However, considering that the eight-item scale proved to be reliable, as indicated by a Cronbach’s Alpha of 0.70; the different norms are used in combination by several authors (e.g., Stok, 2014; Thompson et al., 2008, Wood Baker et al., 2003); and the variable would be split into healthy and unhealthy norms groups, it was decided to compute an overall scale score (MTotal=5.10, SD=0.78) and perform a median split on these scores. The healthy social norm group comprised 110 participants (Mhealthy=5.76; SD=0.42), and the unhealthy social norm group comprised 129 participants (Munhealthy=4.54; SD=0.52).
Intention to eat healthy. The dependent variable, intention to eat healthy, was assessed with four items adapted from the statements mentioned above (Ajzen n.d.; World Health Organization, 2018a), for example, “I intend to drink two litres of water every day for the next three months” ranging from 1=strongly agree to 7=strongly disagree.
The healthy items were recoded to ensure that the direction was similar to the unhealthy items. Consequently, a higher score indicated a higher intention to eat healthy.
An exploratory factor analysis with oblimin rotation indicated that the scale was loading on two factors, explaining 71.25% of the variance. These two factors divided the healthy and unhealthy food and drink intentions from each other. The four-item scale proved not to be reliable (α=0.54). Since the two healthy-intention items significantly correlated with each other (r=0.40, p=.010), as well as the two unhealthy-intention items (r=0.44, p=.010), the healthy- and unhealthy-intention statements seem to measure different constructs. Since the hypotheses of this study concerned the intention to eat healthy, it was decided to calculate a mean score with only the two healthy-intention items (M=5.64, SD=1.08). Higher scores of
this variable correspond to higher intention to eat healthy, and lower scores correspond to a lower intention to eat healthy.
Attitude towards food. According to Ajzen (2002), attitudes about specific behaviour
are derived from behavioural beliefs. The participants’ attitudes towards eating and drinking either healthy or unhealthy was assessed via four questions in each of five categories. Four questions adapted from the statements above (Ajzen, 2002; Fishbein & Ajzen, 2010;
Magnusson, Arvola, Koivisto Hursti, Åberg & Sjödén, 2001) assessed the attitudes towards food with items such as “Me drinking two litres of water every day for the next three months would be”: 1=very bad to 5=very good, 1=unpleasant to 5=pleasant, 1=unenjoyable to 5=enjoyable, 1=unimportant to 5=important, and 1=harmful to 5=beneficial.
First, the unhealthy statements were recoded, so all answers ranged from positive to negative. An exploratory factor analysis, with oblimin rotation, indicated that the scale was unidimensional, explaining 62.14% of the variance. Since the scale also proved reliable (α=0.84), a mean score was calculated with all items (M=4.18, SD=0.49). Higher scores correspond to a more positive attitude about healthy food, and lower scores correspond to a more negative attitude about healthy food.
Manipulation check
To assess whether the manipulation was perceived as intended, the following question was asked: “To what extent do you believe that the foods and drinks portrayed on these Instagram posts were healthy or unhealthy?”: 1=very healthy to 7=very unhealthy. An independent samples t-test indicated that participants perceived the Instagram post portraying healthy food as significantly healthier (M=1.75, SD=0.73) than the unhealthy Instagram post (M=6.01, SD=1.06), t(237)=36.19, p < .000. This result indicates that the manipulation of the Instagram posts was successful.
Randomisation check
To determine whether the study variables were equally distributed over the two experimental Instagram conditions, chi-square and independent samples t-tests were
performed. The results indicated that the two experimental Instagram conditions did not differ with respect to age (t(237)=-0.51, p=0.959); gender χ2(1)=0.841, p=.359; education
χ2
(5)=4.89, p=.430; and country of long-time residence χ2(27)=21.70, p=.753. Thus, randomisation was successful.
Other variables Control variables.
Next to the background variables of gender, age, and country of long-time residence, several control variables were assessed.
Instagram usage was assessed using a seven-point scale asking, “How often do you
use or have you used Instagram?” ranging from 1=less than once a month to 7=several times
a day.
Past behaviour with regard to food consumption was assessed, according to Ajzen
(2002), by using multiple questions adapted from the four statements such as “In the course of the last months, how often have you eaten fried foods?”: 1=never to 7=every day (World Health Organization, 2018a). Here, the exploratory factor analysis, with oblimin rotation, indicated that the scale was loading on two factors, explaining 62.80% of the variance. A correlation indicated that the two healthy statements were significantly correlating by 0.352, p=0.01. It was therefore decided to only use the two healthy past behaviour statements.
Perceived attractiveness and appeal of the food portrayed on the Instagram posts were
assessed with two items: “To what extent did the food you just saw look attractive/
appealing?”: 1=very much so to 7=not at all. These two items were then computed into one variable revealing the total attractiveness of the food in the post. Adapted from Magnusson et
al. (2001), the price and availability of healthy food was assessed, asking participants to rank healthy food as 1=cheap to 7=expensive and as 1=easily available to 7=difficult to find. Lastly, participants were asked to indicate whether they had any restrictions in their diet, such as vegetarianism, veganism, or religious-based diets (Patience, 2016; Nordqvist, 2017).
Correlation analyses revealed that only healthy past food behaviour correlated significantly with the intention to eat healthy (see Table 2).
An independent samples t-test indicated that the control variables availability (t(237)=-0.09, p=0.931); price of healthy food t(237)=-0.42, p=0.674; and healthy past behaviour t(237)=1.17, p=0.243) were equally distributed among the two experimental Instagram conditions. On the other hand, Instagram usage, diet, and attractiveness of the post were not equally distributed among the conditions. Participants’ mean Instagram usage in the healthy food condition (Mhealthy=5.47, SD=2.05) was significantly higher than the participants’ mean Instagram usage in the unhealthy food condition (Munhealthy=4.72, SD=2.48),
t(237)=2.55, p=0.011. A chi-square analysis indicated that significant differences were found between the conditions and participant’s diet, χ2 (1)=6.61, p=.010. In the healthy Instagram post condition, 30.3% had a restricted diet (versus 69.7% omnivores) and in the unhealthy food condition, 83.8% were omnivores (versus 16.2% restrictive diet). Participants’ mean attractiveness in the healthy food condition (Mhealthy=5.98, SD=1.25) was significantly higher than the participants’ mean attractiveness in the unhealthy food condition (Munhealthy=5.26, SD=1.52), t(237)=-4.02, p=0.000.
This means that randomisation across conditions was not totally successful on the control variables and that Instagram usage, diet, and total attractiveness of the food in the post should be included in the analyses as covariates.
Main analysis Hypothesis testing
An independent samples t-test was performed to determine whether the exposure to an Instagram post portraying healthy food will lead to a greater intention to eat healthy than exposure to an Instagram post portraying unhealthy food (H1). The results indicate that there was no significant difference in participants’ mean intention to eat healthy: t(237)=0.97, p=0.333. For both groups, the intention to eat healthy was relatively high, and H1 could not be supported.
To determine whether the influence of Instagram posts on the intention to eat healthy is mediated by attitudes towards food, in a way that a healthy post leads to more positive attitudes towards food (H2), which in turn leads to a higher intention to eat healthy (H3), a mediation analysis was conducted using Model 4 of the SPSS macro PROCESS. PROCESS conducts an OLS regression-based path analysis to estimate the direct and indirect effects (Hayes, 2013). The model with 5,000 bootstrap samples assessed the expected mediation with Instagram usage, diet, past behaviour, and attractiveness as covariates.
As illustrated in Table 3, the results did not reveal a significant direct effect of Instagram posts portraying food on the participants’ intention to eat healthy (b=-0.09, p=.463). Additionally, there was no significant effect of Instagram posts portraying food on the participants’ attitude towards food (b=0.03, p=.688). Both groups had relatively high and therefore healthy attitudes towards food (Mhealthy=4.17, SD=0.50; Munhealthy=4.19, SD=0.48). Thus, H2 failed to be supported. Nonetheless, a significant result was found regarding the participants’ attitudes towards food on their intention to eat healthy (b=0.40, p=.001). A more positive attitude towards food leads to a higher intention to eat healthy. Thus, H3 is supported. However, the total indirect effect of Instagram posts portraying food on the participants’ intention to eat healthy was not significant and almost zero (Indirect effect=-0.01, boot SE=0.03, BCI [-0.04, 0.07]). Hence, there is no mediation.
The covariate healthy past behaviour was found to have a significant and positive effect not only on attitudes towards food (b=0.14, p=.000), but also on the intention to eat healthy (b=0.36, p=.000). This means that healthy past behaviour leads to a more positive and therefore healthy attitude towards food, as well as to a higher intention to eat healthy.
Table 3. OLS-regression model coefficients for the mediation model (Model 4) Antecedents Attitude towards food Intention to eat healthy
b p CI b p CI Constant 3.69 0.000 [3.40, 3.99] 2.28 0.000 [1.21, 3.35] Instagram post 0.03 0.688 [-0.10, 0.15] -0.09 0.463 [-0.32, 0.15] Attitude - - - 0.40*** 0.001 [0.16, 0.65] Instagram usage -0.01 0.283 [-0.04, 0.11] -0.00 0.912 [-0.05, 0.05] Diet 0.09 0.192 [-0.05, 0.23] 0.16 0.235 [-0.11, 0.43] Attractiveness -0.01 0.556 [-0.05, 0.03] 0.03 0.501 [-0.05, 0.11] Past behaviour 0.14*** 0.000 [0.10, 0.18] 0.36*** 0.000 [0.28, 0.44] R2=.21 F(5, 233)=12.30*** p=.000 R2=.40 F(6, 232)=26.21*** p=.000 ***p=.000
The moderated mediation model (Model 8) of the SPSS macro PROCESS was used to test whether the effect of Instagram post portraying food (healthy vs. unhealthy) on the
participants’ intention to eat healthy (H4) or their attitudes towards food (H5) will be moderated by perceived social norms (healthy vs. unhealthy) (see Table 4). The model with 5,000 bootstrap samples assessed the expected moderation and overall model with Instagram posts as the independent variable, social norms as the moderator, attitude towards food as the mediator, and intention to eat healthy as the dependent variable. Instagram usage, diet, attractiveness, and past behaviour were again included as covariates.
There was no significant interaction effect between Instagram posts portraying food and social norms on intention to eat healthy (b=0.03, p=.892). Regarding the moderated mediation model proposed in H4, the index of moderated mediation was not significant, as the 95% bootstrap confidence interval contained zero (index of moderated mediation=-0.07, boot SE=0.05, BCI [-0.20, 0.01]). Thus, H4 failed to be supported.
A marginally significant interaction effect was found between Instagram posts and social norms on attitudes towards food (b=-0.21, p=.068). This, together with the conditional effects, indicates that Instagram post portraying healthy food in combination with a healthy social norm seem to have a negative effect on attitudes towards food (b=-0.09, p=.292). For the unhealthy social norm there was no effect.
Social norms seem to have a direct and positive effect on attitudes towards food (b=0.35, p=.000). A positive social norm seems to lead to a more positive attitude towards food.
Again, attitude towards food affected the intention to eat healthy (b=0.35, p=.008). This means that a positive attitude towards food will lead to a higher intention to eat healthy.
Lastly, healthy past behaviour did seem to have a direct effect not only on attitudes towards food (b=0.13, p=.000), but also on the intention to eat healthy (b=0.36, p=.000). These results indicate that a healthy past behaviour can lead to a more positive attitude towards food, as well as to a higher intention to eat healthy. See Table 4 for all of the results.
Table 4. OLS-regression model coefficients for a moderated mediation model (Model 8) Antecedents Attitude towards food Intention to eat healthy
b p CI b p CI Constant 3.55 0.000 [3.26, 3.84] 2.43 0.000 [1.34, 3.51] Instagram post 0.11 0.139 [-0.04, 0.27] -0.01 0.520 [-0.40, 0.20] Attitude - - - 0.35** 0.008 [0.09, 0.61] Social norms 0.35*** 0.000 [0.20, 0.51] 0.17 0.314 [-0.16, 0.49] Instagram post* Social norms -0.20 0.068 [-0.42, 0.02] 0.03 0.892 [-0.41, 0.47] Instagram usage -0.01 0.412 [-0.03, 0.01] -0.00 0.900 [-0.05, 0.05] Diet 0.07 0.297 [-0.06, 0.21] 0.14 0.307 [-0.13, 0.41] Attractiveness -0.01 0.672 [-0.05, 0.03] 0.03 0.465 [-0.05, 0.11] Past behaviour 0.13*** 0.000 [0.09, 0.16] 0.36*** 0.000 [0.28, 0.44] R2=.28 F(7, 231)=12.87*** p=.000 R2=.41 F(8, 230)=20.00*** p=.000 ***p=.000; **p < .010; *p < .050
Discussion and conclusion
The aim of this study was to investigate to what extent Instagram posts portraying healthy or unhealthy food can influence the intention to eat healthy and what roles are played by attitudes towards food and perceived social norms.
The results of this study indicate that no significant effect was found between
Instagram posts portraying food and the intention to eat healthy. Additionally, this effect does not seem to be mediated by attitudes towards food. Nonetheless, it was found that a more positive attitude towards food leads to a significant higher intention to eat healthy. This indicates that attitudes towards food are an underlying determinant of the intention to eat healthy (Ajzen, 1991). The Instagram posts portraying food did not influence attitudes towards food and intention to eat healthy. This implies that the Instagram post did not have the effect as expected. This could be due to a lack of social closeness from the Instagram post.
The literature on social modelling of eating emphasises that identification and a social context are key components, which can trigger modelling behaviour. Individuals compare themselves to another person and decide whether his or her behaviour is relevant to them (Higgs & Thomas, 2016). Although studies have suggested that social modelling of eating does occur, even when dining partners are non-familiar (Kaisari & Higgs, 2015), it does seem that this is different in an online scenario. Instagram is perceived as a social media platform, where a participant can follow friends, brands, and people in which he or she is interested. These social ties were possibly not emphasised enough in the stimulus material, and therefore it might be assumed that the participants did not have any social connection to the people who posted the food pictures on Instagram. This implies that the results could have been
significant if the stimulus material had been created more personally. Additionally the
cultivation theory claims, repeated exposure will lead to greater influence; in other words, that the recipient will take over similar beliefs as portrayed in the medium (Gerbner et al., 1994).
Furthermore, it was investigated whether perceived social norms moderated the effect of Instagram posts portraying food on attitudes towards food and the intention to eat healthy. No significant interaction effect was found between Instagram post and perceived social norms on the intention to eat healthy. However, a marginally significant interaction effect was found between Instagram posts and perceived social norms on attitudes towards food. This indicates that Instagram posts showing healthy food, in combination with a healthy social norm, seemed to have a negative effect on attitudes towards food. The elaboration likelihood model claims that when persuasive information is seen as relevant and credible, in other words, in line with someone’s norms, it will be processed centrally and more consciously. Therefore, it seems that this led to a more critical evaluation and a sort of reactance (Petty & Cacioppo, 1986; Bredahl et al., 1998). Reactance theory claims that when people believe that something or someone is trying to persuade them, their freedom is threatened. This creates a boomerang effect and reflects the content in a more negative light (Wang & Jones, 2017).
This phenomenon could have affected people’s perception of the Instagram posts, which in turn can lead to a negative attitude towards food.
Furthermore, a direct effect was found between social norms and attitudes towards food. Additionally, attitude towards food had a direct effect on the intention to eat healthy. These results go hand in hand with the findings of Wood, Baker, Little, and Brownell (2003) and indicate that the assumed moderation of the effect of an Instagram posts on attitudes by social norms is not the case, but that it is rather a direct effect of social norms on attitudes towards food. This finding leads to the conclusion that perceived social norms influence attitudes, and these, in turn, affect the intention to eat healthy. Since social norms seem to have a direct effect on the intention to eat healthy as well, it can be concluded that attitudes towards food mediate the effect of perceived social norms on the intention to eat healthy. Consequently, having a healthy social norm might lead to a more positive attitude towards food, which in turn leads to a higher intention to eat healthy (Ajzen, 1991). Here, Instagram might be used additionally to create a healthy social norm over time by posting an outstanding amount of healthy food. Nevertheless, more research is needed to illustrate this phenomenon.
Managerial implications
The results of this study can be used as an inspiration for professionals in health
communication. It seems important to strive to create a healthy social norm among people, to influence their attitudes towards food and their intention to eat healthy. Although no
significant influences of the Instagram post were found, social media can still be used to create a healthier social norm over time. As this study indicates, no immediate social modelling influence was found. Nonetheless, over time, social media can strive to create a healthier social norm by having more healthy content online. When friends, family, peers, and influencers display mostly healthy food posts online, this might create a healthy social norm in people, which in turn leads to a more positive attitude towards food and then to a higher
intention to eat healthy. With this process, the societal problem of obesity can be tackled, and a healthier society could be created.
Limitations and future research
This study had several limitations, which will be discussed in the following. The sample consisted mainly of well-educated people. It is likely that this fact could have influenced the results, since people with higher education are more health-conscious and enjoy healthier diets (Xie, Gilliland, Li & Rockett, 2003). In contrast, manual workers do more physical labour, therefore have higher energy requirements and prefer foods higher in calories (BBC, 2019; Jfmed, 2019). This characteristic could indicate why the overall sample already had a rather healthy diet, as well as reported a rather healthy social norm, and were positive towards healthy food, likely biasing the results. Future researchers could replicate this study, using a more representative sample.
Another factor to consider is that more than half of the participants in the sample (62.8%) use Instagram every day or several times a day. This repeated behaviour can be seen as a habit that has formed. Habits that form over social media are formed automatically; it has been demonstrated that individuals become less attentive as a result (Larose, Kim & Peng, 2010). The lack of attention towards the Instagram posts could lead to a more peripheral processing of the information and would impede any effect on the intention to eat healthy (Petty & Cacioppo, 1986). When an Instagram post is not consciously processed, there is no possibility that a social modelling effect can occur. Future studies could attempt to make participants more attentive by instructing them to pay attention to the stimulus material.
Although the manipulation of the Instagram post worked as intended, the stimuli material did not have the desired social modelling effect. This might have been due to
perceived attractiveness; the healthy Instagram post was perceived as more attractive than the unhealthy Instagram post. This factor could have biased the overall perception of the
Instagram posts. As discussed earlier, the lack of a social context in the stimulus material might be another reason that there was no social modelling effect. If participants were
exposed to their own Instagram account and saw posts only from people they know or follow, the social context would have been closer and the perception could have been more focussed. Random food posts from unknown people seemed to lack that social context and therefore do not lead to the expected modelling influence (e.g., Cruwys et al. 2015). As mentioned earlier, future research should attempt to use people’s own Instagram pages or make it appear as if the pictures were posted by their peers.
The measurement of intention to eat healthy was based on Fishbein and Ajzen (2010) and used the World Health Organization’s (2018a) four most important behaviours
concerning healthy and unhealthy food. Nevertheless, the unhealthy and healthy statements seem to measure different constructs and were not unidimensional as expected. There could be the assumption that participants, who consider their eating habits as healthy, see unhealthy behaviour as a different construct, because they can see it as a reward and enjoyment rather than simply nutrition (De Andrade Previato & Behrens, 2017). Different cultures can additionally influence the perception of healthy food, for example, tasty food is mostly associated in the United States with unhealthy food, whereas in France, healthy food is linked to tastiness (Werle, Trendel & Ardito, 2013). Considering that this sample had participants from 28 different countries, that might explain a difference in food perception, a fact that could also be tied to the intention to eat healthy. Therefore, for future research, it would be interesting to test what a participant perceives as healthy, unhealthy, and tasty food and if he or she sees unhealthy food as a reward instead of simply nutrition. Of these questions, the computed variable food perception could be added to the proposed model and include factors on how people perceive food.
Overall, this research provides clues to methods for how obesity can be tackled in a creative way to build a healthier, better society.
References
Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50(2), 179–211. doi:10.1016/0749-5978(91)90020-T
Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social behaviour. Englewood Cliff, NJ: Prentice-Hall.
Ajzen, I. (n.d.). Constructing a theory of planned behaviour questionnaire. Retrieved from http://www.midss.org/sites/default/files/tpb.construction.pdf.
Ajzen, I. (2002). Constructing a TpB Questionnaire: Conceptual and Methodological Considerations. Retrieved from
https://pdfs.semanticscholar.org/0574/b20bd58130dd5a961f1a2db10fd1fcbae95d.pdf. BBC. (2019). Diet. Retrieved June 3, from
https://www.bbc.com/bitesize/guides/zdjfr82/revision/6
Beaver, J., Lawrence, A., van Ditzhuijzen, J., Davis, M., Woods, A., & Calder, A. (2006). Individual differences in reward drive predict neural responses to images of food. The Journal of Neuroscience: the Official Journal of the Society for Neuroscience, 26(19), 5160–5166. doi:10.1523/JNEUROSCI.0350-06.2006
Bredahl, L., Grunert, K., & Frewer, L. (1998). Consumer Attitudes and Decision-Making With Regard to Genetically Engineered Food Products – A Review of the Literature and a Presentation of Models for Future Research. Journal of Consumer Policy, 21(3), 251–277. doi:10.1023/A:1006940724167
Chadwick, P., Crawford, C., & Ly, L. (2013). Human food choice and nutritional interventions. Nutrition Bulletin, 38(1), 36–42. doi:10.1111/nbu.12005
Chapman, R. (2019). 105 Best Hashtags For Food on Instagram That’ll Have Your Followers Saying #ForkYeah. Elite daily. Retrieved from: https://www.elitedaily.com/p/105- best-hashtags-for-food-on-instagram-thatll-have-your-followers-saying-forkyeah-15933140.
Christakis, N., & Fowler, J. (2007). The spread of obesity in a large social network over 32 years. The New England Journal of Medicine, 357(4), 370-379.
doi:10.1056/NEJMsa066082
Cialdini, R., Reno, R., & Kallgren, C. (1990). A Focus Theory of Normative Conduct: Recycling the Concept of Norms to Reduce Littering in Public Places. Journal of Personality and Social Psychology, 58(6), 1015–1026.
doi:10.1037/0022-3514.58.6.1015
Cialdini, R. B., Kallgren, C. A., & Reno, R. R. (1991). A focus theory of normative conduct: A theoretical refinement and re-evaluation. Advances in Experimental Social
Psychology, 24(20), 201–234. doi:10.1016/S0065-2601(08)60330-5
Cialdini, R. B., & Trost, M. R. (1998). Social influence: Social norms, conformity and
compliance. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), The handbook of social psychology (pp. 151-192). New York, NY, US: McGraw-Hill.
Cooke, R., & Papadaki, A. (2014). Nutrition label use mediates the positive relationship between nutrition knowledge and attitudes towards healthy eating with dietary quality among university students in the UK. Appetite, 83, 297–303.
doi:10.1016/j.appet.2014.08.039
Covello, V. (2003). Best Practices in Public Health Risk and Crisis Communication. Journal of Health Communication, 8(1), 5–8. doi:10.1080/713851971
Cruwys, T., Bevelander, K., & Hermans, R. (2015). Social modeling of eating: A review of when and why social influence affects food intake and choice. Appetite, 86, 3–18. doi:10.1016/j.appet.2014.08.035
De Andrade Previato, H., & Behrens, J. (2017). Taste-related factors and food neophobia: Are they associated with nutritional status and teenagers’ food choices? Nutrition, 42, 23– 29. doi:10.1016/j.nut.2017.05.006
Dixon, H., Scully, M., Wakefield, M., White, V., & Crawford, D. (2007). The effects of television advertisements for junk food versus nutritious food on children’s food attitudes and preferences. Social Science & Medicine, 65(7), 1311–1323.
doi:10.1016/j.socscimed.2007.05.011
Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior: An introduction to theory and research. London: Addison-Wesley.
Fishbein, M., & Cappella, J. N. (2006). The role of theory in developing effective health communications. Journal of Communication, 56, 1-17. doi:10.1111/j.1460-2466.2006.00280.x
Fishbein, M., & Ajzen, I. (2010). Predicting and changing behavior: The reasoned action approach. New York: Psychology Press.
Gerbner, G., Gross, L., Morgan, M., & Signorielli, N. (1994). Growing up with television: The cultivation perspective. In J. Bryant, & D. Zillman (Eds.), Media effects: Advances in theory and research (pp. 17–42). Hillsdale, NJ: Lawrence Erlbaum Associates.
Guidry, J., Jin, Y., Orr, C., Messner, M., & Meganck, S. (2017). Ebola on Instagram and Twitter: How health organizations address the health crisis in their social media engagement. Public Relations Review, 43(3), 477–486.
doi:10.1016/j.pubrev.2017.04.009
Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. New York, NY: Guilford Press.
Haylock, Z. (2017, April 5). Do Instagram Model Hacks Work for Normal People?. Retrieved from
https://nyulocal.com/do-instagram-model-hacks-work-for-normal-people-2d0f18de908.
Hearty, Á., Mccarthy, S., Kearney, J., & Gibney, M. (2007). Relationship between attitudes towards healthy eating and dietary behaviour, lifestyle and demographic factors in a
representative sample of Irish adults. Appetite, 48(1), 1–11. doi:10.1016/j.appet.2006.03.329
Hefner, D., & Vorderer, P. (2016). Permanent Connectedness and Multitasking In L.
Reinecke & M. Oliver (Eds.), The Routledge handbook of media use and well-being: international perspectives on theory and research on positive media effects (pp. 237– 249). New York, NY: Routledge.
Herman, C. P., Roth, D. A., & Polivy, J. (2003). Effects of the presence of others on food intake: A normative interpretation. Psychological Bulletin, 129, 873–886.
doi:10.1037/0033-2909.129.6.873
Higgs, S., & Thomas, J. (2016). Social influences on eating. Current Opinion in Behavioral Sciences, 9(C), 1–6. doi:10.1016/j.cobeha.2015.10.005
Holmberg, C., E. Chaplin, J., Hillman, T., & Berg, C. (2016). Adolescents’ presentation of food in social media: An explorative study. Appetite, 99, 121–129.
doi:10.1016/j.appet.2016.01.009
Hu, Y., Manikonda, L., & Kambhampati, S. (2014). What we instagram: A first analysis of instagram photo content and user types. In Proceedings of the 8th International Conference on Weblogs and Social Media, ICWSM 2014 (pp. 595-598). The AAAI Press.
Jfmed. (2019). Estimation of daily requirements for energy and nutrients. Retrieved from https://www.jfmed.uniba.sk/fileadmin/jlf/Pracoviska/ustav-fyziologie/6ZH-Estimation.pdf.
Kaisari, P., & Higgs, S. (2015). Social modelling of food intake. The role of familiarity of the dining partners and food type. Appetite, 86, 19–24. doi:10.1016/j.appet.2014.09.020 Kuźbicka, K., & Rachoń, D. (2013). Bad eating habits as the main cause of obesity among
LaRose, R., Kim, J., & Peng, W. (2010). Social networking: Addictive, compulsive,
problematic, or just another media habit. A networked self: Identity, community, and culture on social network sites, 56, 678-685.
Lee, E., Lee, J., Moon, J., & Sung, Y. (2015). Pictures Speak Louder than Words: Motivations for Using Instagram. Cyberpsychology, Behavior, and Social Networking, 18(9), 552–556. doi:10.1089/cyber.2015.0157
Lê, J., Dallongeville, J., Wagner, A., Arveiler, D., Haas, B., Cottel, D., Simon, C., Dauchet, L. (2013). Attitudes toward healthy eating: a mediator of the educational level–diet relationship. European Journal of Clinical Nutrition, 67(8), 808–814.
doi:10.1038/ejcn.2013.110
Magnusson, M., Arvola, A., Koivisto Hursti, U., Åberg, L., & Sjödén, P. (2001). Attitudes towards organic foods among Swedish consumers. British Food Journal, 103(3), 209– 227. doi:10.1108/00070700110386755
Maheshwari, S. & Koblin, J. (2018). Why Traditional TV I in Trouble. The New York Times. Retrieved from https://www.nytimes.com/2018/05/13/business/media/television-advertising.html.
Mejova, Y., Haddadi, H., Noulas, A., & Weber, I. (2015). #FoodPorn: Obesity Patterns in Culinary Interactions. In Proceedings of the 5th International Conference on digital health 2015 (pp. 51–58). ACM. doi:10.1145/2750511.2750524
Mollen, S., Ruiter, R., & Kok, G. (2013). Healthy and unhealthy social norms and food selection: Findings from a field-experiment. Appetite, 65, 83-89.
doi:10.1016/j.appet.2013.01.020
Nordqvist, C. (2017). Nine most popular diets rated by experts 2017. Retrieved from https://www.medicalnewstoday.com/articles/5847.php.
Patience, S. (2016). Religion and dietary choices. Retrieved, from
http://www.independentnurse.co.uk/clinical-article/religion-and-dietary-choices/145719/.
Petty, R. E., & Cacioppo, J. T. (1986). The elaboration likelihood model of persuasion. In Communication and persuasion (pp. 1-24). Springer, New York, NY.
Prentice, D. (2008). Mobilizing and weakening peer influence as mechanisms for changing behavior. Implications for alcohol intervention programs. In M. J. Prinstein & K. A. Dodge [Eds.], Understanding peer influence in children and adolescents. (p. 161– 180.)
Prinsen, S., de Ridder, D., & de Vet, E. (2013). Eating by example. Effects of environmental cues on dietary decisions. Appetite, 70, 1–5. doi:10.1016/j.appet.2013.05.023
Roininen, K., Lähteenmäki, L., & Tuorila, H. (1999). Quantification of Consumer Attitudes to Health and Hedonic Characteristics of Foods. Appetite, 33(1), 71–88.
doi:10.1006/appe.1999.0232
Roininen, K., & Tuorila, H. (1999). Health and taste attitudes in the prediction of use
frequency and choice between less healthy and more healthy snacks. Food Quality and Preference, 10(4), 357–365. doi:10.1016/S0950-3293(98)00057-3
Salvy, S., Kieffer, E., & Epstein, L. (2008). Effects of social context on overweight and normal-weight children’s food selection. Eating Behaviors, 9(2), 190–196. doi:10.1016/j.eatbeh.2007.08.001
Statista. (2018). Number of monthly active Instagram users from January 2013 to June 2018 (in millions). Retrieved from
https://www.statista.com/statistics/253577/number-of-monthly-active-instagram-users/. Stok, F. M. (2014). Eating by the Norm: The Influence of Social Norms on Young People's
Stok, F.M., Ridder, de, D.T.D., Vet, de, E., & Wit, J.B.F. (2014). Don’t tell me what I should do, but what others do: The influence of descriptive and injunctive peer norms on fruit consumption in adolescents. British Journal of Health Psychology, 19(1), 52-64. doi:10.1111/bjhp.12030
Thompson, V., Bachman, C., Watson, K., Baranowski, T., & Cullen, K. (2008). Measures of self-efficacy and norms for low-fat milk consumption are reliable and related to beverage consumption among 5th graders at school lunch. Public Health
Nutrition, 11(4), 421–426. doi:10.1017/S1368980007000547
Trogdon, J., Finkelstein, E., Hylands, T., Dellea, P., & Kamal‐Bahl, S. (2008). Indirect costs of obesity: a review of the current literature. Obesity Reviews. Oxford, UK: Blackwell Publishing Ltd. doi:10.1111/j.1467-789X.2008.00472.x
Vartanian, L., Spanos, S., Herman, C., & Polivy, J. (2015). Modeling of food intake: a meta-analytic review. Social Influence, 10(3), 1–18. doi:10.1080/15534510.2015.1008037 Vaterlaus, J., Patten, E., Roche, C., & Young, J. (2015). #Gettinghealthy: The perceived
influence of social media on young adult health behaviors. Computers in Human Behavior, 45, 151–157. doi:10.1016/j.chb.2014.12.013
Wang, M. B., & Jones, V. (2017). How Instagram content affects brand attitudes and behaviour. Journal of Digital & Social Media Marketing, 5(2), 175-188.
Werle, C., Trendel, O., & Ardito, G. (2013). Unhealthy food is not tastier for everybody: The “healthy=tasty” French intuition. Food Quality and Preference, 28(1), 116–121. doi:10.1016/j.foodqual.2012.07.007
Wood Baker, C., Little, T., & Brownell, K. (2003). Predicting Adolescent Eating and Activity Behaviors: The Role of Social Norms and Personal Agency. Health
Psychology, 22(2), 189–198. doi:10.1037/0278-6133.22.2.189 World Health Organisation. (2018a). Healthy Diet. Retrieved from
World Health Organization. (2018b). Obesity and overweight. Retrieved from http://www.who.int/mediacentre/factsheets/fs311/en/.
Xie, B., Gilliland, F., Li, Y., & Rockett, H. (2003). Effects of Ethnicity, Family Income, and Education on Dietary Intake among Adolescents. Preventive Medicine, 36(1), 30–40. doi:10.1006/pmed.2002.1131
